Avery's Neonatology PDF

Avery's Neonatology, 6th Edition
MacDonald, Mhairi G.;

Seshia, Mary M. K.;
Mullett, Martha D.
Copyright 2005 Lippincott Williams & Wilkins
Editors: MacDonald, Mhairi G.; Seshia, Mary M. K.; Mullett, Martha

D.
Title: Avery's Neonatology, 6th Edition
Copyright 2005 Lippincott Williams & Wilkins
2005
Lippincott Williams & Wilkins
Philadelphia
530 Walnut Street, Philadelphia, PA 19106 USA, http://www.LWW.com
0-7817-4643-4
1st Edition,1975 J.B. Lippincott
2nd Edition,1981 J.B. Lippincott
3rd Edition,1987 J.B. Lippincott
4th Edition,1994 J.B. Lippincott
5th Edition,1999 Lippincott William&Wilkins
All rights reserved. This book is protected by copyright. No part of this
book may be reproduced in any form by any means, including
photocopying, or utilized by any information storage and retrieval system
without written permission from the copyright owner, except for brief
quotations embodied in critical articles and reviews. Materials appearing
in this book prepared by individuals as part of their official duties as U.S.
government employees are not covered by the above-mentioned
copyright.
Printed in the USA
Library of Congress Cataloging-in-Publication Data
Avery's neonatology : pathophysiology & management of the newborn/
edited by Mhairi G. MacDonald, Mary M.K. Seshia, Martha D. Mullett.6th
ed.
p.; cm.
Rev. ed. of: Neonatology / edited by Gordon B. Avery, Mary Ann
Fletcher, Mhairi G. MacDonald. c1999.
Includes bibliographical references and index.

ISBN 0-7817-4643-4
1. Infants (Newborn) Diseases.
2. Neonatology.
I. Avery, Gordon B.
II. MacDonald, Mhairi G. III. Seshia, Mary M. K. IV. Mullett, Martha D. V.

Neonatology. VI. Title: Neonatology. [DNLM: 1. Infant, Newborn,
Diseases physiopathology. 2. Infant, Newborn, Diseases therapy. 3.
Prenatal Diagnosis. WS 421 A9556 2005]
RJ254.N46 2005
618.9201 dc22
2005004348
Care has been taken to confirm the accuracy of the information presented
and to describe generally accepted practices. However, the authors,
editors, and publisher are not responsible for errors or omissions or for
any consequences from application of the information in this book and
make no warranty, expressed or implied, with respect to the currency,
completeness, or accuracy of the contents of the publication. Application
of the information in a particular situation remains the professional
responsibility of the practitioner.
The authors, editors, and publisher have exerted every effort to ensure
that drug selection and dosage set forth in this text are in accordance
with current recommendations and practice at the time of publication.
However, in view of ongoing research, changes in government
regulations, and the constant flow of information relating to drug therapy
and drug reactions, the reader is urged to check the package insert for
each drug for any change in indications and dosage and for added
warnings and precautions. This is particularly important when the
recommended agent is a new or infrequently employed drug.
Some drugs and medical devices presented in the publication have Food
and Drug Administration (FDA) clearance for limited use in restricted
research settings. It is the responsibility of the health care provider to
ascertain the FDA status of each drug or device planned for use in their
clinical practice.
10 9 8 7 6 5 4 3 2 1
EDITED BY
Mhairi G. MacDonald M.B.Ch.B., F.R.C.P.(E.), F.R.C.P.C.H., D.C.H.
Professor of Pediatrics
George Washington University School of Medicine and Health Sciences,
Washington, District of Columbia
Mary M. K. Seshia M.B.Ch.B., F.R.C.P.(E.), F.R.C.P.C.H., D.C.H.

Professor of Pediatrics; Head
Section of Neonatology, Health Sciences Centre, University of Manitoba,
Winnipeg, Manitoba
Martha D. Mullett M.D., M.P.H.

Robert C. Byrd Health Sciences Center, West Virginia University,
Department of Pediatrics, Morgantown, West Virginia
Secondary Editors
Anne M. Sydor
Acquisitions Editor
Jenny Kim
Developmental Editor
Alicia Jackson
Project Manager
Benjamin Rivera
Senior Manufacturing Manager
Kathy Neely
Marketing Manager
Andrew Gatto
Cover Designer
Production Service: TechBooks
Printer: Quebecor World-Taunton
Contributors
Thomas E. Adrian Ph.D., F.R.C.Path.

Professor
Department of Biomedical Sciences Creighton University School of
Medicine, Omaha, Nebraska.
Chapter 40: Gastrointestinal Disease
Ruben E. Alvaro M.D., F.A.A.P.

Associate Professor
Department of Pediatrics University of Manitoba; Acting Head of
Neonatology, Department of Pediatrics, St. Boniface General Hospital,
Winnipeg, Manitoba, Canada.
Chapter 17: Cardiorespiratory Adjustments at Birth
Marianne S. Anderson
Associate Professor of Pediatrics and Neonatology
Perinatal Research Center, University of Colorado HSC, Denver, Colorado.
Chapter 26: Intrauterine Growth Restriction and the Small-ForGestational-Age Infant
Robert J. Arceci M.D., Ph.D.

Director and King Fahd Professor of Pediatric Oncology; Professor of
Pediatrics and Oncology
Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins,
Baltimore, Maryland.
Chapter 53: Neoplasia
John H. Arnold M.D.

Assistant Professor
Department of Anaesthesia (Pediatrics), Harvard Medical School; and
Senior Associate, Anesthesia and Critical Care, Children's Hospital,
Boston, Massachusetts.
Chapter 57: Anesthesia and Analgesia in the Neonate
Gordon B. Avery M.D., Ph.D.
Emeritus Professor of Pediatrics; Retired Chair

Department of Pediatrics, The George Washington University School of
Medicine and Health Sciences; and Chief Academic Officer, Emeritus,
Department of Pediatrics and Neonatology, Children's National Medical
Center, Washington, D.C.
Chapter 1: Neonatology: Past, Present and Future
Stephen Baumgart M.D.

Professor of Pediatrics Elect
George Washington University School of Medicine; Senior Staff Physician,
Department of Neonatology, Children's Hospital National Medical Center,
Washington, D.C.
Chapter 24: Thermal Regulation
David A. Beckman Ph.D.

Associate Professor
Department of Pediatrics, Thomas Jefferson University; and Nemours
Research Programs, Alfred I. duPont Hospital for Children, Wilmington,
Delaware.
Chapter 15: The Effects of Maternal Drugs on the Developing Fetus
Edward F. Bell M.D.

Professor
Department of Pediatrics, University of Iowa; and Director, Divison of
Neonatology, Children's Hospital of Iowa, Iowa City, Iowa.
Chapter 21: Fluid and Electrolyte Management
Joseph A. Bellanti M.D.

Professor
Department of Pediatrics and Microbiology-Immunology, Georgetown
University Medical Center, Washington, D.C.
Chapter 45: Immunology of the Fetus and Newborn
Corinne Benchimol D.O.

Clinical Assistant Professor
Department of Pediatrics, Mount Sinai School of Medicine; Clinical

Assistant Professor, Department of Pediatrics, Mount Sinai Hospital, New
York, New York.
Chapter 42: Renal Disease
Forrest C. Bennett M.D.

Professor
Department of Pediatrics, University of Washington School of Medicine,
Seattle, Washington.
Chapter 60: Developmental Outcome
Judy Bernbaum M.D.

Professor
Department of Pediatrics, University of Pennsylvania School of Medicine;
and Director, Neonatal Follow-Up Program, The Children's Hospital of
Philadelphia, Philadelphia, Pennsylvania.
Chapter 59: Medical Care After Discharge
Victor Blanchette M.D., B.Ch., F.R.C.P.

Professor
Department of Pediatrics, University of Toronto; and Chief, Division of
Hematology/Oncology, The Hospital for Sick Children, Toronto, Ontario,
Canada.
Chapter 46: Hematology
Carl L. Bose M.D.

Professor
Department of Pediatrics, University of North Carolina at Chapel Hill; and
Chief Division of Neonatal-Perinatal Medicine, UNC Hospitals, Chapel Hill,
North Carolina.
Chapter 4: Neonatal Transport
Michael J. Boyajian
Program director plastic and reconstructive surgery; Director craniofacial
surgery program
Division: plastic and reconstructive surgery, Children's National Medical

Center, Washington, D.C.
Chapter 44: Surgical Care of Conditions Presenting in the Newborn
Robert J. Boyle M.D.

University of Virginia Health System, Charlottesville, Virginia.
Chapter 2: Current Moral Priorities and Decision Making in NeonatalPerinatal Medicine
Robert L. Brent M.D., Ph.D., D.Sc.

Distinguished Professor of Pediatrics, Radiology, and Pathology
Department of Pediatrics, Jefferson Medical College, Philadelphia,
Pennsylvania; and Head, Clinical and Environmental Teratology
Laboratory, Nemours Research Programs, Alfred I. duPont Hospital for
Children, Wilmington, Delaware.
Luc P. Brion M.D.

Professor
Department of Pediatrics, Albert Einstein College of Medicine; and
Attending Neonatologist, Department of Pediatrics, Children's Hospital at
Montefiore, Bronx, New York.
Richard L. Bucciarelli M.D.

Associate Vice President for Health Affairs and Professor of Pediatrics
Department of Pediatrics, University of Florida; Neonatologist, Department of Pediatrics and Neonatology, Shands at University of Florida,
Gainesville, Florida.
Chapter 3: Neonatology in the United States: Scope and Organization
Dorothy I. Bulas M.D.

Professor of Pediatrics and Radiology
Departments of Radiology and Pediatrics, The George Washington
University Medical Center; and Director, Program in Diagnostic Imaging,

Division of Diagnostic Imaging, Children's National Medical Center,
Washington, D.C.
Chapter 12: Fetal Imaging: Ultrasound and Magnetic Resonance Imaging
Barbara K. Burton M.D.

Professor and Head
Division of Genetics and Metabolism, Department of Pediatrics, University
of Illinois College of Medicine; and Director, Center for Medical and
Reproductive Genetics, Michael Reese Hospital, Chicago, Illinois.
Chapter 41: Inherited Metabolic Disorders
Anthony J. Camerota M.D.

Department of Critical Care Medicine, Naval Regional Center, Portsmith,
Virginia.
Chapter 57: Anesthesia and Analgesia in the Neonate
Joyce Campbell R.N., M.S.

Service Director
Emergency Department Infection Control Practitioner, Children's National
Medical Center, Washington, D.C.
Chapter 49: Healthcare-Associated Infections
Anthony Chan M.B.B.S., F.R.C.P.C.

McMaster University, Hamilton, Canada; and Consultant, Coagulation
Laboratory, Department of Pediatric Laboratory Medicine, The Hospital for
Sick Children, Toronto, Canada.
Sukgi S. Choi M.D.

Associate Professor
Department of Otolaryngology and Pediatrics, The George Washington
University School of Medicine; and Vice Chairman, Department of
Otolaryngology, Children's National Medical Center, Washington, D.C.

John M. Costello M.D.

Instructor
Department of Pediatrics, Harvard Medical School, Division of Cardiac
Intensive Care, Department of Cardiology, Children's Hospital Boston,
Boston, Massachusetts.
Chapter 34: Preoperative and Postoperative Care of the Infant with
Critical Congenital Heart Disease
Gary L. Darmstadt M.D.

Assistant Director
Department of International Health, Bloomberg School of Public Health,
Johns Hopkins University, Baltimore, Maryland; and Senior Research
Advisor, Saving Newborn Lives Initiative, Office of Health, Save the
Children Federation-USA, Washington, D.C.
Chapter 55: Dermatologic Conditions
Jonathan M. Davis M.D.

Professor
Department of Pediatrics, State University of New York at Stony Brook
School of Medicine, Stony Brook; and Director, Neonatology; Director,
Cardiopulmonary Research Institute, Department of Pediatrics, WinthropUniversity Hospital, Mineola, New York.
Chapter 30: Bronchopulmonary Dysplasia
James G. H. Dinulos M.D.

Assistant Professor of Medicine and Pediatrics (Dermatology)
Children's Hospital at Dartmouth, Dartmouth Medical School, Lebanon,
New Hampshire.
Chapter 55: Dermatologic Conditions
Yigal Dror M.D.

Division of Hematology/Oncology and Bone Marrow Transplantation;
Director, Marrow Failure and Myelodysplasia Program, Scientist Track,

Research Institute, The Hospital for Sick Children, University of Toronto,
Canada.
Arie Drugan M.D.

Associate Professor
Department of Obstetrics and Gynecology, Wayne State University, Hutzel
Hospital, Detroit, Michigan; and Vice-Chairman, Depart-ment of
Obstetrics and Gynecology, Rambam Medical Center, Haifa, Israel.
Chapter 10: Prenatal Diagnosis in the Molecular Age Indications,
Procedures, and Laboratory Techniques
Chester M. Edelmann Jr. M.D.

Professor
Department of Pediatrics, Albert Einstein College of Medicine; Attending
Physician, Department of Pediatrics, Montefiore Medical Center, Bronx,
New York.
Martin R. Eichelberger M.D.

Professor
Department of Surgery and Pediatrics, The George Washington University
School of Medicine; and Director, Emergency Trauma Services, Children's
National Medical Center, Washington, D.C.
Mark I. Evans M.D.

Professor
Department of Obstetrics and Gynecology, Mount Sinai School of
Medicine; Director, Institute for Genetics, New York, New York.
Lynda B. Fawcett Ph.D.
Assistant Professor of Pediatrics

Department of Pediatrics, Jefferson Medical College, Philadelphia,
Pennsylvania; and Assistant Professor, Nemours Biomedical Research,
Alfred I. duPont Hospital for Children, Wilmington, Delaware.
Michael F. Flanagan M.D.

Department of Pediatrics, Section of Pediatrics Cardiology, Dartmouth
Medical School; and DartmouthHitchcock Medical Center, Lebanon, New
Hampshire.
Chapter 33: Cardiac Disease
Alan W. Flake M.D.

Department of Pediatric Surgery, Children's Hospital of Philadelphia,
Philadelphia, Pennsylvania.
Chapter 13: Fetal Therapy
John C. Fletcher Ph.D.

Deceased; Prior Professor Emeritus of Biomedical Ethics, Department of
Internal Medicine, University of Virginia School of Medicine,
Charlottesville, Virginia.
Chapter 2: Current Moral Priorities and Decision Making in NeonatalPerinatal Medicine
Mary Ann Fletcher M.D.

Clinical Professor of Pediatrics
The George Washington University School of Medicine and Health
Sciences, Washington, D.C.
Chapter 19: Physical Assessment and Classification
Joseph T. Flynn M.D., M.Sc.

Associate Professor of Clinical Pediatrics
Director, Pediatric Hypertension Program, Department of Pediatric
Nephrology, Montefiore Medical Center, New York, New York.
Wayne H. Franklin M.D., M.P.H.

Associate Professor
Department of Pediatrics, Feinberg School of Medicine at Northwestern
University, Division of Pediatric Cardiology, Children's Memorial Hospital,
Chicago, Illinois.
Chapter 34: Preoperative and Postoperative Care of the Infant with
Bishara J. Freij M.D.

Department of Pediatrics, Wayne State University School of Medicine; and
Chief, Division of Infectious Diseases, Department of Pediatrics, William
Beaumont Hospital, Royal Oak, Detroit, Michigan.
Chapter 47: Bacterial and Fungal Infections
Michael Friedman M.D.

Division of Neonatology, Stoneybrook University; and Senior Staff
Physician, Department of Neonatology, Stoneybrook University Hospital,
Stoneybrook, New York.
Chapter 24: Thermal Regulation
David M. Frim M.D., Ph.D.

Assistant Professor
Department of Surgery and Pediatrics, The University of Chicago; and
Chief, Section of Pediatric Neurosurgery, The University of Chicago
Children's Hospital, Chicago, Illinois.
Chapter 51: Neurosurgery of the Newborn
Michael K. Georgieff M.D.

Professor
Department of Pediatrics, University of Minnesota; and Staff
Neonatologist, Department of Pediatrics, Fairview-University Medical
Center, Minneapolis, Minnesota.

Chapter 22: Nutrition
Harold M. Ginzburg M.D., J.D., M.P.H.

Clinical Professor
Departments of Psychiatry and Neurology, Tulane Medical Center; and
Clinical Professor, Department of Psychiatry, Louisiana State University,
New Orleans, Louisiana.
Chapter 8: Law, Quality Assurance and Risk Management in the Practice
of Neonatology
Penny Glass Ph.D.

Associate Professor of Pediatrics
Department of Pediatrics, George Washington University School of
Medicine and Health Sciences; and Director, Child Development Program,
Department of Psychology, Children's National Medical Center,
Washington, D.C.
Chapter 9: The Vulnerable Neonate and the Neonatal Intensive Care
Environment
Sergio G. GoloMbek M.D.

Department of Pediatrics, Division of Neonatology, Westchester Medical
Center, New York Medical College, Valhalla, New York.
Chapter 31: Principles of Management of Respiratory Problems
Stanley N. Graven M.D.

Professor
Community and Family Health, Department of Pediatrics, University of
South Florida; and Staff Physician, Department of Pediatrics, Tampa
General Hospital, Tampa, Florida.
Chapter 6: Newborn Intensive Care Unit Design: Scientific and Practical
Considerations
Kathy Hamelin R.N. B.A, M.N., I.B.C.L.C.

Clinical Nurse Specialist; Lactation Consultant
Women's Health Program, Health Sciences Centre; Adjunct Professor,
Faculty of Nursing, University of Manitoba, Winnipeg, Manitoba, Canada.
Chapter 23: Breastfeeding and the Use of Human Milk in the Neonatal
Intensive Care Unit
Anne R. Hansen M.D.

Assistant Professor
Department of Pediatrics, Harvard Medical School; and Medical Director
Neonatal Intensive Care Unit, Department of Medicine, Children's Hospital
Boston, Boston, Massachusetts.
Michael R. Harrison M.D.

Department of Pediatric Surgery, University of California, San Francisco,
San Francisco, California.
Gary E. Hartman M.D.

Professor of Surgery and Pediatrics
Department of Surgery, The George Washington University School of
Medicine; and Chairman, Department of Pediatric Surgery, Children's
William W. Hay Jr. M.D.

Professor
Department of Pediatrics, University of Colorado Health Sciences Center,
Aurora, Colorado.
Chapter 26: Intrauterine Growth Restriction and the Small-ForGestational-Age Infant
Alan Hill M.D., Ph.D.

Professor
Department of Pediatrics, University of British Columbia; and Head,

Division of Neurology, Department of Pediatrics, British Columbia's
Children's Hospital, Vancouver, British Columbia, Canada.
Chapter 50: Neurological and Neuromuscular Disorders
Carl E. Hunt M.D.

Department of Pediatrics, Medical College of Ohio, Toledo, Ohio; and
Director, National Center on Sleep Disorders Research, National Heart,
Lung, and Blood Institute, Bethesda, Maryland.
Chapter 28: Control of Breathing: Development, Apnea of Prematurity,
Apparent Life-Threatening Events, Sudden Infant Death Syndrome
Nelson B. Isada M.D.

Alaska Perinatology Associates, Anchorage, Alaska.
Sherwin J. Isenberg M.D.

Lantz Professor of Pediatric Ophthalmology and Vice-Chairman
Department of Ophthalmology, Jules Stein Eye Institute, HarborUCLA
School of Medicine, Los Angeles, California.
Chapter 54: Eye Disorders
Mark P. Johnson M.D.

Associate Professor
Departments of Obstetrics/Gynecology, Pathology, Molecular Medicine,
and Genetics, Wayne State University; and Associate Director,
Reproductive Genetics, Department of Obstetrics and Gynecology, Hutzel
Hospital/The Detroit Medical Center, Detroit, Michigan.
Lauren A. Johnson-Robbins M.D.

Assistant Professor
Department of Pediatrics, Albany Medical College; and Attending
Neonatologist, Albany Medical Center, Albany, New York.
Chapter 27: Multiple Gestations
George W. Kaplan M.D., M.S., F.A.A.P., F.A.C.S.

Clinical Professor
Department of Surgery and Pediatrics, School of Medicine, University of
California, San Diego; and Chairman, Division of Urology, Children's
Hospital, San Diego, California.
Chapter 43: Structural Abnormalities of the Genitourinary Tract
Winston W. K. Koo M.B.B.S.

Professor
Department of Pediatrics, Wayne State University; and Neonatologist,
Department of Pediatrics, Hutzel Hospital, Detroit, Michigan.
Chapter 36: Calcium and Magnesium Homeostasis
Lajos Kovacs M.D.C.M., F.R.C.P.C., F.A.A.P.

Associate Professor
Department of Pediatrics; Faculty of Medicine, McGill University; and Staff
Neonatologist, Department of Neonatology, Sir Mortimer B. Davis Jewish
General Hospital, Montreal, Quebec, Canada.
Chapter 25: The Extremely Low-Birth-Weight Infant
Jukka K. Korpela M.D., Ph.D.

Adjunct Professor of Pediatrics
Department of Infectious Diseases, The George Washington University
Medical Center; and Faculty, Department of Infectious Diseases,
Children's National Medical Center, Washington, D.C.
Chapter 49: Healthcare-Associated Infections
Ian Laing M.D., F.R.C.P.C.H., F.R.C.P.E.

Clinical Director
Neonatal Unit, Simpson Centre for Reproductive Health, Royal Infirmary,
Edinburgh, Great Britain.
Chapter 20: General Care
HelaIn J. Landy M.D.

Professor and Chair
Department of Obstetrics and Gynecology, Georgetown University; and
Chief, Department of Obstetrics and Gynecology, Georgetown University
Hospital, Washington, D.C.
Chapter 14: The Impact of Maternal Illness on the Neonate
Mary M. Lee M.D.

Department of Pediatrics, Duke University Medical Center, Durham, North
Carolina.
Chapter 39: Endocrine Disorders of the Newborn
Judith Littleford M.D., B.Sc., F.R.C.P.C.

Associate Professor
Department of Anesthesia, University of Manitoba; and Section Head,
Women's Hospital of Anesthesia, Department of Anesthesia, Winnipeg,
Manitoba, Canada.
Chapter 16: Obstetric Anesthesia and Analgesia: Effects on the Fetus and
Newborn
Liz Loewen R.N., B.F.A., M.N.

Lecturer; Faculty of Nursing
University of Manitoba; and Network Researcher, MB Telehealth,
Winnipeg, Regional Health Authority, Winnipeg, Manitoba, Canada.
Chapter 5: Telehealth in Neonatology
Ralph A. Lugo M.D.

Associate Professor
Department of Pharmacotherapy, University of Utah College of Pharmacy,
Salt Lake City, Utah.
Chapter 56: Drug Therapy in the Newborn
Carolyn Houska Lund R.N., M.S.N, F.A.A.N.

Intensive Care Nursery, Children's Hospital, Oakland, California.
Chapter 7: Organization of Care and Quality in the NICU
George H. McCracken Jr. M.D.

Professor; The Sarah M. and Charles E. Seay Chair in Pediatric Infectious
Diseases
Department of Pediatrics, University of Texas, Southwestern Medical
Center; and Attending Physician, Children's Medical Center, Dallas, Texas.
Mhairi G. MacDonald M.B.Ch.B., F.R.C.P.(E.), F.R.C.P.C.H., D.C.H.

The George Washington University School of Medicine and Health
Sciences, Washington, D.C.
Chapter 8: Law, Quality Assurance and Risk Management in the Practice
of Neonatology
Joseph R. Madsen M.D.

Assistant Professor
Department of Surgery, Harvard Medical School; and Associate Professor,
Department of Neurosurgery, Children's Hospital, Boston, Massachusetts.
M. Jeffrey Maisels M.B., B.Ch.

Wayne State University School of Medicine, University of Michigan Medical
Center; and Chairman, Department of Pediatrics, William Beaumont
Hospital, Royal Oak, Michigan.
Chapter 35: Jaundice
Kathleen A. Marinelli M.D., T.R.C.L.C., F.A.B.M.

Department of Pediatrics, University of Connecticut School of Medicine,
Hartford, Connecticut.
Chapter 23: Breastfeeding and the Use of Human Milk in the Neonatal
Intensive Care Unit
Gilbert I. Martin M.D.

Clinical Professor
Department of Pediatrics, University of California, Irvine, Irvine; and
Director, Neonatal Intensive Care Unit, Citrus Valley Medical Center, West
Covina, California.
Chapter 6: Newborn Intensive Care Unit Design: Scientific and Practical
Considerations
Irene M. McAleer M.D.

Department of Surgery, School of Medicine, University of California, San
Diego; and Division of Urology, Children's Hospital, San Diego, California.
Chapter 43: Structural Abnormalities of the Genitourinary Tract
Scott Douglas McLean M.D.

Assistant Professor
Department of Pediatrics, Uniformed Services University of the Health
Sciences, Bethesda, Maryland; and Chief, Medical Genetics, San Antonio
Military Pediatrics Center, Brooke Army Medical Center, Ft. Sam Houston,
Texas.
Chapter 38: Congenital Anomalies
Thomas Moshang Jr. M.D.

Professor
Department of Pediatrics, University of Pennsylvania; and Chief,
Department of Pediatrics, Children's Hospital of Philadelphia, Philadelphia,
Pennsylvania.
Chapter 39: Endocrine Disorders of the Newborn
Martha D. Mullett M.D.

Robert C. Byrd Health Sciences Center, West Virginia University,
Department of Pediatrics, Morgantown, West Virginia.
Sarah C. Muttitt M.D., F.R.C.P.C., F.A.A.P., M.B.A.

Assistant Professor
Department of Pediatrics and Child Health, University of Manitoba; and
Director, MB Telehealth, Winnipeg RHA, Winnipeg, Manitoba, Canada.
Suhas M. Nafday M.D., M.r.C.P.I., F.A.A.P.

Assistant Professor
Attending Neonatologist, Section of Neonatology, Children's Hospital at
Montefiore, Bronx, New York.
Michael Narvey M.D., F.R.C.P.C.

Neonatal Fellow
Department of Pediatrics, University of Alberta, Edmonton, Alberta,
Canada.
Chapter 19: Physical Assessment and Classification
Kurt D. Newman M.D.

Departments of Surgery and Pediatrics, The George Washington
University School of Medicine, Department of Pediatric Surgery, Children's
Edward S. Ogata M.D., M.M.

Professor
Department of Pediatrics, Northwestern University Medical School; and
Chief Medical Officer, Children's Memorial Hospital, Chicago, Illinois.
Chapter 37: Carbohydrate Homeostasis
William Oh M.D.
Professor
Department of Pediatrics, Brown Medical School; Attending Neonatologist,
Department of Pediatrics, Women and Infants' Hospital, Providence,

Rhode Island.
Chapter 21: Fluid and Electrolyte Management
Enrique M. Ostrea Jr. M.D.

Professor
Department of Pediatrics, Wayne State University School of Medicine; and
Neonatologist, Department of Neonatology, Hutzel Women's Hospital,
Detroit, Michigan.
Chapter 58: The Infant of the Drug-Dependent Mother
Apostolos Papageorgiou M.D., F.R.C.P.C., F.A.A.P.

Professor
Departments of Pediatrics and Obstetrics and Gynecology, McGill
University; and Chief, Department of Pediatrics and Neonatology, Sir
Mortimer B. Davis Jewish General Hospital, Montreal, Quebec, Canada.
Ermelinda Pelausa M.D., F.R.C.P.C.

Assistant Professor of Pediatrics (Neonatology)
McGill University Faculty of Medicine; and Staff Neonatologist; Medical
Director (Neonatal Follow-up), Department of Neonatology, Sir Mortimer
B. Davis Jewish General Hospital, Montreal, Quebec, Canada.
Roderic H. Phibbs M.D.

Department of Pediatrics, University of California, San Francisco, San
Francisco, California.
Chapter 18: Delivery Room Management
J. Edgar Winston Cruz Posecion M.D.

Division of NeonatalPerinatal Medicine, Department of Pediatrics, Wayne
State University School of Medicine, Hutzel Hospital; and Children's
Hospital of Michigan, Detroit, Michigan.
Chapter 58: The Infant of the Drug-Dependent Mother
David M. Powell M.D.

Assistant Professor
Department of Surgery, The George Washington University School of
Medicine; and Attending Surgeon, Department of Pediatric Surgery,
Children's National Medical Center, Washington, D.C.
Richard Powers M.D.

Associate Neonatologist
Division of Neonatology, Children's Hospital and Research Center at
Oakland, Oakland, California.
Chapter 7: Organization of Care and Quality in the NICU
Gabriella Pridjian M.D.

Professor and Chairman; The Ernest A. and Elizabeth Miller-Robin Chair in
Obstetrics and Gynecology
Tulane University School of Medicine; and Assistant Dean, Lakeside
Hospital and Clinic; Adjunct Professor of Pediatrics, Clinical Geneticist in
Human Genetics Program, New Orleans, Louisiana.
Chapter 11: Feto-Maternal Interactions: Placental Physiology, the In
Utero Environment, and Fetal Determinants of Adult Disease
Gloria S. Pryhuber M.D.

Associate Professor of Pediatrics (Neonatology) and Environmental
Medicine Associate; and Director
Neonatal-Perinatal Fellowship Program, Golisano Children's Hospital at
Strong, Rochester, New York.
Chapter 29: Acute Respiratory Disorders
Yung-Hao Pung M.D., M.P.H.

Assistant Professor
Department of Pediatrics, Georgetown University School of Medicine,
Washington, D.C.
Virender K. Rehan M.D., M.R.C.P. (UK), M.R.C.P.I. (Dublin)

Associate Professor
Department of Pediatrics, David Geffen School of Medicine at UCLA; and
Director, Neonatal Intensive Care, Department of Pediatrics, Harbor UCLA
Medical Center, Torrance, California.
Chapter 18: Delivery Room Management
Mary E. Revenis M.D.

Department of Neonatology, The George Washington University School of
Medicine and Health Sciences, and Children's National Medical Center,
Washington, D.C.
Chapter 27: Multiple Gestations
Ward R. Rice M.D, Ph.D.

University of Cincinnati College of Medicine, Children's Hospital Medical
Center, Cincinnati, Ohio.
Chapter 29: Acute Respiratory Disorders
Henrique Rigatto M.D.

Professor
Department of Pediatrics, Physiology and Reproductive Medicine,
University of ManitobaWinnipeg; and Director of Neonatal Research,
Health Sciences Centre, Winnepeg, Manitoba, Canada.
Chapter 17: Cardiorespiratory Adjustments at Birth
William W. Robertson Jr. M.D.

Professor
Departments of Orthopaedic Surgery and Pediatrics, The George
Washington University School of Medicine, Washington, D.C.
Chapter 52: Orthopedics
Warren N. Rosenfeld M.D.

Professor
Department of Pediatrics, State University of New York at Stony Brook
School of Medicine, Stony Brook; and Chairman, Department of
Pediatrics, Winthrop-University Hospital, Mineola, New York.
Chapter 30: Bronchopulmonary Dysplasia
Lisa M. Satlin M.D.

Professor
Department of Pediatrics, Mount Sinai School of Medicine; Chief, Division
of Pediatric Nephrology, Department of Pediatrics, Mount Sinai Medical
Center, New York, New York.
Robert L. Schelonka M.D.

Assistant Professor
Department of Pediatrics, University of Alabama at Birmingham; and
Associate Medical Director, Regional Newborn Intensive Care Unit,
Division of Neonatology, University of Alabama at Birmingham,
Birmingham, Alabama.
Mary M. K. Seshia M.b.Ch.B., F.R.C.P.

Department of Pediatrics, University of Cincinnati College of Medicine; and
Director, Division of Pulmonary Biology, Children's Hospital Medical
Center, Cincinnati, Ohio.
Karen S. Wood M.D.

Assistant Professor
Department of Pediatrics, University of North Carolina at Chapel Hill;
Medical Director, Pediatric Transport, Carolina Air Care, UNC Hospitals,
Chapel Hill, North Carolina.
Chapter 4: Neonatal Transport
Yuval Yaron M.D.

Sourasly Medical Center, Tel Aviv, Israel.
Scott B. Yeager M.D.

Associate Professor
Department of Pediatrics, University of Vermont; and Chief, Division of
Pediatric Cardiology, Medical Center Hospital of Vermont, Burlington,
Vermont.
Chapter 33: Cardiac Disease
Terence L. Zach M.D.

Associate Professor
Department of Pediatrics, Creighton University, Omaha, Nebraska.
Chapter 40: Gastrointestinal Disease
Barbara J. Zeligs
Research Associate
Department of Pediatrics, Georgetown University School of Medicine; and
International Center for Interdisciplinary Studies of Immunology,
Washington, D.C.
Foreword
It was in 1972 that I began planning the first edition of Neonatology:
Pathophysiology and Management of the Newborn. In the preface of the
first edition, I wrote that Knowledge in this area has so expanded that it
now seems important to collect this material in a multi-author reference
work. The newborn is heir to so many problems, and his or her
physiology is so unique and rapidly changing, that all conditions in the
newborn should come within the concern of the new and expanding
discipline of neonatology. This thinking still guides the sixth edition,
which you hold in your hands.
In the meantime, science and practice in neonatology have grown so
rapidly that it is impossible for anyone to retain all the information that
this text contains. It is more an anthology, a collection of major state-ofthe-art papers, than a single book. We just do not have enough personal
disk space and RAM to encompass it all. Many neonatologists have
become sub-sub-specialists, while still retaining a broad view of the field.
We also have become adept at using remote information, via electronic
searches and consultation. Perhaps one of our all-too-frequent
weaknesses rests in failing to fulfill the role of the general practitioner
who puts it all together for individual families. In this respect, the chapter
on ethical considerations has been an important aspect of each edition. As
we go to publication, we mourn the death of John C. Fletcher, a pioneer
and leader in medical ethics and co-author of the current Ethical
Considerations chapter, who died May 27, 2004, at the age of 72. Dr.
Fletcher was an Episcopal priest a man of deep personal convictions who
was chief ethics officer at the NIH Clinical Center and a member of the
faculty of the Virginia Theological Seminary. He served on many NIH
committees as an ethicist, and he was a frequent consultant, author, and
lecturer on ethics. We have learned much from his wisdom and
compassion, and we offer him our admiration and respect.
There are 60 chapters in the current edition. All of those previously
published have been extensively rewritten. Eight chapters have new
authors, and there are three entirely new chapters regarding breast
feeding, infection control in the NICU, and telehealth. Something old,
something new, something borrowed, something blue, as the saying
goes. Concerns about thermoregulation, delivery room resuscitation,

nutrition, and infection have been there since the beginning. New
contributions regarding molecular genetics, imaging and monitoring
technology, and respiratory support are now important. Borrowed
information comes from experts in perinatal obstetrics, radiology and
imaging technology, pharmacology, laboratory medicine, cardiology,
surgery, and outcomes research. And something blue is still with us: the
critical problems of oxygenation attending delivery, lung immaturity,
cardiac malformations, and pulmonary hypertension.
This is my last hurrah with this textbook. I have long been out of the
nursery, even longer out of the laboratory, and I find life as an Emeritus
Professor pleasant indeed. But I still delight in the wonderful interplay of
science and art in the now-robust field of neonatology. Permit me, then, a
few predictions. We will devote more energy to understanding and
preventing preterm birth. We will find ways to deal with adverse
behaviors that result in poor outcomes for the neonate: smoking, alcohol,
drugs, poor prenatal care, inadequate nutrition, excess stress, prenatal
infections, and others not yet described. We will apply health outcomes
research to provide efficient feedback loops so that our best practices will
be more quickly adopted, and variation among NICUs will be reduced. We
will understand how to use our new molecular biology tools to modify
growth and aid healing and repair. We will become more adept at
neuroprotection and will intervene in destructive spirals such as
reperfusion injury to the brain. And most importantly, we will learn
wonderful and unexpected things about the plasticity of the brain and
nervous system. Geoffrey Raisman, a founder of the concept of
neuroplasticity, has said that current theory is the enemy of new scientific
discovery. We need to marvel at the feats of regeneration achieved by
salamanders, for one, and not rest until we understand which factors turn
on, and which inhibit, regrowth of neural tissue. If newts can do it, we
can do it. Enjoy!!
Preface
Early in the 1970s, the major medical postgraduate examining bodies in
the United Kingdom, the Royal College of Physicians and the Royal
College of Surgeons, recognized two specialties: adult medicine and
general surgery. Pediatrics could be practiced only after completion of
training and examination in adult medicine. The subspecialty of
neonatology did not yet exist, and newborns weighing less than 1500
grams at birth were considered unlikely to survive. In the United States,
subspecialty apprenticeships (fellowships) were available in neonatology;
however, the treatment provided to sick newborn infants was largely
extrapolated by miniaturization of that provided to adults and larger
children. Earlier misadventures with benign therapeutic agents, such as
oxygen, had resulted in calamitous complications, leading to the concept
that premature infants stood a greater chance of intact survival if
interventions were limited to the minimum.
In 1972, when the founding editor of Neonatology: Pathophysiology and
Management of the Newborn, Gordon B. Avery, M.D. Ph.D., was planning
the first edition, Dr. Mhairi MacDonald was in residency training in
Edinburgh, Scotland; Dr. Martha Mullett in Morgantown, West Virginia,
USA; Dr. Molly Seshia was practicing pediatrics in India.By the time that
Drs. MacDonald and Mary Ann Fletcher joined Dr. Avery as co-editors for
the fourth edition in the early 1990s, neonatology was an established
pediatric subspecialty in both the United Kingdom and the United States,
with a track record at once admirable for the pace of advancement of
knowledge and technology and cautionary for the significant incidence of
long-term morbidity in NICU survivors.
As the sixth edition goes to press, with two new editors, Drs. Mullett and
Seshia, it is clear that the optimal practice of neonatology continues to
require a careful balance between the aggressive use of modern
therapeutic technology and the prevention of unintended damage to a
fragile organism. The prefaces to previous editions reflect an ongoing
concern with ethical issues surrounding the increasing ability to treat
newborns previously considered nonviable.
The Accreditation Council for Graduate Medicine (ACGME) in the United
States has defined 6 general medical competencies that must be achieved
by trainees in any branch of clinical medicine and surgery; namely:

patient care (appropriate, effective, and compassionate), clinical science
(application of evolving biomedical and clinical science to patient care),
practice-based learning and improvement (ongoing investigation,
evaluation, and improvement of patient care practices), interpersonal
skills and communication (with patients, patients' families, and other
members of the healthcare team), professionalism (vocation, ethical
practice, self-evaluation, demeanor), and systems-based medicine
(having the awareness that all healthcare is provided in the context of a
larger system and the ability to effectively utilize system resources to
support the care of patients). We hope that the sixth edition of this text
will substantially assist those in training to acquire all six competencies.
ACGME emphasis on communication skills and systems-based medicine,
in particular, provides a timely reminder that the NICU team cannot exist
in a vacuum; sound decisions must be made with the assistance of
consultants in the community, and follow-up care for NICU graduates
must be planned and monitored as carefully as acute intensive care.
In this edition of Neonatology there are three new chapters and eight new
first authors. The addition of the chapter on telemedicine is based on
advances in knowledge and technology. The addition of the chapters on
breast feeding and infection control in the NICU are long overdue.
Mhairi G. MacDonald M.B.Ch.B., F.R.C.P.(E.), F.R.C.P.C.H., F.A.A.P., D.C.
H.
Martha D. Mullett M.D., M.P.H.
Mary M. K. Seshia M.B.Ch.B., F.R.C.P.(E.), F.R.C.P.C.H., D.C.H.
Preface to the First Edition

Neonatology means knowledge of the human newborn. The term was
coined by Alexander Schaffer, whose book on the subject, Diseases of the
Newborn, was first published in 1960. This book, together with Clement
Smith's Physiology of the Newborn Infant, formed cornerstones of the
developing field. In the past fifteen years, neonatology has grown from
the preoccupation of a handful of pioneers to a major subspecialty of
pediatrics. Knowledge in this area has so expanded that it now seems
important to collect this material into a multiauthor reference work.
Although the perinatal mortality rate has declined over the past fifty
years, the best presently attainable survival rates have not been achieved
throughout the world, and indeed the United States lags behind fifteen
other countries, despite its vast resources. New knowledge and
improvement in the coordination of services for mother and child are
needed to drive down perinatal mortality further. And finally, far greater
emphasis must be placed on morbidity, so that surviving infants can lead
full and productive lives. One hopes that in the future the yardstick of
success will be the quality of life and not the mere fact of life itself.
In this past decade, neonatology, as a recognized subspeciality of
pediatrics, has come into being around the intensive care-premature
nursery. Needless to say, the problems of prematurity are far from
solved. But neonatology is ripe for a broadening-out from its prematurityhyaline membrane disease beginnings. The newborn is heir to so many
problems, and his or her physiology is so unique and rapidly changing
that all conditions of the newborn should come within the concern of the
new and expanding discipline of neonatology. It has long since become
standard practice to admit to premature nurseries other high-risk infants
such as those of diabetic or toxemic mothers. Here the criterion is the
need for intensive care. However, the neonatologist's specialized
knowledge should give him a significant role in the care of other infants in
the first two to three months of life, whether they require intensive care
and whether they are readmitted for problems unrelated to prematurity
and birth itself. Detailed knowledge of newborn physiology can assist in
the management of congenital anomalies, surgical conditions of the
neonate, failure to thrive, nutritional problems, genetic, neurologic, and
biochemical diseases, and a host of conditions involving delayed

maturation. Thus one can conceive of a subspecialty sharply limited in
age to early infancy but broad in its study of the interaction of normal
physiology and disease processes.
Neonatology must also grow in its relationship to obstetrics and fetal
biology. In the best centers, an active partnership has developed between
obstetrics and pediatrics around the management of high-risk
pregnancies and newborns. Sometimes training has been cooperative, but
in only a few instances have basic scientists concerned with fetal biology
been brought into this effort. Important beginnings have been made in
studying the fetomaternal unit, such as the endocrine studies of Egon
Diczfalusy, the cardiopulmonary studies of Geoffrey Dawes, and the
immunologic studies of Arther Silverstein. But fundamental processes
such as the controls of fetal growth and the onset of labor are not
understood at this time. Centers or institutes bringing together workers of
diverse points of view are needed to wrestle with the profound problems
of fetal biology. At the clinical level, the interdependence of obstetrics and
neonatology is obvious. As an ultimate development, these two specialties
may one day be joined as a new entity-perinatology-at least at the level
of training and certification. In the meantime, far greater mutual
understanding and daily interaction are needed for the optimal care of
mothers and their infants.
This book is organized around problems as they occur, as well as by
organ systems. It hopes to achieve a balance between presentation of the
basic science on which rational management must rest, and the advice
concerning patient care which experts in each subarea are qualified to
give. Individual chapter authors have approached their subjects in various
ways, and no attempt has been made to achieve a completely uniform
format. In some instances, there is overlap of subject material, but the
somewhat different viewpoints presented, and the desire to spare the
reader from hopscotching through the book after cross references, have
persuaded me to leave small overlaps undisturbed.
It is appreciated that no volume such as this can have more than a finite
useful lifetime. Yet while its currency lasts, I hope it will serve as a
practical guide to therapy and an aid in the understanding of
pathophysiology for those active in the care of newborns.
Table of Contents
Part I - General Considerations
Chapter 1 - Neonatology: Past, Present and Future
PAST: THE ROOTS
PRESENT: THE TRUNK
THE BRANCHES
THE PRESENT
THE FUTURE
REFERENCES
Chapter 2 - Current Moral Priorities and Decision Making in Neonatal-Perinatal

Medicine
GOALS OF NEONATAL-PERINATAL MEDICINE
SOCIETAL AND PROFESSIONAL ISSUES
ETHICS AND MORAL DELIBERATION
CULTURAL AND RELIGIOUS DIVERSITY
HIV SCREENING OF PREGNANT WOMEN AND NEWBORN INFANTS
SUBSTANCE ABUSE IN PREGNANCY
GENETIC TESTING
MAKING DECISIONS REGARDING THE APPROPRIATENESS OF INTERINSTITUTIONAL TRANSFER
REFERENCES
Chapter 3 - Neonatology in the United States: Scope and Organization
THE CAPACITY TO DELIVER IN-HOSPITAL NEONATAL CARE
COSTS OF DELIVERING NEONATAL CARE
MAJOR MEASURES OF OUTCOME
MAJOR MORBIDITIES
REGIONALIZATION TODAY
REFERENCES
Chapter 4 - Neonatal Transport
HISTORY
ORGANIZATION AND ADMINISTRATION
TECHNICAL ASPECTS
DOCUMENTATION
QUALITY ASSURANCE
PSYCHOSOCIAL CONSIDERATIONS
LEGAL CONSIDERATIONS
BACK TRANSPORT
REFERENCES
Chapter 5 - Telehealth in Neonatology
THE HISTORY OF TELEHEALTH
TECHNOLOGY
MEDICO-LEGAL AND REGULATORY ISSUES
THE ECONOMICS AND EVALUATION OF TELEHEALTH
TELEHEALTH APPLICATIONS IN NEONATOLOGY
PRACTICAL STEPS IN GETTING STARTED
IMPLICATIONS FOR THE FUTURE
REFERENCES
Chapter 6 - Newborn Intensive Care Unit Design: Scientific and Practical

Considerations
DEVELOPING A MISSION STATEMENT
CREATING THE TEAMS
SITE VISITS
SPECIFIC DESIGN ISSUES
EQUIPMENT SELECTION
REVIEW AND APPROVAL PROCESS
HUMAN DYNAMICS CRUCIAL TO A SUCCESSFUL MOVE
CHANGE-IN-PLACE
REFERENCES
Chapter 7 - Organization of Care and Quality in the NICU
ORGANIZATION OF CARE
QUALITY IN THE NEONATAL INTENSIVE CARE UNIT
CONCLUSION
REFERENCES
Chapter 8 - Law, Quality Assurance and Risk Management in the Practice of

Neonatology
BASIC LEGAL AND REGULATORY CONCEPTS
QUALITY ASSURANCE, PERFORMANCE IMPROVEMENT, ANDRISK MANAGEMENT
CONCLUSIONS
REFERENCES
Chapter 9 - The Vulnerable Neonate and the Neonatal Intensive Care

Environment
NEONATAL SENSORY SYSTEMS: DEVELOPMENT, DISORDERS, ENVIRONMENT,

AND INTERVENTION
GENERAL PRINCIPLES
ACKNOWLEDGMENTS
REFERENCES
Part II - The Fetal Patient

Chapter 10 - Prenatal Diagnosis in the Molecular Age Indications,
GENE ORGANIZATION
ELEMENTS OF MOLECULAR GENETICS
PRENATAL DIAGNOSIS
REFERENCES
Chapter 11 - Feto-Maternal Interactions: Placental Physiology, the In Utero

Environment, and Fetal Determinants of Adult Disease
HUMAN PLACENTATION
PLACENTAL TRANSFER
PLACENTAL METABOLISM
PLACENTA AS AN ENDOCRINE ORGAN
AMNIOTIC FLUID
FETAL MEMBRANES
UMBILICAL CORD
PLACENTAL PHYSIOLOGY IN DISEASE STATES
THE IN UTERO ENVIRONMENT AND FETAL DETERMINANTS OF ADULT DISEASE
BIRTH-WEIGHT AND CORONARY HEART DISEASE
BIRTH-WEIGHT AND HYPERTENSION
BIRTH-WEIGHT AND DIABETES
REFERENCES
Chapter 12 - Fetal Imaging: Ultrasound and Magnetic Resonance Imaging
ASSESSMENT OF FETAL AGE, GROWTH, AND MATURITY
DETERMINING FETAL AGE
FETAL WEIGHT ESTIMATES
ASSESSMENT OF FETAL GROWTH
FETAL ANATOMY
DETECTION OF FETAL ASPHYXIA
DOPPLER ASSESSMENT OF BLOOD VELOCITIES IN UMBILICAL AND FETAL

VESSELS
SUMMARY
REFERENCES
Chapter 13 - Fetal Therapy
SURGICAL THERAPY
ENDOCRINE DISORDERS
INBORN ERRORS OF METABOLISM
PRENATAL HEMATOPOIETIC STEM CELL TRANSPLANTATION
CONCLUSION
References
Chapter 14 - The Impact of Maternal Illness on the Neonate
PRETERM DELIVERY
MATERNAL NUTRITION
MATERNAL ILLNESSES
SUMMARY
REFERENCES
Chapter 15 - The Effects of Maternal Drugs on the Developing Fetus
CHARACTERIZATION OF ADVERSE REPRODUCTIVE OUTCOMES
FACTORS THAT AFFECT SUSCEPTIBILITY TO THE DELETERIOUS EFFECTS OF

DRUGS
ACKNOWLEDGMENTS
REFERENCES
Chapter 16 - Obstetric Anesthesia and Analgesia: Effects on the Fetus and

Newborn
EVALUATION OF WELL-BEING
PAIN MANAGEMENT
ANALGESIC TECHNIQUES FOR LABOR: EFFECTS ON THE FETUS AND NEWBORN
SURGERY DURING PREGNANCY
RESUSCITATION ISSUES DURING PREGNANCY
ANESTHETIC TECHNIQUES FOR CESAREAN SECTION: EFFECTS ON THE FETUS

AND NEWBORN
POSTPARTUM PAIN MANAGEMENT
CONCLUSION
ACKNOWLEDGEMENT
REFERENCES
Part III - Transition and Stabilization

Chapter 17 - Cardiorespiratory Adjustments at Birth
PULMONARY ADAPTATION
RESPIRATORY ADAPTATION
CIRCULATORY ADAPTATION
CONCLUSION
REFERENCES
Chapter 18 - Delivery Room Management
PATHOPHYSIOLOGY OF INTRAPARTUM ASPHYXIA AND RESUSCITATION
HIGH-RISK PREGNANCIES
RESUSCITATION OF THE ASPHYXIATED INFANT
DISCONTINUATION OF RESUSCITATION
SPECIAL PROBLEMS
ACKNOWLEDGMENTS
REFERENCES
Chapter 19 - Physical Assessment and Classification
NEWBORN HISTORY
GESTATIONAL AGE ASSESSMENT
GROWTH
EXAMINATION
REFERENCES
Chapter 20 - General Care
ANTENATAL
INTRAPARTUM
NEONATAL
SPECIFIC CONGENITAL INFECTIONS
POST-DISCHARGE
REFERENCES
Chapter 21 - Fluid and Electrolyte Management
BODY COMPOSITION OF THE FETUS AND NEWBORN INFANT
INSENSIBLE WATER LOSS
NEUROENDOCRINE CONTROL OF FLUID AND ELECTROLYTE BALANCE
RENAL FUNCTION IN RELATION TO FLUID AND ELECTROLYTE THERAPY
PRINCIPLES OF FLUID AND ELECTROLYTE THERAPY
ACID-BASE BALANCE
FLUID AND ELECTROLYTE PROBLEMS ASSOCIATED WITH SPECIFIC CLINICAL

CONDITIONS
TECHNICAL ASPECTS OF PARENTERAL FLUID THERAPY
REFERENCES
Chapter 22 - Nutrition
NUTRITIONAL CAPABILITIES OF THE NEWBORN INFANT
NUTRIENT REQUIREMENTS FOR TERM AND PRETERM INFANTS
NUTRIENT DELIVERY
WHAT TO FEED NEWBORN INFANTS
NUTRITIONAL MONITORING
REFERENCES
Chapter 23 - Breastfeeding and the Use of Human Milk in the Neonatal

Intensive Care Unit
OVERVIEW
BACKGROUND
INCIDENCE OF BREASTFEEDING OR PROVISION OF HUMAN MILK
ADVANTAGES OF HUMAN MILK FOR PREMATURE AND COMPROMISED BABIES
CHALLENGES TO THE PROVISION OF HUMAN MILK
THE USE OF MOTHER'S MILK IN NEONATAL INTENSIVE CARE
FORTIFICATION OF HUMAN MILK FORVERY LOW BIRTH WEIGHT BABIES
THE DEVELOPMENTAL PROGRESSION TOWARD BREASTFEEDING
ONGOING BREASTFEEDING SESSIONS
DISCHARGE PLANNING
SUPPORTING BREASTFEEDING IN THE NEONATAL INTENSIVE CARE UNIT AND

AFTER DISCHARGE
SPECIAL CIRCUMSTANCES
CONTRAINDICATIONS TO BREAST FEEDING
MEDICATIONS AND DRUGS
CONCLUSION
REFERENCES
Chapter 24 - Thermal Regulation
A HISTORICAL PERSPECTIVE
THERMAL BALANCE AT THE BEGINNING OF LIFE
CONVECTION-WARMED INCUBATORS
THE RADIANT WARMER BED
HEAT SHIELDING
EPIDERMAL BARRIER PROTECTION
HYBRID INCUBATOR/RADIANT WARMER DESIGN
FEVER
PHYSIOLOGIC CONTROL OF BODY TEMPERATURE
MECHANISMS PRODUCING NEONATAL FEVER
REFERENCES
Part IV - The Low-Birth-Weight Infant

Chapter 25 - The Extremely Low-Birth-Weight Infant
EPIDEMIOLOGY
PERINATAL MANAGEMENT
FLUIDS AND ELECTROLYTES
MAJOR MORBIDITIES OF THE EXTREMELY LOW-BIRTH-WEIGHT INFANT
FOLLOW-UP OF THE EXTREMELY LOW BIRTH WEIGHT INFANT
FUTURE DIRECTIONS AND ETHICAL ISSUES
ACKNOWLEDGMENTS
REFERENCES
Chapter 26 - Intrauterine Growth Restriction and the Small-For-GestationalAge Infant
INTRODUCTION
DEFINITIONS
INTERPRETATION OF FETAL GROWTH CURVES
INTRAUTERINE GROWTH RESTRICTION AND PRETERM BIRTH
FETAL GROWTH
GROWTH OF BODY COMPONENTS IN THE FETUS
REGULATION OF FETAL GROWTH
FETAL NUTRIENT UPTAKE AND METABOLISM AND REGULATION OF FETAL

GROWTH
ANTENATAL CARE OF THE INTRAUTERINE GROWTH-RESTRICTED FETUS
CLINICAL EVALUATION AND TREATMENT OF THE SMALL-FOR-GESTATIONAL AGE

INFANT
CLINICAL PROBLEMS OF THE SMALL-FOR-GESTATIONAL-AGE NEONATE (
OUTCOMES AND LONG-TERM CONSEQUENCES OF SMALL-FOR-GESTATIONALAGE INFANTS
REFERENCES
Chapter 27 - Multiple Gestations
EPIDEMIOLOGY
IMPACT OF REPRODUCTIVE TECHNOLOGY
ZYGOSITY
PLACENTATION
ANTEPARTUM COMPLICATIONS
ANTENATAL MANAGEMENT
LABOR AND DELIVERY
MORTALITY
TWIN-TO-TWIN TRANSFUSION SYNDROME
STUCK TWIN
ASPHYXIA
GROWTH
CONGENITAL ANOMALIES
NEONATAL DISORDERS
POSTNEONATAL CARE AND FOLLOW-UP
REFERENCES
Part V - The Newborn Infant

Chapter 28 - Control of Breathing: Development, Apnea of Prematurity,
Apparent Life-Threatening Events, Sudden Infant Death Syndrome
NORMAL CONTROL OF BREATHING
CLINICAL DISORDERS IN CONTROL OF BREATHING
IDIOPATHIC APPARENT LIFE-THREATENING EVENTS
SUDDEN INFANT DEATH SYNDROME
REFERENCES
Chapter 29 - Acute Respiratory Disorders
HUMAN LUNG DEVELOPMENT
DEVELOPMENTAL ANOMALIES
THE SURFACTANT SYSTEM
RESPIRATORY DISTRESS SYNDROME
MECONIUM ASPIRATION SYNDROME
PERSISTENT PULMONARY HYPERTENSION OF THE NEWBORN
AIR LEAKS
REFERENCES
Chapter 30 - Bronchopulmonary Dysplasia
DEFINITION AND INCIDENCE
PATHOGENESIS
PATHOPHYSIOLOGIC CHANGES
MANAGEMENT
OUTCOME
PREVENTION STRATEGIES
REFERENCES
Chapter 31 - Principles of Management of Respiratory Problems
OXYGEN THERAPY
ASSESSMENT OF GAS EXCHANGE
STRATEGIES FOR RESPIRATORY SUPPORT
MANAGEMENT IN THE INTENSIVE CARE SETTING
REFERENCES
Chapter 32 - Extracorporeal Membrane Oxygenation
INDICATIONS
PROCEDURE
EQUIPMENT AND SYSTEMS
PATIENT MANAGEMENT
SUMMARY
REFERENCES
Chapter 33 - Cardiac Disease
Incidence
Infant Mortality
Long-Term Survival
Etiology
Fetal Cardiology
Fetal Cardiac Intervention
Prematurity
Recognition of Clinical Features
Diagnostic Tools
Diagnostic Cardiac Catheterization and Angiography
Management Procedures for Severe Cardiac Disease
Cyanotic Lesions
Acyanotic Lesions
Acyanotic Anomalies with Abnormal Cardiac Function or Structure
Arrhythmias
Sinus Arrhythmia
Accelerated Ventricular Rhythm
Junctional Ectopic Tachycardia
Atrioventricular Block
Second-Degree Atrioventricular Block
REFERENCES
Chapter 34 - Preoperative and Postoperative Care of the Infant with Critical

Congenital Heart Disease
INTRODUCTION
PREOPERATIVE CARE
PREOPERATIVE PHYSIOLOGY AND MANAGEMENT: SPECIFIC LESIONS
POSTOPERATIVE CARE
CONCLUSION
REFERENCES
Chapter 35 - Jaundice
Introduction
Formation, Structure, and Properties of Bilirubin
Fetal Bilirubin Metabolism
Neonatal Bilirubin Metabolism
Physiologic Mechanisms of Neonatal Jaundice
Physiologic Jaundice
Bilirubin Toxicity
Clinical Features of Bilirubin Encephalopathy
Clinical Sequelae of Hyperbilirubinemia
The Clinical Approach to the Jaundiced Newborn
Jaundice in the Healthy Newborn
Pathologic Causes of Jaundice
Newer Developments that Affect Our Approach to the Jaundiced Newborn
Prevention, Identification, and Management of Neonatal Hyperbilirubinemia
Laboratory Measurements of Bilirubin
Preventing Extreme Hyperbilirubinemia and Kernicterus
Treatment
Phototherapy
Exchange Transfusion
Pharmacologic Treatment
Physiologic Role of Bilirubin
Acknowledgment
REFERENCES
Chapter 36 - Calcium and Magnesium Homeostasis
TISSUE DISTRIBUTION
CIRCULATING CONCENTRATION
PHYSIOLOGIC CONTROL
NONCLASSIC CONTROL OF CALCIUM AND MAGNESIUM HOMEOSTASIS
DISTURBANCES IN SERUM MINERAL CONCENTRATIONS
SKELETAL MANIFESTATIONS OF DISTURBED MINERAL HOMEOSTASIS
REFERENCES
Chapter 37 - Carbohydrate Homeostasis
MATERNAL METABOLISM DURING PREGNANCY
DEVELOPMENT OFGLUCOSE-PRODUCING AND GLUCOREGULATORYCAPABILITIES

IN THE FETUS
NEONATAL GLUCOSE REQUIREMENTS
DIAGNOSIS AND TREATMENT
CONSEQUENCES
REFERENCES
Chapter 38 - Congenital Anomalies
DEFINITIONS AND CLASSIFICATIONS
MANAGEMENT STRATEGY
SELECTED EXAMPLES
GENETIC COUNSELING
REFERENCES
Chapter 39 - Endocrine Disorders of the Newborn
DISORDERS OF SEXUAL DIFFERENTIATION
DISORDERS OF CHROMOSOMAL SEX
DISORDERS OF GONADAL DETERMINATION
DISORDERS OF PHENOTYPIC SEX
OTHER CONDITIONS INVOLVING GENITOURINARY DEVELOPMENT
EVALUATION
DISORDERS OF THE HYPOTHALAMUS AND PITUITARY
DISORDERS OF THE ANTERIOR PITUITARY
DISORDERS OF THE POSTERIOR PITUITARY
DISORDERS OF THE ADRENAL GLAND
ADRENAL INSUFFICIENCY
DISORDERS OF THE THYROID
THYROID FUNCTION TESTS
RESISTANCE TO THYROID HORMONES
UNRESPONSIVENESS TO THYROID-STIMULATING HORMONE
SECONDARY AND TERTIARY HYPOTHYROIDISM
SYMPTOMS OF HYPOTHYROIDISM
DIAGNOSIS
TREATMENT AND PROGNOSIS
EUTHYROID SICK SYNDROME
CONSUMPTIVE HYPOTHYROIDISM SECONDARY TO GIANT HEMANGIOMA
CONGENITAL THYROTOXICOSIS
REFERENCES
Chapter 40 - Gastrointestinal Disease
DEVELOPMENT OF THE GASTROINTESTINAL TRACT
GASTROINTESTINAL HORMONES AND ENTERIC NEUROPEPTIDES
ABNORMALITIES OF THE GASTROINTESTINAL TRACT
REFERENCES
Chapter 41 - Inherited Metabolic Disorders
CLINICAL MANIFESTATIONS OF INBORN ERRORS OF METABOLISM
HYPERAMMONEMIA
METABOLIC ACIDOSIS
EMERGENCY TREATMENT OF THE INFANT WITH AN ACUTE METABOLIC

ENCEPHALOPATHY
HYPOGLYCEMIA
JAUNDICE AND LIVER DYSFUNCTION
FINDINGS SUGGESTIVE OF A STORAGE DISEASE
ABNORMAL ODOR
DYSMORPHIC FEATURES
ABNORMAL EYE FINDINGS
SAMPLES TO OBTAIN FROM A DYING CHILD WITH A SUSPECTED INBORN ERROR

OF METABOLISM
NEWBORN SCREENING FOR INHERITED METABOLIC DISORDERS
MATERNAL METABOLIC DISORDERS
REFERENCES
Chapter 42 - Renal Disease
Developmental Physiology
Clinical Evaluation of Renalfunction and Disease
Physical Examination
Acute Renal Failure and Oligoanuria
Nephrotoxicity
Nonsteroidal Antiinflammatory Agents
Management of CKD in Infancy
Summary
Hypertension
Bacteriuria and Urinary Tract Infections
Tubular Dysfunction
Congenital Nephrotic Syndrome
REFERENCES
Chapter 43 - Structural Abnormalities of the Genitourinary Tract
IMAGING
ANOMALIES OF THE KIDNEY
ANOMALIES OF THE URETERS AND BLADDER
GENITAL ABNORMALITIES
URINARY TRACT INFECTION
UROLOGIC ASPECTS OF MYELODYSPLASIA
PRUNE-BELLY SYNDROME
UROLOGIC IMPLICATIONS OF IMPERFORATE ANUS
FEMALE GENITAL ABNORMALITIES
ACKNOWLEDGMENT
REFERENCES
Chapter 44 - Surgical Care of Conditions Presenting in the Newborn
LESIONS OF THE HEAD AND NECK
MASSES IN THE NECK
UPPER AIRWAY OBSTRUCTION CAUSING RESPIRATORY DISTRESS
THORACIC LESIONS CAUSING RESPIRATORY DISTRESS
LESIONS OF THE ESOPHAGUS
ABDOMINAL SURGERY
DISORDERS OF THE GENITALIA
OBSTRUCTIVE JAUNDICE
HERNIA AND HYDROCELE
ABNORMALITIES OF THE UMBILICUSAND ABDOMINAL WALL
UMBILICAL GRANULOMA
SACROCOCCYGEAL TERATOMA
VASCULAR ACCESS
REFERENCES
Chapter 45 - Immunology of the Fetus and Newborn
DEVELOPMENT OF THE IMMUNE SYSTEM
INFLAMMATORY RESPONSE
CELLULAR COMPONENTS
POLYMORPHONUCLEAR
PHAGOCYTOSIS
MONOCYTES AND MACROPHAGES
HUMORAL COMPONENT
FIBRONECTIN ANDADHESION MOLECULES
CYTOKINES
ANTIBODIES
OPSONIC CAPACITY
ADAPTIVE IMMUNE MECHANISMS
HEMATOPOIETIC DIFFERENTIATION
LYMPHOPOIETIC DIFFERENTIATION
T-CELL SYSTEM
T CELLS
B-CELL SYSTEM
MATERNAL, FETAL, AND NEONATAL INTERACTIONS
IMMUNOLOGIC CONSEQUENCES OF INTRAUTERINE INFECTIONS
IMMUNOLOGIC EVALUATION
NEONATAL IMMUNE RESPONSE TO HUMAN IMMUNODEFICIENCY VIRUS
EVALUATION OF THE HUMORAL IMMUNE SYSTEM
EVALUATION OF THE CELL-MEDIATED IMMUNE SYSTEM
IMMUNOLOGIC THERAPY
PREVENTION OF INFECTION IN LOW BIRTH WEIGHT INFANTS
TREATMENT OF PRESUMED NEONATAL INFECTION
GROUP B STREPTOCOCCAL INFECTION
SPECIFIC ANTIBODY REPLACEMENT THERAPY
SUMMARY OF RECENT CONSIDERATIONS CONCERNING THE USE OF

INTRAVENOUS IMMUNOGLOBULIN IN NEONATAL SEPSIS
ACTIVE IMMUNIZATION OF PREGNANT WOMEN
USE OF RECOMBINANT HUMAN GRANULOCYTE-COLONY STIMULATING FACTOR

OR GRANULOCYTE-MACROPHAGE COLONY-STIMULATING FACTOR
PLASMA COMPONENT TRANSFUSION
PNEUTROPHIL TRANSFUSION THERAPY
REFERENCES
Chapter 46 - Hematology
Anemias
Bleeding Disorders
Platelet Disorders
Leukocytes
Transfusion of Blood and Blood Products
Thrombosis and Embolism
Acknowledgments
REFERENCES
Chapter 47 - Bacterial and Fungal Infections
EPIDEMIOLOGY
PHARMACOLOGIC BASIS OF ANTIMICROBIAL THERAPY
SEPSIS NEONATORUM
SYPHILIS
BACTERIAL MENINGITIS
OSTEOMYELITIS AND SEPTIC ARTHRITIS
CUTANEOUS INFECTIONS
URINARY TRACT INFECTION
NEONATAL OPHTHALMIA
DIARRHEAL DISEASE
LOWER RESPIRATORY TRACT INFECTION
OTITIS MEDIA
PERITONITIS
REFERENCES
Chapter 48 - Viral and Protozoal Infections
Introduction
Rubella
Herpes Simplex Virus Infections
Cytomegalovirus Infections
Varicella
Other Herpesvirus Infections
Parvovirus B19 Infections
Hepatitis Virus Infections
Lymphocytic Choriomeningitis Virus
West Nile Virus
Human Immunodeficiency Virus Type 1 Infection
Toxoplasmosis
Miscellaneous Viral and Protozoal Pathogens
REFERENCES
Chapter 49 - Healthcare-Associated Infections
EPIDEMIOLOGY
HEALTHCARE-ASSOCIATED INFECTION RATE
LATE-ONSET SEPSIS
VENTILATOR-ASSOCIATED PNEUMONIA
URINARY TRACT INFECTIONS
INFECTIONS DUE TO MULTI-DRUG RESISTANT ORGANISMS
VIRAL INFECTIONS
INFECTION CONTROL MEASURES
REFERENCES
Chapter 50 - Neurological and Neuromuscular Disorders
NEUROLOGIC EVALUATION OF THE NEWBORN
NEONATAL SEIZURES
HYPOXIC-ISCHEMIC CEREBRAL INJURY
MAJOR NEUROPATHOLOGIC PATTERNS OF INJURY
INTRACRANIAL HEMORRHAGE
SUBDURAL HEMORRHAGE
PRIMARYSUBARACHNOID HEMORRHAGE
INTRACEREBELLAR HEMORRHAGE
INTRAVENTRICULAR HEMORRHAGE IN THE TERM NEWBORN
BRAIN DEATH IN NEONATES
NEUROMUSCULAR DISORDERS
PERIPHERAL NERVE DISORDERS
DISORDERS OF THENEUROMUSCULAR JUNCTION
REFERENCES
Chapter 51 - Neurosurgery of the Newborn
THE PATHOPHYSIOLOGY OF NEONATAL NEUROSURGERY
FLUID COLLECTIONS AND THEIR MANAGEMENT
MANAGEMENT OF THE INFANT WITH AN OPEN DEFECT OF THE NEURAL AXIS
REMOVAL OF EXCESS INTRACRANIAL MASS
OPENING PREMATURE FUSIONS IN THE NEONATE
CEREBROVASCULAR ANOMALIES OF THE NEWBORN
PROGNOSIS AND LONG-TERM OUTLOOK FOR THE NEUROSURGICAL NEONATE
REFERENCES
Chapter 52 - Orthopedics
PHYSICAL EXAMINATION
MUSCULOSKELETAL ANOMALIES
GENERALIZED MUSCULOSKELETAL ANOMALIES
BIRTH FRACTURES
OBSTETRIC PALSY
BONE AND JOINT INFECTIONS
REFERENCES
Chapter 53 - Neoplasia
EPIDEMIOLOGY
DEVELOPMENTAL GROWTH DISTURBANCES, GENETIC ABERRATIONS, AND

CANCER PATHOGENESIS
EXPOSURE TO MATERNAL MALIGNANCY
PRENATAL EXPOSURE TO MATERNAL GENOTOXINS
TUMORS OF NEUROEPITHELIAL ORIGIN
CONGENITAL LEUKEMIA
NEOPLASMS OF THE KIDNEY
TUMORS OF GERM CELL ORIGIN
ENDODERMAL SINUS OR YOLK SAC TUMORS
PRIMARY HEPATIC NEOPLASMS
SOFT TISSUE SARCOMAS
VASCULAR NEOPLASMS AND MALFORMATIONS
HISTIOCYTOSES
THERAPEUTIC ISSUES AND LATE EFFECTS OF THERAPY
REFERENCES
Chapter 54 - Eye Disorders
GENERAL CONSIDERATIONS
EXAMINATION TECHNIQUES
ACQUIRED DISORDERS
TRAUMA
PHOTOTHERAPY FOR HYPERBILIRUBINEMIA
RETINOPATHY OF PREMATURITY
COMMON DIFFERENTIALDIAGNOSTIC PROBLEMS
REFERENCES
Chapter 55 - Dermatologic Conditions
INTRODUCTION
SKIN DEVELOPMENT, STRUCTURE, AND FUNCTION
EXAMINATION OF NEWBORN SKIN
SCALY RASHES
VESICLES AND PUSTULES
DISORDERS OF PIGMENTATION (BROWN SPOTS/WHITE SPOTS)
BROWN SPOTS
ATROPHIC LESIONS
TRANSIENT VASCULAR LESIONS
VASCULAR GROWTHS
CYSTS AND NODULES
EDEMA
PURPURA
YELLOW-ORANGE LESIONS
REFERENCES
Part VI - Pharmacology
Chapter 56 - Drug Therapy in the Newborn
PRINCIPLES OF PHARMACOLOGY APPLIED TO NEONATES
PHARMACOKINETICS
CLINICAL TOXICOLOGY
SPECIFIC CLASSES OF DRUGS
DRUG EXCRETION IN BREAST MILK
REFERENCES
Chapter 57 - Anesthesia and Analgesia in the Neonate
PAIN PERCEPTION
ANESTHESIA
ANALGESIA
NONPHARMACOLOGIC TECHNIQUES
SEDATION
CLONIDINE
SUMMARY
REFERENCES
Chapter 58 - The Infant of the Drug-Dependent Mother
EPIDEMIOLOGY
NARCOTICS
HISTORY
ANTENATAL PROBLEMS
NEONATAL PROBLEMS
NONNARCOTIC HYPNOSEDATIVES
BARBITURATES
CHLORDIAZEPOXIDE AND DIAZEPAM
OTHER HYPNOSEDATIVES
DIFFERENTIAL DIAGNOSIS
COCAINE
ALCOHOL
MARIJUANA
NICOTINE AND SMOKING
PHENCYCLIDINE
AMPHETAMINES
CLUB DRUGS (ECSTASY, GHB, KETAMINE)
CAFFEINE
DIAGNOSIS OF DRUG EXPOSURE
TREATMENT
DRUG-ADDICTED WOMEN AS INFANT CAREGIVERS
FOLLOW-UP
REFERENCES
Part VII - Beyond the Nursery

Chapter 59 - Medical Care After Discharge
EXPECTATIONS OF GROWTH
FEEDING PROBLEMS
IMMUNIZATIONS
SPECIALIZED CARE
DISCHARGE PLANNING
REFERENCES
Chapter 60 - Developmental Outcome
ORGANIZATION OF A HIGH-RISK INFANT FOLLOW-UP PROGRAM
NEURODEVELOPMENTAL OUTCOME OF LOW-BIRTH-WEIGHT PREMATURE

INFANTS
SPECIFIC PERINATAL-NEONATAL COMPLICATIONS IN HIGH-RISK INFANTS
REFERENCES
Appendices
Appendix A: The Healthcare Matrix
Appendix B: HemoglobinOxygen Dissociation Curves
Appendix C: Breast Milk in the NICU
Appendix D: Growth Parameters
Appendix E: Jaundice
Appendix F: Blood Pressure
Appendix G: Renal and Urinary Tract Anomalies
Appendix H: Drug Formulary
Appendix I: Nutrients Values
Appendix J: Technical Procedures
Color Plate
Index
Chapter 1
Neonatology: Past, Present and Future
Gordon B. Avery
Rapid change has characterized neonatology since the name was coined in 1960 by Alexander
Schaffer. Structurally, it can be compared with a tree (Fig. 1-1). Its rootsobstetrics, pediatrics,
and physiologybegan at the turn of the century. A sturdy trunk has developed in the intensive
care nurseries (ICNs) scattered across the United States and around the world. The branches
have spread so widely that it is difficult for a single person to be expert in all the areas of activity
required for a tertiary neonatology service. Important interactions have gone beyond allied
disciplines such as obstetrics, anesthesiology, cardiology, radiology, and surgery. Neonatologists
today struggle with hospital administrators, pediatric training program directors, legislatures,
Congress, the courts, the federal government, malpractice lawyers, right-to-life groups, and
ethicists in an effort to determine their proper roles and limits. Caught in a cross-fire between
strenuous cost-containment measures and regulations mandating the vigorous treatment of all
newborns regardless of prognosis, many neonatologists wonder when a stable situation will be
reached. Yet stimulating growth has occurred, mainly since 1960.
PAST: THE ROOTS

One of the main roots from which neonatology grew was supplied by obstetricians such as Pierre
Budin and Sir Dugald Baird, who were interested in the babies they delivered and not merely in
the immediate welfare of the mother. It may be a serious oversimplification to imply that, in
former times, obstetricians were content if the baby was born alive. Childbirth, however, was the
cause of a significant number of maternal deaths and for many was a fearful experience.
Premature infants were expected to die, as were most neonates with malformations. There was a
feeling that natural selection should be allowed to discard the runt of the litter, as suggested by
the designation of premature babies as weaklings. It was Budin and his pupil Couney who
pioneered incubator care of premature infants and thus helped change some of the early,
pessimistic attitudes toward these babies.
Another significant root of neonatology can be found in the quiet premature nursery, such as
that operated by Julius Hess and Evelyn Lundeen in Chicago in the early 1900s (1). Only
premature infants were admitted to these nurseries, and gentleness with minimal intervention
was the policy. To prevent infection, staff wore gowns, caps, and masks and set up a scrub
routine that excluded parents and minimized traffic in the area. Feedings of breast milk by
eyedropper were delayed for up to 72 hours, and the infants were handled as little as possible.
Yet the supportive conditions needed to allow the body to recover, as described by Florence
Nightingale, were present: warmth, rest, diet, quiet, sanitation, space, and others (2). Perhaps
some of the high-tech nurseries of today could benefit from an infusion of this superb nurturing
orientation.
Physiology is a taproot of neonatology. Advances in neonatal care rest directly on descriptions of
the changing body processes of the newborn infant. Men such as Barcroft and Dawes began
delineation of fetal circulation and placental function. These studies in turn led to the
establishment of the fetal lamb model, which subsequently was widely exploited. Neonatal
metabolic, gastrointestinal, respiratory, and central nervous system functions were studied by
Levine, Smith, Peiper, and others. The 1945 publication of the first edition of Clement Smith's
textbook, The Physiology of the Newborn Infant, was a signal event in our evolving ability to care
for sick newborns in a rational manner (3).
A final anchoring root of neonatology is the therapeutic trial. Innumerable traditional teachings
about premature infants eventually have been proved false. Without scientific testing as a guide,
neonatologists would constantly be off course. As it is, several dangerous misadventures have
been averted by clinical trials. An example is prophylactic sulfonamide treatment of premature
infants, which was
P.3
found to cause increased kernicterus (4). Silverman, Gordon, and Day were pioneers who
insisted on rigor in such trials.
Figure 1-1 The neonatal-perinatal tree shows the roots, trunk, and branches of the specialty.
PRESENT: THE TRUNK

The ICN is the institution that forms the sturdy trunk of endeavors in neonatology. The ICN is a
therapeutic environment, a collection of equipment, and a multidisciplinary team that is guided
by dedicated leadership, by a group of specific protocols, and by a body of relevant scientific
knowledge. It is the ICN as an integrated organism, rather than any single person or collection of
people, that takes care of the sick neonatein a sense, this is holistic medicine turned inside out.
The ICN is very effective in the care of desperately ill babies, but its rationale remains difficult to
explain to hospital administrators, insurance companies, and the U.S. Congress.
Over the years, there has been a steady increase in the intensity of illness observed in the ICN.
With this rise has come an increase in the number and variety of personnel and the amount of
technically sophisticated equipment. In the 1950s, premature care was a major concern. The
principal interventions were resuscitation, thermoregulation, careful feeding, simple and
exchange transfusion, and supportive care of respiratory distress. By the 1960s, electronic
monitors came into use, and blood gases began to be measured. Feedings were aided by
nasogastric tubes, and in-creased laboratory monitoring became possible. Antibiotics became
available for treatment of neonatal sepsis.
By the 1970s, the use of umbilical catheters and arterial pressure transducers was routine, and
respirator therapy for hyaline membrane disease began to succeed. Nutritional support for sick
infants was aided by transpyloric feeding tubes and finally by complete intravenous alimentation.
Microchemistry tests for most necessary parameters became widely available. Neonatal surgery
was shown to be feasible for many congenital abnormalities, including serious cardiac defects.
With the 1980s came the advent of computed tomography and ultrasonography. Significant
concern centered around ventricular hemorrhages and consequent post hemorrhagic
hydrocephalus in small premature infants. Transcutaneous electrodes became available first for
measurement of oxygen and then for carbon dioxide. Pulse oximetry was used increasingly for
continuous physiologic monitoring. Nutritional and metabolic supports were significantly refined.
Surfactant replacement has reduced the severity of lung disease in premature infants.
Extracorporeal membrane oxygenation permitted the survival of some previously unsalvageable
infants. In the 1990s, magnetic resonance imaging improved visualization of lesions, and
positron emission tomography and magnetic resonance spectroscopy promise to reveal the
physiology of the intact brain. As survival has become common for infants who weigh as little as
600 g (1.3 lb) at birth, increased attention has swung to assuring intact survival, and the 1990s
have been dubbed the decade of the brain.
THE BRANCHES
Perinatology
A body of specialized knowledge, a group of subspecialized professionals, the advent of
technically advanced equipment, and the formation of special care units all contributed to the
development of neonatology. In obstetrics, these same elements came together approximately
10 years later and resulted in the specialty of maternofetal medicine. Perinatologists developed
high-risk prenatal clinics and special delivery facilities for unstable patients. A steadily growing
body of literature from animal and clinical investigations allowed improved management of
pregnancy complications and monitoring of fetal status. Ultrasonography detected fetal
abnormalities and determined fetal size, anatomy, activity, breathing, and response to stress. For
the first time, mothers were given drugs designed to treat fetal conditions, and intrauterine
transfusions were performed in cases of threatened hydrops. Fetal surgery has yet to find its
proper place, but prenatal shunts have been inserted for hydrocephalus, obstructed urinary tracts
have been drained, and diaphragmatic hernia repair has been attempted.
In many major teaching hospitals, perinatal obstetric and neonatal services have joined forces to
form perinatal centers. Often with codirectors from the two disciplines,
P.4
these centers foster cooperation in the best interests of the high-risk patient. Integrated planning
and management in optimal cases consist of a high-risk prenatal clinic, timing and management
of labor and delivery, resuscitation, and intensive care in the nursery. Statistics on morbidity and
mortality, reviewed in periodic joint conferences, permit constant refinement of policy and
technique. These centers facilitate training programs in both perinatology and neonatology and
are important sources for research. They demonstrate the best survival rates for small premature
and other categories of infants at highest risk, and they provide the standard by which perinatal
care is judged.
The natural alliance between perinatal obstetrics and neonatology is so great that some have
suggested it receive department status within the medical complex. In some ways, the perinatal
center accomplishes this on an ad hoc basis. However, perinatologists and neonatologists are
parts of traditional departments of obstetrics and pediatrics, with surgical and medical
orientations, respectively. Allocation of resources and ranking of faculty occur predominantly
through the parent departments, where the priorities of the chairpersons are paramount.
Regional Connections
The medical community found that to optimize function of the perinatal center, it was necessary
to form complex relationships among hospitals and community resources (see Chapter 3). First,
neonatal and then maternal transport services were developed to move patients safely and
efficiently to the center and back. Hot lines were established to provide consultation and bed
allocation, which could be coordinated through a single phone number. Affiliations for continuing
education were formed, and standard protocols for referral were worked out. In some instances,
exchange of personnel and joint review of statistics helped cement the connection.
Government agencies extended a helping hand in this networking. Some state legislatures were
convinced early of the public health advantage of regional systems. The government began to
award grants to perinatal centers to underwrite some of the cost of outreach and system
activities. Training and research grants from the federal government often were given to the
same perinatal centers and served to buttress the resources of these centers. With passage of
the National Health Planning and Resources Development Act (Public Law 93-641) in 1974, the
federal government mandated regional planning of expensive resources in the interest of
efficiency and economy. Designation of care levels I, II, and III was adopted, and publication of
regional plans for perinatal care became widespread.
Unfortunately, during the 1980s regionalization suffered striking reversals. The mandated state
health planning mechanisms were emasculated. Hospitals became increasingly competitive and
wished to offer full services to prepaid health plans. Large numbers of neonatologists moved into
suburban hospitals and set up level II nurseries. The result was a decentralization of perinatal
high-risk care, with loss of efficiency and economy of scale.
Finally, I wish to call attention to the involvement of neonatologists in a mixture of educational,
administrative, and political activities that have been very taxing and time-consuming. Although
neonatology training has emphasized bedside care, a business school or public health degree
would be directly useful to today's neonatologist. A thorough training in counseling, group
dynamics, and law also would be invaluable.
THE PRESENT
Ethics
During the newborn period, babies with congenital malformations, asphyxia, and extreme
immaturity are seen. Modern, powerful life-support systems provide the technology for relatively
prolonged continuation of futile care. These circumstances have thrust neonatologists into the
center of a national debate on medical ethics. In the midst of all this, neonatologists must keep
their heads and care for babies and their families as best they can. Chapter 2 deals extensively
with the moral issues raised by neonatal intensive care.
Care of the Family

In general, there has been striking improvement in care of the distressed families of critically ill
newborn infants. Formerly excluded from the ICN because of fear of infection, parents are
brought to the bedside and encouraged to touch their babies. Limited sibling visitation often is
permitted. Nurses help parents get used to intrusive medical equipment, and, as soon as
possible, parents perform simple caretaking procedures with their child. Social workers, parent
support groups, chaplains, and literature for parents have become increasingly available. Starting
with the work of Klaus and Kennell, a growing body of publications has appeared that suggests
that bonding of parents to the newborn infant is important to later functioning of the family unit.
The change of the title of Klaus and Kennell's monograph from Maternal-Infant Bonding (5) in
1976 to Parent-Infant Bonding (6) in 1982 reflects recognition of the role of fathers in the
parenting of neonates.
Home Care
The aftercare of discharged high-risk newborns has become an increasingly important area of
concern for neonatologists. Although many problems are resolved quickly and require only
routine follow-up, a significant number of babies with chronic problems are taken home from the
ICN. As a small premature infant approaches discharge, it becomes clear that parent instruction;
sleep apnea testing; cardiopulmonary resuscitation training; concern with feeding and growth;
schedules for testing sight, hearing, and speech; physical medicine; and developmental
psychology
P.5
foreshadow a first year of life crowded with clinic visits and special needs. Infants with chronic
lung disease often go home on oxygen therapy, multiple medications, and special feeding
regimens. During the first year or two of life, intercurrent infections may require several
readmissions. The best outcomes have been achieved with early diagnosis of developmental
difficulties and the mobilization of community resources. All this necessitates input from a
physician with appreciation of the stormy neonatal period and an understanding of the continuing
problems of infants born prematurely with other perinatal insults. Often, a multidisciplinary team
at the tertiary center collaborates with the family's pediatrician to provide this special care.
Although oriented to neonatal intensive care, neonatologists have begun to assume responsibility
in this aftercare.
The Intensive Care Nursery as a Space Station

The quiet, simple, gentle premature nursery of 50 years ago has been transformed into a
bustling space station. Suspended halfway up a tall building, too often it is drenched in bright
light and alive with activity throughout the night. The tiny baby in the incubator is dwarfed by a
procedure light, a respirator, a multichannel monitor, four intravenous pumps, and a
transcutaneous oxygen module. The baby's chest is covered with electrodes, and tubes lead into
and out of his or her body at several points. Eight to 12 electrical cords, 2 oxygen lines, and a
suction catheter attach to the bedside console. Medications and flush solutions are drawn up in
syringes close at hand. A resuscitation bag is at the head of the baby's bed in case he or she
should stop breathing. Alarms sound, knots of people move busily about, and large machines are
pushed among the incubators. Ultrasonography elements are touched to the infant's head, chest,
and abdomen. Images flash on a screen and are recorded on tape for later analysis. House
officers speak rapidly in an acronymic code barely intelligible to residents who graduated 2 years
previously. Hess and Lundeen, coming some morning for a visit, might wonder if they had missed
the address and arrived on the wrong planet!
At such a bedside, today's neonatologist must cope with a constantly enlarging body of new
literature on cardiopulmonary physiology; nutritional and metabolic support; new antibiotics and
infectious disease conditions; ventricular hemorrhage and asphyxial syndromes; and surgical,
genetic, and cardiac problems. It is no longer possible for a physician to read all the new
publications, even when restricted to the area of neonatology. New therapeutic approaches are
proposed much faster than they can be tested in an orderly way; many such theories never will
be tested.
THE FUTURE
In the future, I believe neonatology will move in two divergent directions. In the realm of
neonatal intensive care, the contributions of basic science, and particularly genetics, will produce
powerful diagnostic and therapeutic tools, which will significantly improve results. However, the
global and public health dimensions of dealing with distressed newborns will require simpler, less
expensive interventions that can be applied in developing nations. Clearly, the bang for the buck
in neonatology is greatest where care is the least developed, and is least where the standard of
care is already high. The challenge will be to keep the locomotive moving forward without leaving
the caboose too far behind.
Of course, I cannot report to you with any precision what new developments in neonatology will
occur in the next 20 years. Nevertheless, Table 1-1 is offered to illustrate, side by side, several
areas of practice: past, present, and future. A few themes are highlighted. The past was
characterized by minimum intervention and a nurturing sort of care dominated by gentle bedside
nursing. The present era seizes on maintenance of normal physiologic and biochemical
parameters. The babybeset with indwelling lines, ventilators, and monitors, who dies in
balanceis in line with this approach. The future will include accepting or deliberately producing
some abnormal values in recognition of the baby's unique disease state. Examples include
cooling for neuroprotection, hyperthermia for tumor inhibition, permissive hypercapnia to
minimize pulmonary barotrauma, and cardiovascular unloading to treat neonatal lung failure. We
will develop gene products and receptor blockers as medications, and use gene promoters and
inhibitors to manipulate the cell's own genetic apparatus to treat diseases. Examples are the use
of vascular endothelial growth factor (VEGF) and elastase inhibitors for reducing chronic lung
disease (7,8,9).
Worldwide, measures to combat perinatal and neonatal infections, lower the incidence of
prematurity, and reduce maternal malnutrition will benefit outcomes without the need for new
technology. The impact of basic public health measures is so dramatic that one World Health
Organization (WHO) official described infant mortality as a public health problem with medical
aspects, not a medical problem with public health aspects (10,11).
Finally, the way we learn will evolve in the years ahead. Our scientific method emphasizes
comparisons of groups differing by a single variable, ideally in a single disease state. Studies
using tissue culture or the laboratory mouse are the models we try to approximate in the
intensive care nursery. But our babies typically have many abnormal states simultaneously, and
many therapies running at the same time. The staff is caring for many infants at the same time,
and sometimes several investigative protocols are running alongside one another. The research
study is never the first priority where life-and-death illness is involved. While all this is going on,
there are changes in policy and equipment that affect care in the whole neonatal intensive care
unit (NICU), across all diseases and birth weights.
We will need new types of statistics to aid in benchmarking interventions and studies of the
impact of benchmarking practices. Large, computer-supported databases will help with
hypothesis generation. We will become more adept at recognizing secular trends in NICU care
P.6
P.7
versus the specific interventions we intend to study. And some of our interventions will be
outside the nursery, in programs of education and/or pre- and post-NICU care. The years ahead
will be as fast paced and challenging as those that have gone before.
TABLE 1-1 NEONATOLOGYPAST, PRESENT, AND FUTURE. THE AUTHOR'S THOUGHTS
AND REPRESENTATIVE EXAMPLES
PAST
Temperature
Nutrition
Infection
Brain injury
Metabolic funct
Gene function
Physical bundling
Laissez-Faire
PRESENT
FUTURE
Cooling for brain

protection
Hyperthermia for
tumor inhibition
Small frequent
Early tinyfeeds: priming Growth factors
feeds
I-V alimentatiaon
Gene manipulation
Nasal spoon
gavage/gastrostomy
anti-stress endocrine
Human milk
adjustment
Isolation
Antibiotics
Immunotherapy
Aseptic technique
Some anitvirals
Gene manipulation
Resuscitation triage Trials of phenobarb, Mg, Anti-oxidants, reduce
cooling
reperfusion injury
DR resuscitation,
Gene manipulation;
support
inhibitors and
expression triggers
Reduce programmed
cell death
Minimal lab work
Microchemistries
patterns of gene
Minimal intervention Adjust towards normal activation
values
microarrays, 10,000
Errors of
metabolism
described
Thermoneutral
maintenance
Servocontrols
genes RNA
inhibitors, releasers,
gene products
Enzyme ontogeny
Selective gene
Neonatal pharmacology activation
Rx with gene products
or inhibitors
Ontogeny of gene
activation
Birth defects
Description and
Neonatal surgery
Preconception risk
observation
Fetal ultrasound Dx
identification
Support
Intrapartum
management of risk
Fetal genetic diagnosis
and treatment
Molecular study of
normal/abnormal
development
Lung function
oxygen
respirators, HFO
Vascular endothelial
humidity
inhaled nitrous oxide
growth factor (VEGF)
surfactant replacement
VEGF activation or
ECMO
Scope of practice
How we learn
instillation
elastase inhibitors
chronic inhaled NO
cardiovascular
unloading
One patient, one
NICU: large team of
World-wide, public
doctor
attendings and support health point of view
One disease, one
Multiple abnormal
Cutting edge biotransaction
conditions
technical vs simple,
The controlling unit simultaneously
practical
nurse
Multiple subspecialities, Prevention vs drastic
consultants
support (prematurity,
Support services: lab,
infections, malnutrition
imaging social work
Low tech, lay trainee
Ethics, legal
interventions
requirements, HMOs
Education, change of
Funding, privacy
traditional practices
regulations
Benchmarking, care
Family rights,
systems research
autonomy, advocacy
Telemedicine, regional
centers
Laboratory models: Multi-institutional
Dealing with multiple
animals, tissue
controlled clinical trials variables and secular
culture, reports of Fundamental basic lab change
institutional results research
New statistical
The published series Molecular mechanisms methodology for
Focus on a single
dominate grants
benchmarking
variable
Outcome studies
Large outcome
Physiologic research
databases
The computer in the
NICU: care, record
keeping, study
Multiple levels of
phenomenology
The single variable in a
single disease state

The impact of changed
practices across
diseases
The result of societal
pressures on practice
The impact of prehospitalization care
The impact of posthospitalization care
Public health and
health of the baby
It takes a village to
raise a child
REFERENCES
1. Hess JH, Lundeen EC. The premature infant: its medical and nursing care. Philadelphia: JB
Lippincott, 1941.
2. Nightingale F. Notes on nursing: what it is and what it is not. (A facsimile of the first edition
printed in London, 1859, with a foreword by Annie W. Goodrich.) Philadelphia: JB Lippincott,
1969.
3. Smith CA. The physiology of the newborn infant. Springfield, IL: Charles C Thomas, 1945.
4. Silverman WA, Anderson DH, et al. A difference in mortality rate and incidence of
kernicterus among premature infants allotted to two prophylactic antibiotic regimens.
Pediatrics 1956;18:614624.
5. Klaus MH, Kennell JH. Maternal-infant bonding. St. Louis: CV Mosby, 1976.
6. Klaus MH, Kennell JH. Parent-infant bonding. St. Louis: CV Mosby, 1982.
7. Rabinovitch M, Bland M. Novel notions on newborn lung disease. Nat Med 2002;8:664666.
8. Zaidi SH, You XM, Ciura S Husain M, et al. Overexpression of the serine protease elafin
protects transgenic mice from hypoxic pulmonary hypertension. Circulation 2002;105:516521.
9. O'Blenes SB, Fischer S, McIntyre B, et al. Hemodynamic unloading leads to regression of

pulmonary vascular disease in rats. J Thorac Cardiovasc Surg 2001;121:279289.
10. Vidyasagar D. A global view of advancing neonatal health and survival. J Perinatol
2002;22:513515.
11. Coco G, Darmstadt GL, Kelley LM, et al. Perinatal and neonatal health interventions
research. J Perinatol 2002;22:S1-S4.
Chapter 2
Current Moral Priorities and Decision
Making in Neonatal-Perinatal Medicine
Robert J. Boyle
John C. Fletcher
(deceased)
GOALS OF NEONATAL-PERINATAL MEDICINE

The first part of this chapter discusses broad goals and moral priorities of neonatalperinatal medicine (NPM) in relation to patients, American society, and developing
nations. The second part reviews some specific ethical issues that frequently arise in
clinical decision making in the neonatal period.
Medicine is a goal-oriented profession. One approach might argue that healing is the
sole and overriding goal of medicine. This view is unconvincing, because some valid
goals of medicine (e.g., prevention) cannot be collapsed into healing. The practice of
NPM illustrates that medicine has multiple, complex, and sometimes conflicting goals:
To save life and cure disease
To relieve pain, suffering, and disability
To rehabilitate and restore function
To prevent disease
To improve the quality of living and dying
To seek new knowledge
The highest moral priority of these subspecialties is to implement medicine's goals in

the perinatal period. In a recent review, Carter and Stahlman broadly discuss the
successes (and limits) of neonatal intensive care in the United States over the past
40 years (1).
SOCIETAL AND PROFESSIONAL ISSUES

Medicine is also accountable to society and to the international community. Critics
may validly argue that American medicine and public policy have neglected primary
prevention and fostered the technological imperative. To counter these criticisms,
the moral priorities of NPM specialists need to include advocacy for comprehensive
prenatal care, improved nutrition, prevention of premature delivery and appropriate
family support and services for the infant after discharge. NPM in the United States is
also part of a global community of science, medicine, and humanitarian outreach.
The moral priorities of the field ought to include commitment to global infant health.
An editorial by Sidney Brenner, a leading figure in genetics, also applies to neonatal
medicine. He wrote: Somehow, in our abstruse discussions about the high
technology of the developed world, we need to balance the equation by thinking
about the rest of the world as well (2). Nations where newborns are in greatest
danger of death or lifelong disability could benefit from transferable knowledge and
technology, for example, maternal folic acid supplementation, prevention of HIV
transmission, vitamin K prophylaxis, eye prophylaxis for gonorrhea, and prevention
of neonatal tetanus.
Preterm deliverylive birth prior to 37 completed weeks' gestationgrew in the
United States from 8.9% in 1980 to 11.9% in 2001, more than 476,000 annual
preterm births (3,4). The increase is caused by an 11% rise in moderately preterm
births (32 to 36 weeks' gestation) in the 1990's. In 2001; charges for hospital stays
for prematurity-related illness reached $13.6 billion. There are large differences in
preterm births by race and ethnicity, although non-Hispanic white women give birth
to the majority of the 4 million annual live births.
Paradoxically, both poverty and economic success are harmful to many infants.
Poverty, race, class, and substance abuse are linked to a large percentage of
prematurity, extremely low-birth-weight, and brain injury. However,
P.9
multifetal pregnancies after assisted reproductive technologies (ART) and non-ART
procedures have reached epidemic proportions (5,6,7,8). Epidemiologic studies also
suggest an increased risk of major birth defects after assisted reproduction (9).
Neonatologists see first hand the human damage and costs of these polar social
determinants of gestational harm. Infertile individuals who want their own genetic
child suffer psychologically and spiritually. Fertility treatments have allowed literally
millions of women to have children who otherwise would not have been able to
conceive. The primary responsibility to reduce these harms falls on reproductive
medicine and policy makers. NPM specialists can facilitate research and remedies for
the problem. To date, reliance on professional self-regulation has not been effective
in reducing multifetal pregnancy in the United States. There is currently a debate
about whether social controls of infertility treatment and research are fitting for a
pluralistic society that highly values reproductive liberty. The President's Council on
Bioethics is now considering a set of very strong federal controls to protect human
embryos and constrain infertility treatment (8). At this point, all could agree that
women undergoing ART should be informed about the risks of multiple gestation and
preterm delivery.
What perspective ought to be taken on the social ethics of the medical profession
and its mission? Many in the United States believe that the answer is simply to let
markets work in the interests of liberty (10). In this model, health care is a twotiered marketplace. The costs for those who can pay include a pool of money and
charity care for those who cannot pay. When the overriding value is justice, medicine
ought to be a largely subsidized service in a society that guarantees security in
health care for all to increase solidarity among its people. However, even when
health policy guarantees universal access, disparities persist in disease and health
care that reflect class, race, and sex.
Following Buchanan, we propose that in a pluralistic democracy, the relation between
society and medicine ought to be understood as a self-correcting contract or bargain
(11). Instead of resting on one supreme value easily captured by ideology, the
contract reflects a commonwealth of values. These values govern the complex goals
of medicine and interests of society in health. A commonwealth has a primary locus
of loyalty. In this society, the preeminent value is fidelity to patients or loyalty to the
patient at hand. Using the right procedural principles, patient-centered medicine is
not inconsistent with viewing the patient in a population with needs for a finite
supply of expensive community health care resources. (12) Within constraints of
fidelity to patients and their rights, physicians ought to promote the welfare of the
many, to practice fairness in access and distribution of resources, and to be efficient
and effective in their practices. Physicians in this society are also members of a
scientific community with a high standard of evidence in practicing medicine. Guided
by these values, society grants the profession a privileged place, permits its
members to earn high incomes, and subsidizes their training. In return, society
expects the net health benefits of pursuit of the goals of medicine to outweigh the
net costs.
In the United States, the increasing investment in perinatal-neonatal care has not
produced proportional improvements in crude infant survival and low-birth-weight.
Thompson and associates (13) compared the neonatal intensive care (NICU)
resources of the United States with those of Australia, Canada, and the United
Kingdom. Compared to the other three nations, the United States has far greater
resources but higher rates of low-birth-weight and death among newborns. The
United States has 6.1 neonatologists per 10,000 live births compared with 3.7, 3.3,
and 2.7 in the other nations, respectively. NICU beds in the United States are 3.3
per 10,000 live births, compared with 2.6 in the first two nations and 0.67 in the UK.
Resources are not linked with low-birth-weight. In the United States, 1.45% of all
neonates have a very low-birth-weight (<1500 g) compared with approximately 1%
in the other nations. Newborns weighing less than 2500 g are born more frequently
in the United States. Low-birth-weight is the single most accurate predictor of
neonatal death. The crude neonatal mortality rate is 4.7% deaths per 1000 births in
the United States, compared with 3.0% in Australia, 3.7% in Canada, and 3.8% in
England and Wales.
Why does success not follow from greater resources? Thompson and associates (13)
point to cultural and political differences between the nations. Australia, Canada, and
the United Kingdom provide health insurance for all children under 18 years of age
and all women ages 18 to 44 years. In the United States only 86% of children and
78% of women had health insurance at the time of the study. These three nations
provide free family planning advice and prenatal and perinatal care, and the United
States does not. The influence of race, poverty, and poor nutrition must also be
factored into the answers. Longstanding recommendations of the Committee on
Perinatal Health to address the root problems have not been implemented (14). The
current March of Dimes campaign on prematurity is a valuable step but does not
replace missing federal support (3).
ETHICS AND MORAL DELIBERATION

Ethics is a body of practical knowledge comprised of principles and values,
judgments about cases and policies, and beliefs and theories about the world and
persons (15). John Dewey's view of the main purpose of ethics is integration of
beliefs about the world with beliefs about the values and purposes that should direct
human conduct (16). Decision making for infants and children should follow the
same general approaches taken for all patients who are unable to make decisions for
themselves.
John Dewey wrote, life is a moving affair in which old truth ceases to apply (16).
Making judgments in novel situations can lead to change of background beliefs and
practices. Since the advent of modern neonatal care in the 1960's, rapid advances in
technology and aggressive care of smaller and sicker neonates have focused
philosophical, theological, professional, legal, consumer, and media attention (both
P.10
positive and negative) on the use of technology and the processes for decision
making in life or death choices for newborns. One can trace changes in background
beliefs and values in the history of ethics in neonatal decision making using the
landmark cases in which medical decisions for infants have been scrutinized by the
United States government or the legal system. Table 2-1 gives samples of important
cases, events, articles, and studies in that history. In addition, several recent
reviews trace this evolution and multinational differences in approach (1,17,18,19).
TABLE 2-1 LANDMARK CASES IN NEONATAL ETHICS
Year Name of Case, Event
1963
1974
1973
Clinical Issues
Trisomy 21, duodenal
Hopkins Baby
Outcome
Died after 15 days of
atresia; parental refusal no feeding/ hydration.

of surgery.
Muliple malformations;
b,c
Baby Houle
Court orders surgery;
parental refusal.
infant died.
Selective nontreatment Controversial report
Duff & Campbell
of multiple infants with

variety of medical
from a NICU.
conditions.
1981
The Danville Twins
Siamese twins, joined at The parents and

the abdomen, with
three legs and a
physicians were found

not guilty.
common pelvis. The

parents and physicians
were criminally charged
with ordering that the
infants not be fed.
1981
Premature infant; 800

g, 26 wks; parents
Stinson baby
Infant dies at 6 mo.
object to unwanted
treatment.
1982
Bloomington Baby
g
(Baby Doe)
1983
Baby Jane Doe
1983
Baby Doe
regulations
1983
President's
surgery.
Spina bifida,
hydrocephalus, and
microcephaly; parental
refusal of surgery.
U.S. Dept. of HHS
issues regulations.
Indiana Supreme Court

rules for parents.
NY Court of Appeals
rules for parents.
Requires life-sustaining
treatment for every
infant.
Commission
Clarifies decisions to
forgo treatment in
newborns.
More moderate than U.

S. Dept. of HHS
regulations.
Child Abuse
Federal law.
To receive federal
funds for child
protection, states must
have procedures for
such cases.
1984
Trisomy 21, tracheal

esophageal fistula;
parental refusal of
Protection Act
1990
1994
Anencephaly.
l m,n
Baby K ,
Mother demands life

supports and federal
courts rule for her.
o,p,q
Messenger
Premature infant; 780
Father disconnects
g, 25 wk. Parents
respirator and jury
request no resuscitation. finds not guilty of
manslaughter.
2000
r,s
HCA v. Miller
Premature (629 g, 23
Court of Appeals
wk) infant resuscitated reverses prior

against parents wishes. judgment awarding
parents$60 million.
Gustafson JM. Mongolism, parental desires, and the right to life. Perspect in
Bio Med 1973;16:524.

b
Maine Medical Center v Houle, No 74145, 1974 (Super. Ct. Cumberland
Co. Me. Feb. 14, 1974).

c
McCormick RA. To save or let die: the dilemma of modern medicine. JAMA
1974;229:172.
d
Duff RS, Campbell AGM, Moral and ethical dilemmas in the special care
nursery. N Engl J Med 1973;289:890.
Stinson R, Stinson P. The long dying of Baby Andrew. Boston, MA: Little
Brown, 1983.
f
MurrayTH, Caplan AL. Beyond Babies Doe. In: MurrayTH, Caplan AL, eds.
Which babies shall live: humanistic dimensions of the care of imperiled
newborns. Clifton, NJ: Humana Press, 1985:3.
g
State ex rel. Infant Doe v Baker, No. 482 S 140 (Ind. May 27, 1982).
Weber v Stony Brook Hosp, 476 NY.S. 2d 685, 686 (App. Div.); Bowen v
American Hospital Association, 476 US.610 at 611(1986).
U.S. Department of Health and Human Services. Nondiscrimination on the

basis of handicaps: procedures and guidelines relating to health care for
handicapped infants. Federal Register 1984 Jan 12:49:622654.
U.S. President's Commission for the Study of Ethical Problems in Medicine

and Biomedical and Behavioral Research. Seriously ill newborns, in deciding
to forego life-sustaining treatment: a report on the ethical, medical and legal
issues in treatment decisions. Washington: US. Government Printing Office;
1983;197.
Child Abuse Protection Act, 42 U.S.C. 5103 (1982).
In re Baby K, 832 F. Supp. 1022 (E.D. Va. 1993); In re Baby K, 16 F. 3d

5900 (4th Cir.).
Annas G. Asking the courts to set the standard of emergency carethe

case of Baby K. N Engl J Med 1994;330:1542.
Paris JJ, Crone RK, Reardon FE. Physician refusal of requested treatment:
the case of Baby K. N Engl J Med1990;322:1012.

o
State v Messenger, file 9467694-FY, Clerk of the Cir. Ct. County of
Ingram, Mich.
p
Clark FI. Making sense of State v Messenger. Pediatrics 1996;97:579.
Paris JJ. Manslaughter or a legitimate parental decision? The Messenger
case. J Perinatol 1996;16:60.

r
HCA v Miller, 2000 WL 1867775, Tex. App. Hous. (Dec. 28, 2000).
Paris JJ, Schreiber MD, Reardon F. The emergent circumstances exception
to the need for consent: the Texas Supreme Court ruling in Miller v HCA. J
Perinatol 2004;24:337.
Abbreviations: HHS, Health and Human Services; NICU, neonatal intensive
care unit.
Neonatologists, nurses, and parents of newborns today stand on the cumulative
moral experience of the past. The
P.11
process for decision making in NPM is more transparent and shared between
clinicians and parents than in the past. However, it is never immune from confusion,
especially when cultural beliefs about medicine collide. Practitioners are arguably
better trained to identify ethical issues, participate in shared decision making, and
seek help with ethical problems. For example, most try to be empathetic, nonbiased,
and honest in disclosing a poor or uncertain prognosis to anxious parents.
Neonatologists, nurses, and their colleagues are fallible. They have made serious
mistakes in excluding parents from decisions and imputing guilt to them for wanting
to withdraw treatment (20). Some events and cases test the limits of moral
concepts, for example, the best interests of the infant, parental autonomy,
professional integrity, futility, and fairness in use of resources, or quality of life.
These cases brew storms that can throw experienced neonatologists, nurses, other
staff, parents, and administrators off balance and into conflict with one another. In
these situations, Martin Buber's metaphor of a narrow rocky ridge between the
gulfs (21) can ring true. Ethics helps human beings to keep enough balance, when
taking hazardous, puzzling, or new paths in social and personal life, to maintain
moral insight and equilibrium. It is a self-correcting, constantly evolving body of
practical knowledge about human problemsconflicts of moral principles, duties, or
loyalties in relationships.
CULTURAL AND RELIGIOUS DIVERSITY

Because ethics is a self-correcting body of knowledge, open societies require a public
process of debate about continuity and change in the moral aspects of social and
professional practices. The United States is now the world's most diverse religious
and cultural society. These differences are very difficult to sort out in debate about
ethics. To change public policy in this society, faith-based reasons standing alone are
insufficient. Secular and presumably rational reasons are required to persuade a

court, legislature, or commission. In debatable health care decisions, we see cases
involving newborns that break down the limits of rationality. For example, should
physicians continue to respect parents' faith-based rejection of futility of treatment
when the baby has total intestinal necrosis or renal agenesis and pulmonary
hypoplasia? In our view, the case of Baby K (22) shows that the federal courts can
approve parental faith-based demands as a legally justified reason to continue futile
treatment of an anencephalic newborn, because the society has been unwilling to
ration resources for this category of newborns.
Cross-cultural issues deeply complicate debate on ethics. For example, should any
moral weight be given to a request to withdraw life support if the parents of a
neonate are immigrants from a society with very different moral beliefs about
medicine and what constitutes a seriously impaired newborn? What if the parents'
culture requires that a traditional healer participate in the care of the infant? Should
weight be given to parents' religious views when the infant is in pain and not
benefiting from additional procedures and the parents flatly reject the concept of
futility? What is best, all things considered, for the infant? Does concern for the
infant absolutely override competing concerns?
The authors take a broadly pragmatic approach to ethics and to cross-cultural issues
in ethical debates. Moral norms and ways of life are, in fact, relative to culture and
personality. However, cultural and psychological relativity does not drain ethics of
content or moral force. Although ethics is a phenomenon within and not independent
of human experience, it has objectivity in two senses. First, ethics is objective as a
growing and evolving body of knowledge about practices in society and the
professions. Secondly, the basic imperatives of ethics are no less objective than law
or medicine (23) when it comes to destructive acts like murder, torture, genocide,
rape, cruelty for its own sake, and so forth. Such acts violate our most basic
convictions about the intrinsic value of human life and community. As Benjamin
writes, If these things aren't wrong, nothing is (24). Some argue that moral
judgments are not binding across cultural boundaries or that ethics is merely a
matter of personal taste. If ethics is merely subjective or relative, there is no basis
for a moral judgment about any horrific act that occurs in subcultures or beyond
one's own culture, such as forcing children to wage war against their own people or
to suffer loss of limb or life. The case of Baby K ought not be judged as horrific
acts because the parents' intentions are loving and not destructive. The moral
failure in these cases is societal in its failure to assume responsibility for setting
democratically debated limits on health care resources. Societal failures can result in
horrific consequences, but the parents and staff involved in these cases do not
deserve moral blame.
Working Through Ethical Problems

Although important moral values and principles are binding, there is no prefabricated
way to work through ethical problems in cases or events when principles collide.
Different methods of moral deliberation compete for selection. When compared with
highly theoretical or case-by-case approaches, the method of wide reflective
equilibrium (25) has proven useful in bioethics. In this approach, one weighs ethical
problems in cases by creating a dialogue (internally or externally) that moves
between three interactive elements: the values and principles at stake, the problem
at hand, and relevant background beliefs and theories.
Principlism is a dialectical method that analyzes ethical problems in a framework of
prima facie principles of biomedical ethics (10). This approach is widely used in
ethics courses and in the literature of neonatal ethics. In our view, a method that
moves only between values or principles and the problem at hand lacks larger
constraints and correctives. Cultural influences and personal bias are at work in
selecting principles and putting them to work in cases. Judgments made in past
cases, such as Baby K (22)
P.12
and Baby Doe (26), are fallible and need evaluation. Background beliefs and theories
are sources of critical distance and constraintbeliefs and theories about the nature
of personhood, community, the world as revealed by science and metaphysics,
human psychology, sociology, and political and economic behavior; the nature of
nonhuman animals, and so forth. In short, our uses of principles and judgments in
cases ought to make sense in terms of intelligible background beliefs and in-depth
knowledge of the issues at hand.
When parents and clinicians are unable to reach agreement on an approach to care,
additional resources should be available to assist the process. Many ethical
conflicts are actually problems in communication and misunderstanding.
Involvement of social workers, chaplains, parent support groups, or other clinical
staff often improves the interaction and leads to a resolution of the conflict. One of
the outcomes of the Baby Doe controversy (27) in the early 1980's was the
recommendation that infant bioethics review committees be established to
prospectively and retrospectively review cases where life-sustaining therapy was
withheld or withdrawn. While the function of these committees was short-lived, they
were the forerunners of the current ethics committees in most institutions. Ethics
committees serve multiple functions including policy development and review, staff
education, and review of difficult cases brought to them. In many institutions, both
medical staff and families may bring cases to the committee. Committee process
varies widely from institution to institution in terms of membership, involvement of
family, and decision-making models (28,29,30). Other institutions use a less
intimidating model of an ethics consultation service, where an individual consultant,
or, more often, a team of consultants (who usually have multidisciplinary
backgrounds), will review the situation with those involved, provide background or
policy information, facilitate communication, and, when necessary, bring the parties
together to work toward a consensus. In most institutions the committee or
consultation service recommendations are advisory rather than binding. Finally,
when consensus is cannot be reached and the clinician, family or institution feels the
issues are sufficiently serious, the court system can be petitioned to intervene. While
the court system is seen as the voice of society weighing in on a difficult issue, this
is often a time-consuming, not necessarily objective process that by its nature is
adversarial and may negatively impact the relationship between the clinician and
parents. A court decision may be appealed and overturned or may result in an
outcome that with greater societal scrutiny may be seen as questionable. Most would
argue that approaching the court should be a last resort.
Moral Status of the Newborn

Part of the evolution in clinical ethics in neonatal-perinatal medicine has involved
ascribing greater moral status and protection at birth to seriously ill newborns. There
has been a dramatic transition from frequent nontreatment of trisomy 21 (31) to the
Baby Doe Rules (22) and support for appropriate assessment of extremely low-birthweight (ELBW) infants.(32) However, there remains a philosophical impression that
decisions to withhold or withdraw treatment for newborns, especially the extremely
premature newborn, may be less troublesome than decisions for older children and
adults, because the newborn does not have the same moral worth or is not fully a
person. Tooley (33) argued that because a newborn lacks advanced brain function,
has no self-consciousness, ability to suffer, or sense of the future, it should not be
considered a person. Engelhardt (34) agrees that neonates are not persons in the
strict sense at the time but affirms that they are persons in a social and cultural
sense. Others have further defined this view, identifying the bonding, affection, and
care given newborns by parents and other adults (35).
While detailed plans for conception, antenatal ultrasound, and fetal movement may
support parental attachment early in and progressing through gestation, it cannot be
denied these emotions are not as strong in the perinatal period, especially when the
infant is extremely preterm, as they are at several months or years of age. In
addition, many preterm infants are the result of unplanned, often unwanted, or
sometimes unknown pregnancies. Blustein describes the neonate as not born into
the family circle so much as outside it, awaiting inclusion or exclusion. The moral
problem the family must confront is whether the child should become part of the
family unit (36). Ross notes that respect for the person is owed to all individuals on
the basis of the individual's personhood (and developing personhood) proportionate
to the actualized capacities of the individual and his or her potential to attain full
personhood (37). It is therefore problematic that the infant's moral worth is
variable, dependent on acceptance by its family rather than on any innate socially or
even legally accepted character. This worth may change over a short period of time,
and this change may then affect differences in or difficulties with decision making
prior to birth versus at a week or month of life. Decisions made in the labor room
regarding a child who has never been seen by the parent and who may have an
increased risk of mental retardation become much more difficult when the 6-monthold child has confirmed severe developmental delay.
Decision Making
Sound clinical decision making should be based on sound data, a careful and
thorough diagnostic assessment, and, based on that diagnostic assessment, accurate
prognostic estimates. Sometimes this is relatively easily accomplished: chromosomal
analysis has documented trisomy 13 and the natural history of this disease is wellknown. In other situations the diagnosis may not be well-defined or the prognosis is
uncertain. For both the ELBW infant about to be born or the premature infant with
grade III intraventricular hemorrhage, there is an increased statistical risk of
developmental delay, but it is not certain how this particular
P.13
infant will progress. Rhoden (38) has defined strategies that have been or might be
used when there is uncertainty about prognosis:
Wait until certain: Continue until the patient is actually dying or will survive
but with definite severe disability. There is an underlying discomfort with
infants, admittedly few in number, who are dying but who might survive with
aggressive care. There is little attention paid to suffering, burden-benefit
ratios, or the number of infants needed to be treated for one additional intact
survivor.
Statistical prognosis: Use statistical cutoffs and aggressively treat all those
selected. This might be described as the evidence-based approach. Selection
might be by birth weight or gestational age. This approach may be used when
resources are limited. Professional, regional, or national guidelines might exist
to define these cutoffs. This approach ignores individual variation and may
sacrifice some potentially normal infants who may behave outside of the norm.
It relies on data that may or may not accurately reflect the clinical situation at
hand. Decision making is psychologically easier, because it is allegedly
objective. Most clinicians justifiably use some criteria for limiting treatment
for some populations, for example, resuscitation at 21 to 22 weeks gestation.
Individualized prognosis: Decide for each infant using the available data, the
present condition, and a benefit-burden analysis. This approach allows for
clinical change, evaluation and reevaluation, and ongoing communication.
There is more of a role for the family in decision making. It also can be a
source of confusion, uncertainty, error, and agony. However, Rhoden believes
that this is justified, given the tragic nature of the situations. Fischer and
Stevenson (39), and more recently Kraybill (40), have expanded on this
approach beginning with a nonprobabilistic paradigm of attempting to save
every ELBW infant's life (provisional intensive care for all) modified by an
individualized prognostic strategy when prognosis could be better defined.
The American Academy of Pediatrics has endorsed this approach (41).
Withholdin/Withdrawing
Most agree that it is ethically superior to withdraw a therapy compared to
withholding it (42,43). If therapy is begun and is effective, the patient benefits. If
the therapy is begun but is not effective, it can be stopped. If the therapy is never
initiated, the patient can never benefit. Initiation of therapy also provides the
clinician with additional time to collect data, which may lead to a more accurate
diagnosis and therefore more reliable prognosis and allows the family more time to
understand the situation. While a preference to withdrawing rather than withholding
is sound philosophically, in actual clinical settings there are often emotional
responses and in some settings religious restrictions to withdrawal of a therapy. It is
much easier emotionally to be passive than to make an active decision to withdraw
aggressive care. Pulling the plug and killing my baby are not infrequently heard.
Letting God decide and waiting for a miracle are examples of the same
phenomenon. Doron and associates (44) document the relatively frequent
disagreement of parents when clinicians recommend withholding aggressive care.
There is a potential for weeks or months of care, and possible pain and suffering,
before the infant's death or an eventual decision to withdraw. The clinical situation
may reach the point that there is no aggressive care to withdraw, and the outcome
is poor. Some clinicians are reluctant, based on this experience, to initiate care as
freely. If a clinician is less willing to withdraw care than not to initiate it because of
uncertainty or philosophical reasons, this should influence their decision process and
be made clear during discussions with parents.
Clinically the withdrawal of life-sustaining therapy does not require that the patient
have multiorgan system failure or meet criteria for brain death. If the organ that has
failed or been irreversibly injured is a vital one, decision making should be based on
the infant's prognosis for recovery, long-term survival, quality of life, and so forth.
Technology would be withdrawn, and the infant would be allowed to die. Brain
death or death by neurological criteria is a clinical and legal definition of a type of
death. Criteria for brain death in the newborn are somewhat different than for older
children and adults (45,46,47). Some jurisdictions allow for a religious exemption.
However, in most circumstances, once the criteria have been met, the patient is
pronounced dead. These families need to be compassionately informed about what
this means ethically and legally and appropriately prepared for withdrawal of the
ventilator. There should be no discussion of removing life-sustaining treatment or
keeping the baby alive because the patient is legally dead.
The term euthanasia generates a great deal of confusion and debate in legal,
legislative, media, and clinical spheres. Active versus passive euthanasia, voluntary
versus involuntary euthanasia, physician-assisted suicide, and other descriptors have
created unfortunate ambiguity about the actual issue at hand. If one defines
euthanasia or active euthanasia as directly and actively causing the death of a
patient who may not be imminently dying or is dependent on life-sustaining
technology, usually by administering a lethal dose of a drug, most state law and
policies of the American Medical Association and the American Academy of Pediatrics
would prohibit that action (48,49). How often active euthanasia of neonates actually
occurs in the United States is not known. However, there is a significantly different
approach in Europe. Recent studies suggest that while active euthanasia is illegal
and rarely occurs in most countries, it appears to be acceptable practice in the
Netherlands and France. Approximately 70% of neonatal physicians in France and
45% in the Netherlands have been involved in a decision of active euthanasia (50).
Confusion arises when the term passive euthanasia is used for decisions to
withdraw life-sustaining therapy with the expectation that the patient will die (51).
The use of medication to treat symptoms of pain or dyspnea or other suffering in the
context of comfort or palliative care further compounds the confusion,
P.14
in spite of an ethical duty to provide this type of care. Some would argue that the
overall intention may be the same as or may certainly blur into active euthanasia.
However, others would suggest that intention is an important determinant (52).
Describing what actually is being considered as a plan of care and avoiding the
terminology may prevent the confusion and the associated emotion.
The Role of Parents

Parents are the decision makers for their children. This role is understood to have
social, legal, and ethical facets; some are better defined than others, some are
potentially in conflict with others. Parents should be in the best position to judge
what is in their child's best interest. They will be the ones who will continue to raise
the child after the newborn period. Numerous studies and commentaries have
reflected on the difficulties families may have to face following the birth of an infant
who will have long-term medical and educational needs and how they perceive their
situations (53,54,55,56). They will be the ones who will have to deal with the
consequences of the decisions. Ethically, this right and responsibility of parents has
been the focus of considerable analysis and commentary since the Baby Doe era.
The extremes of no one, including parents, should make these decisions and
absolute right of parents to make medical decisions are simplistic and unrealistic
(57). Parents should be seen as guardians of their child's welfare, not owners.
Batholome (58) suggests using the language of permission, a term somewhat less
rigid than consent. He sees the parents' role as a duty rather than a right, a duty
to ensure the provision of necessary medical care. Parental permission for
interventions into children's lives must not be seen as the unconditional right to
demand or refuse a particular intervention because it is a proper exercise of parental
authority over the lives of children. That children are largely dependent, at least for
a time, on their parents is to be affirmed, but that dependency does not warrant the
second-class social standing implied by a parental right of consent in decisions
affecting the health care of their children. Child abuse and neglect statutes
frequently result in challenges to parental decision and control. Finally, decision
making for another individual is not identical to making decisions for oneself. While
the legal system and ethical consensus give an adult patient's personal decisions
wide leeway, surrogate decision making for the adult patient requires that decisions
be solely in the patient's best interests or what the patient would have wanted.
Surrogates are often questioned, especially when life-sustaining therapies are at
issue. The infant and child should be afforded this same protection.
Weir (59) proposes that the parents as decision makers should:
Have relevant knowledge and information about medical facts, prognosis, and
family setting
Be impartial
Be emotionally stable
Be consistent
This final quality should assure that the process ends with the same result in similar
cases. However, what information the parents receive and from whom may clearly
impact their decision. The clinician has an obligation to present accurate, up-to-date
information. The story of Baby Doe resulted, in part, from outdated impressions
about Down syndrome (26). The large number of studies, many now outdated, with
small sample sizes or preselected samples, clouds prognostication about extremely
premature infants. Several studies document the different prognoses presented by
obstetricians versus pediatricians/neonatologists for the extremely premature infant
(60). Others reflect major differences in clinical approach and parental counsel
between intensivists and rehabilitation physicians for children who are or might
become ventilator dependent (61).
Other issues complicate the process relative to the parents' emotional stability and
consistency. Especially in the immediate newborn period, the infant and parents
especially the mothermay have been separated, sometimes by hundreds of miles.
In some cases the parents may not even have seen the baby before it was
transferred to a neonatal unit. The mother may at times be ill herself or recovering
from anesthesia. The parents may have varying degrees of support from other family
members. The father of the baby may or may not be involved. The birth of a
critically ill or severely malformed infant is often seen as a loss (of the normal-term
infant they expected); the family may actually be grieving or dealing with anger,
depression, denial, or fear. Often the information presented to the family is complex
and difficult to understand, especially at first hearing. Time constraints may further
complicate the process, for example, in the delivery suite or for surgical emergencies.
Standards for Decision Making

The usual broad standard for decision making for this population is to decide what is
in the child's best interests. However, defining best interests can be difficult
and reflects a basic problem with the use of terms that may carry very different
meanings for different individuals.
Weir and Bale (62) suggest eight variables for evaluating best interest:
The severity of the patient's medical condition
The achievability of curative or corrective treatment
The important medical goals in the case (such as prolongation of life, relief of
pain, or amelioration of disabling conditions)
The presence of serious neurologic impairments
The extent of the infant's suffering
The multiplicity of other serious medical problems
The life expectancy of the infant
The proportionality of treatment-related benefits and burdens
Is the best-interests standard not, in fact, quite subjective; potentially defined

differently by clinicians and family? How do we define how much potential pain and
suffering
P.15
is acceptable for a given outcome, especially when the outcome may not be precisely
predicted? How do we define an acceptable outcome? Is being alive but vegetative
acceptable? Is a 20% or 40% or 70% risk of death or developmental disability
acceptable? Is it in the best interests of a child with trisomy 21 to have surgery for
duodenal atresia, or the child with trisomy 18 to have open heart repair for
congenital heart disease? How differently should one consider a 25% risk of a poor
outcome (e.g., for the very-low-birth-weight [VLBW] infant) versus a known, defined
poor prognosis (e.g., for trisomy 13)? Silverman (63,64) has proposed abandoning
the term best interests and adopting the concept of a standard of reasonableness
proposed by Veatch. Here the issue is how far the technology is pursued to
accomplish reasonable ends defined by those most directly affected by the decisions
the parents. Well-being extends beyond organic and physiologic function into the
realms of social, legal, occupational, religious, aesthetic, and other aspects of life.
Are best interest or reasonableness standards the same as quality of life
standards? The latter term is often viewed negatively because of the even greater
subjective element. How does one define a good or bad quality of life? Does defining
a quality of life that is poor reflect potential for discrimination against individuals
with disabilities? In response to the death of an infant with trisomy 21 whose parents
refused surgery for tracheoesophageal fistula, the federal government in 1982,
under pressure from right-to-life advocates and advocacy groups for the
handicapped, proposed regulations to prohibit hospitals from withholding care from
newborns. The Baby Doe rules stated:
All such disabled infants must, under all circumstances, receive appropriate
nutrition, hydration, and medication.
All such disabled infants must be given medically indicated treatment.
There are three exceptions to the requirement that all disabled infants must
receive treatment, or, stated in other terms, three circumstances in which
treatment is not considered medically indicated. These circumstances are:
The infant is chronically and irreversibly comatose.
The provision of such treatment would merely prolong dying, not be

effective in ameliorating or correcting all of the infant's life-threatening
conditions, or otherwise be futile in terms of the survival of the infant.
The provision of such treatment would be virtually futile in terms of the

survival of the infant, and the treatment itself under such circumstances
would be inhumane.
The physician's reasonable medical judgment concerning the medically

indicated treatment must be one that would be made by a reasonably prudent
physician who is knowledgeable about the case and the treatment possibilities
with respect to the medical conditions involved. It is not to be based on
subjective quality of life or other abstract concepts (27).
In the commentary for the rules, certain conditions were identified as not requiring
treatment, including anencephaly, trisomy 13, and extremely low-birth-weight.
Quality-of-life terminology carries negative connotations from the Baby Doe era.
Quality of life, just as best interests, can be very subjective; what one clinician
sees as a good quality of life may be unacceptable to another professional or a
parent. A family may see the life of a child who is profoundly visually handicapped
but with normal intelligence as qualitatively poor. Although many families do not feel
burdened with a moderately mentally retarded child, others consider learning
disability with normal intelligence unacceptable. Clinicians who work with
developmentally impaired children often have a very different appraisal than do
laypersons of the quality of life in this patient population.
However, it may be possible to use quality of life as a basis for decision making
when considering more fundamental issues. Richard McCormick (65) proposes a
minimal condition for defining quality: the capacity for experience or social
interrelating. If the condition is not met, as with anencephaly, treatment is not
required. Coulter and associates (66) define interests that would constitute a
minimal quality of life as:
Freedom from intractable pain and suffering. Mental retardation, paralysis,

cerebral palsy would not be considered physical suffering; dyspnea or
intractable physical pain would.
Capacity to experience and enjoy lifethe ability to enjoy food, warmth, or the
caring touch of another; the ability to give or receive love.
Expectation of continued lifeheroic treatment, when death will likely occur in

a few weeks or months, may be cruel.
There has been considerable debate about how much attention should be paid in the
decision-making process to interests other than the child's. Parents may be
overwhelmed with the prospects of chronic medical care, financial burdens, difficulty
in raising a handicapped child, need for special education, and harm to other children
in the family. Some parents may focus on their own psychological and financial
interests, protecting their lifestyle and other children at home. Fost (67) suggests,
The history of childhood is one that does not support idyllic notions of parents as
decision makers for their children. It is nave to posit an identity of interest between
infant and parent [in all situations]. Parents guard their own interests, those of the
family as a unit, and those of current and future siblingsall of which may be
gravely threatened by the newborn (68). Most would argue that the parents should
not refuse treatment that would be in the infant's interests in order to avoid burdens
to the family (59,69,70). These considerations are rarely allowed to play a role in
surrogate decisions for older children and adults. Others suggest that the impact of a
decision on the welfare of the family may be taken into account. Silverman (71)
comments, parents of a badly damaged baby often resent that their family is
P.16
required to pass a sacrifice test to satisfy the moral expectations of those who do not
live, day by day, with the consequences of diffuse idealism. It is easy to demand
prolongation of eachlife that requires none of [the clinician's] own resources to
maintain that life later. Ross (72), using the concept of the intimate family,
proposes a model of constrained parental autonomy, where the parent should be
guided by the child's well being, but is not obligated to disregard all personal
interests of themselves or other children in order to fulfill the child's needs and
interests. This is an issue that requires sensitivity to the parents and family situation,
but at the same time balancing the short- and long-term needs of the child. A New
York Academy of Medicine conference reached the following conclusion:
Although parents may have legitimate concerns about the
effect of treatment decisions on themselves and their other
children, the desire to avoid emotional, financial or other
hardships cannot justify the denial of clearly beneficial
medical care to an ill or injured child.If parents are unable
or unwilling to provide essential medical treatment,
healthcare professionals should first assure that social
counseling and supports are made available to the family to
assist them. If the parents remain unwilling to consent to
the needed medical treatment, then we must utilize legal
mechanisms to ensure social support or supervision to
provide those treatments which are clearly in the best
interests of the child. (73)
What is the clinician's role in decision making for the infant? Do pediatric
practitioners have a stronger responsibility for the decisions made for the child than
the clinician caring for an adult patient? Bartholome (74) defines a role that imposes
legal and ethical duties and obligations which exist independently of any parental
wishes, desires or consentings. The New York Academy of Medicine states the
clinician must maintain an independent obligation to protect the child's
interest (73). The Academy of Pediatrics defines providers' responsibilities as
follows:
Proxy consent poses serious problems for providers.
[They] have legal and ethical duties to their child patients to
render competent medical care based on what the patient
needs, not what someone else expresses The
pediatrician's responsibilities to his or her patient exist
independent of parental desires or proxy consent (75).
Futility
Conflict also may arise when the parents demand care that the clinician feels is
inappropriate, futile, and potentially harmful to the child. The clinician should not be
required to provide care that he or she considers harmful or unethical. The problem
again is one of definition: What is futile care? Who defines futility? What may be
futile in the eyes of the clinician may be beneficial for the child from the parents'
point of view (76). Being kept alive on the ventilator with no chance of recovery and
minimal or no social interaction may be sufficient for the parents to continue. The
case of Baby K complicated this question even further when the federal court upheld
the mother's demand for resuscitation and aggressive treatment of her anencephalic
infant (22). While the application of this case to other futility cases may be limited
because the court's decision was based on the Federal Emergency Treatment and
Labor Act, which requires emergency care for life-threatening situationsit has
clearly raised the level of sensitivity to parental demands. Again, ethics consultation
or ethics committee involvement may facilitate resolution. The clinician may transfer
the patient to another clinician (or institution) who is willing to provide the care.
Some institutions have developed procedures for case review that may conclude that
life-sustaining treatments can be withdrawn, even against the objections of the
family (77). As noted earlier, the health care system and society have usually been
unwilling to consider the important justice issues of the financial costs to society,
both immediate and long-term, and of limitation of resources.
The Extremely Low-birth-weight Infant

Infants who are born at gestational ages of less than 27 weeks or who weigh less
than 800 g have been the focus of considerable professional, legal, and media
debate about withholding resuscitation in the delivery room, parental requests for
aggressive care or no resuscitation, mortality rates, and quality of outcome (both
short-term into later infancy and long-term to school age). There are hundreds of
outcome studies and reviews regarding this population, several of which highlight
the critical issues for this population (78,79,80). The problems of decision making
(discussed previously) are painfully apparent. For example, the clinician has an
obligation to accurately inform the parents expecting an ELBW infant about the
clinical situation, prognosis, anticipated clinical course, and so forth. While there is
certainly no lack of outcome data, it is clear that there are wide variations in
outcome from one study to another. The studies are also faulted by the sample size
of infants at any gestation, by the selection criteria (were all live born infants
included or only those admitted to the NICU, inborn or outborn, etc), and by
decisions about viability and nonresuscitation that then impact mortality of the
sample as a whole. Mortality will, by definition, be higher if most infants were not
resuscitated because of clinician or family preference. Drop out and incomplete
followup, small numbers, and variation in definitions of morbidity impact the
usefulness of long-term outcome studies. With such a wide range of study results,
there is a risk or tendency for an individual clinician to choose those that support his
or her philosophical approach (conservative or aggressive). Fortunately, there are

several large, collaborative studies now available that lessen the impact of
uncontrolled variables and provide more reliable outcome data (81,82,83,84,85,86).
El-Metwally and associates (87) report significantly different survival rates when all
ELBW infants are aggressively managed. Ideally, clinicians should have mortality and
morbidity data for their own institution in addition to national statistics to share with
the parents.
P.17
Use of gestational criteria for decision making is reasonable because studies show a
very dramatic correlation to survival (81,82,88,89,90,91,92,93,94,95). Each
additional gestational week adds a significant increase in survival percentage.
However, in many cases both obstetrical estimates using ultrasound and neonatal
estimates using physical exam scoring are not accurate enough to define 23 versus
24 or 24 versus 25 weeks gestation (96,97,98). The current practice of quoting
gestational age in fractions of a week (e.g., 23 5/7) provides a sense of accuracy
that is not justified by the existing data.
Birth weights are accurately measured and have been correlated to outcome
(35,88,89,99,100,101,102,103). However, for gestations in the 22- to 25-week
range, there is an extremely wide variation in birth weights. Using birth weights also
ignores the impact of intrauterine growth restriction. If study populations include
growth-restricted infants, outcome data may be overly optimistic. If the decision
makers consider only birth weight, the predicted prognosis may be overly pessimistic.
Guidelines for Resuscitation/Care

Several professional groups have developed criteria or guidelines for decision making
for this population. In 1994, the Canadian Paediatric Society and the Society of
Obstetricians and Gynaecologists of Canada defined relatively specific
recommendations:
At 22 completed weeks of gestation, they suggest
treatment should be started only at the request of fully
informed parents or if it appears that the gestational age
has been underestimated. At 23 to 24 completed weeks of
gestation, they emphasize a significant role for parental
wishes, the option of resuscitation, and the importance of
discussion with parents about the need for flexibility in
deciding to start or withhold resuscitation, depending on the
infant's condition at birth. Finally, at 25 weeks gestation
they state resuscitation should be attempted for all infants
without fatal anomalies (104).
Discussions by neonatologists in Colorado, in collaboration with the Colorado
Collective for Medical Decisions, developed guidelines that define comfort care as the
only appropriate choice for the 22 weeks' gestation infant. At 23 weeks, most would
advise comfort care; but, if the parents understood the high risks involved, would be
willing to initiate a course of intensive care. At 24 weeks, the neonatologists were
able to support either decision, as long as a collaborative process with good
information occurred. At 25 weeks, they were uncomfortable with withholding
intensive care, and some, but not all, were willing to support a parental request for
comfort care, if there had been a good parent education process and an effort to
collaborate with the parents (Hulac P. Colorado Collective for Medical Decisions,
personal communication, January 2000). Tyson and associates (80) have suggested
developing fairly detailed guidelines based on outcome data. Female and/or smallfor-gestational-age infants would be resuscitated at lower birth weights than male
and/or appropriately grown infants. Antenatal steroids would lower the
recommended weight further. They recommend mandatory resuscitation when the
data reveal a greater than 50% chance of survival without severe sequelae and
optional resuscitation when the chance is 25% to 49%. Resuscitation for infants with
a less than 25% chance of survival without severe sequelae would be seen as
investigational. Neonatal Resuscitation Program, which is supported by the American
Academy of Pediatrics and the American Heart Association, suggests that
noninitiation of resuscitation for newborns of less than 23 weeks' gestational age
and/or 400 g in birth weight is appropriate (105). Interestingly, clinicians have
overinterpreted these criteria as either demanding resuscitation at 23 weeks or 400
g or as stating it is inappropriate to resuscitate at less than 23 weeks or 400 g (80).
Discussions with the family should include the uncertainty of gestational age if that is
at issue, the advantage of assessing and then making decisions, the possibility of
withdrawal of support if it is apparent that there will be a bad outcome, and the
importance of ongoing assessment and ongoing communication. To avoid confusion
and conflict, the obstetrician and the pediatrician/neonatologist should coordinate
their approaches. For the extremely immature infant, the parents should be aware
that there may be nothing that can be done. There may be physical limitations to the
resuscitation. In most circumstances, firm decisions about how to proceed should be
avoided. Likewise, vague terms such as no heroic measures or do everything can
cause confusion and conflict. Several recent papers have discussed the importance of
antenatal and intrapartum discussions (106,107). When possible, the use of
educational materials, tours of the NICU, or videos of NICU experiences may further
the parents understanding of the situation. The Colorado Collective for Medical
Decisions has developed a video especially for the parents of an ELBW infant (108).
Infants with Severe Encephalopathy/Brain Death

Infants with perinatal hypoxic-ischemic injury may have lethal multisystem failure.
However, more often they are able to survive despite severe central nervous system
(CNS) injury. Prolonged, difficult-to-control seizures, hypotonia; poor feeding;
apnea; and inability to maintain body temperature are poor prognostic signs for
developmental outcome. Imaging studies (computed tomography [CT], magnetic
resonance imaging [MRI]) as well as an electroencephalogram (EEG) may add

additional prognostic data. Some may be ventilator dependent because of poor
respiratory drive. Others may require gastrostomy feedings. Prognosticating may be
difficult early after the insult, but in some circumstances the clinical data
convincingly show that the prognosis is extremely poor. Do not resuscitate orders
and/or withdrawal from the ventilator are appropriate decisions at this time.
Infants with Severe Congenital Malformations/

Chromosomal Anomalies
In a series of articles and commentaries beginning in 1973, pediatricians Duff and
Campbell (20) reported on a practice
P.18
of selective nontreatment at Yale-New Haven Hospital that resulted in the deaths of
43 infants over a period of 30 months. Physicians and parents decided together, with
the physicians sometimes yielding to parental wishes. Surveys at that time
confirmed that many pediatricians and pediatric surgeons accepted selective
nontreatment of seriously impaired newborns. In a survey of New England
pediatricians, 54% did not recommend surgery for an infant with Down syndrome
and duodenal atresia, and 66% would not recommend surgery for an infant with a
severe case of spina bifida (109). Twenty-two percent of surveyed pediatricians in
the San Francisco area favored nontreatment for infants with trisomy 21 who had no
complications, and more than 50% recommended nontreatment for Down syndrome
with duodenal atresia (110). With the Baby Doe case, advocates for the
handicapped, developmental pediatricians, and family support groups promoted a
change in approach for infants with trisomy 21 and meningomyelocele. Treatment
and approaches to rehabilitation and special education have improved; the current
standard of care in the United States is that infants with trisomy 21 receive the same
treatment that is indicated for a child with a similar medical or surgical condition who
does not have trisomy 21.
There are other conditions associated with extremely poor prognoses for survival or
for reasonable quality of life, including anencephaly and other severe CNS
developmental anomalies, trisomy 13, and trisomy 18 (111). Most of these infants
will not survive beyond the first few months of life. Those who do survive have
severe neurodevelopmental delays. Many have life-threatening conditions in the
immediate newborn period and require life-sustaining treatment to prolong survival.
Most would agree that a recommendation to withdraw or withhold aggressive
therapy and provide comfort care is appropriate. The Neonatal Resuscitation
Program guidelines suggest that noninitiation of resuscitation is appropriate for
confirmed diagnoses of anencephaly or trisomies 13 or 18 (105). Hopefully, every
parent of such an infant would have appropriate counseling about the diagnosis; and
in this circumstance, a definite plan for care in the delivery room (including no
cesarean section for fetal distress) and during the newborn period can be very
helpful and comforting to the parent. Again, the guideline does not state that it is
inappropriate to resuscitate. When a condition is not diagnosed until after birth,
there is the usual obligation for the clinician to provide the parents with appropriate,
accurate information in a timely fashion. Some families wish to pursue aggressive
treatment, and their wishes should be considered. Internet sites and parent support
groups have expanded the options available to families for treatment (112). While
cardiac surgery for infants with trisomy 18 was formerly quite uncommon, there are
now increasing requests for such procedures. The parents should be well informed,
and a plan should be developed for the care, including how to proceed if the infant
does not tolerate the procedure, becomes ventilator dependent, and so forth. If the
clinician is not willing to offer this level of care, efforts should be made to transfer
care to another physician or facility.
Artificial Nutrition and Hydration

While most agree that withdrawing artificial nutrition and hydration from hopelessly
ill, persistently vegetative or dying patients may be appropriate (113,114), this issue
becomes less well-defined in the newborn population, who are all dependent on
others for feeding and where tube feeding is the standard of care for many until they
develop the ability to feed. There are circumstances, however, where discontinuing
intravenous nutrition or tube feedings may be appropriate (115,116,117,118). If the
infant is actively dying, there is no benefit to feeding or hydration, and experience
would suggest that there may be less suffering from excess secretions if the patient
is somewhat dehydrated (113). Coma and persistent vegetative state are rarely
defined in the newborn, but severe neurologic impairment is, unfortunately, not rare.
However, exact prognostication about recovery is difficult, and some infants who
need feeding support initially may develop the ability to feed normally after a period
of recovery. When clinical evaluation strongly suggests an extremely poor neurologic
outcome, it may be reasonable to withhold feedings (118). Factors to consider
include whether the infant is able to feed at all or whether the infant demonstrates
hunger or satiety. Infants may have medical conditions that preclude enteral feeding
(e.g., short gut syndrome), and, therefore, they are dependent on total parenteral
nutrition (TPN). TPN has many potential complications including loss of venous
access, infection, and liver toxicity. If the possibility of slow recovery of absorptive
ability or successful bowel transplant is poor, withdrawal of TPN may be an option
after consideration of the criteria discussed earlier. Withdrawal of feeding often
presents more psychological and emotional angst than withdrawal of other lifesustaining technology because of the cultural connotations of feeding and
starving (119). While the Baby Doe regulations specifically list hydration and
nutrition as required, the term appropriate is also used and therefore open to
interpretation.
Palliative/Comfort Care and Pain Control

When aggressive care or life-sustaining technology is withheld or withdrawn, care
does not stop. The clinician has an obligation to the patient and the parents to
provide comfort care, warmth, feedings, if desired, and emotional support. The
clinical staff must attend to relief of symptoms and pain control (120). Care may in
fact become more intensive in terms of time spent with the family. Catlin and
Carter (121) have developed a protocol for neonatal end-of-life palliative care.
Clinicians and, more recently, regulators have highlighted the importance clinically
and ethically of identification and treatment of pain in the neonate (122,123,124).
High-Technology Home Care

With technological advances in the hospital has come the use of this same
technology in the hometube feedings,
P.19
oxygen, and mechanical ventilators. This development has raised numerous ethical
issues both on an individual patient basis and related to the general issues of
resource allocation, funding, and health planning (125, 126). How a family and the
clinician reach the conclusion to proceed or not proceed with long-term home care,
especially mechanical ventilation, requires careful attention to the details of
adequate information, parent education and training, planning for professional
assistance in the home, adequate support and funding, opportunity for respite care,
and so forth. Studies suggest that there is wide variation in physician practice
regarding offering home ventilation, with families then receiving unbalanced
information (61). With the growing incidence of ventilator use in the home, parent
support groups and the Internet offer a very positive bias. Physicians may feel that
this is fiscally inappropriate, or that the child's quality of life may be poor, or that the
family will be unable to manage at home. Home ventilation for older children and
adults with a variety of diseases provides a great deal of satisfaction for many of
them (127). Every child and family is unique in their clinical situation, prognosis,
family goals and stamina, and resources. On the other hand, just because the
technology for home ventilation is now available does not require its use in every
circumstance. If the family is unable or unwilling to provide less intensive care in the
home (e.g., tube feeding or oxygen), other options must be identified. It is
inappropriate to discharge an infant with heavy ongoing health care needs home to a
family that is ill prepared and does not have adequate support from outside services.
HIV SCREENING OF PREGNANT WOMEN AND

NEWBORN INFANTS
The diagnosis and treatment of HIV infection has changed dramatically in a short
time period. A decade ago, diagnosis often involved a screening test with
confirmation testing to follow; diagnosis can now be quite accurately made in an
hour or less. Antibody testing of the newborn is, in reality, antibody testing of the
mother; it is now possible to test the newborn for viral DNA. The critical treatment
element relates to the ability to prevent transmission of infection from mother to
baby, by treatment of the mother during the pregnancy and labor and of the
newborn in the first several weeks after birth. There is also evidence that treatment
limited to labor and the postnatal period only, or even to the infant postnatally, may
significantly decrease the risk of transmission. It has therefore become significantly
more important to know the mother's HIV status early in the gestation and certainly
during labor. Public Health recommendations promote universal, voluntary, routine
testing of the mother, using an opt-out strategy for consentwe will test you for
HIV unless you specifically say you refuse (128,129,130,131). Mandatory testing
potentially violates the mother's rights to privacy and self-determination and, even
today, may expose her to social consequences and discrimination. Being pregnant
should not require an individual to forgo rights. Once the infant has been born, some
would argue that the advantages of early diagnosis and treatment outweigh the risks
to the mother. Some jurisdictions have mandated newborn screening if the mother's
status is not known. However, most professionals would recommend routine, but
voluntary, testing. The mother's status will be defined, but the benefits to the baby
would outweigh the risks to the mother (131).
SUBSTANCE ABUSE IN PREGNANCY

Substance abuse during pregnancy has incited a great deal of media and legal
attention. Prosecutors and governmental agencies have charged mothers with
homicide and child abuse, imprisoned them for the duration of their pregnancies,
and removed their children from their custody (132,133,134). Complicating the
ethical and legal issues is a wide range of scientific opinion about the short- and longterm effects of drug exposure in utero on the child (135,136). Recent studies
suggest that the rearing environment is much more critical for the child's
development (135). Routine drug screening of pregnant women without their
consent has been ruled unconstitutional (137). Punitive approaches in general are
counterproductive, potentially keeping women away from prenatal care (132).
Routine screening of newborns should be limited to situations in which there is a
clinical history or signs or symptoms compatible with drug exposure or withdrawal.
Testing infants solely to identify substance-abusing mothers should be avoided. The
approach to mothers and babies should be supportive, voluntary, and nonpunitive
(138).
GENETIC TESTING
In the past two decades, there has been a dramatic increase in knowledge about the
human genome and the ability to screen for the presence of certain genetic diseases,
carrier states, traits, or predispositions to disease. Genetic testing of a newborn
raises different issues than routine medical testing performed as part of clinical care.
The genetic information relates not only to the individual tested but also to other
members of the family. Results of testing may have psychological (guilt, anxiety),
social (stigmatization, discrimination), and financial impact (insurability, employment
considerations) with long-term consequences. Some genetic information defines
risks only and does not predict with any certitude a specific condition or outcome.
Finally, many defined conditions do not necessarily have effective therapy. For these
reasons, concerned professional organizations have developed policies and
guidelines related to these issues. Testing should involve effective counseling,
informed consent, and attention to confidentiality. Genetic testing to confirm a
medical diagnosis would be an appropriate component of medical care, for example,
DNA analysis for cystic fibrosis in an infant with meconium ileus or chromosomal
analysis for a newborn with clinical
P.20
features of Down syndrome. Testing for conditions that may benefit from monitoring,
prophylaxis, or treatment in an otherwise healthy individual (e.g., familial
hyperlipidemia) may also be in the child's best interest. However, carrier screening
for diseases with no risk to the pediatric patient should be avoided. Likewise,
screening for adult-onset conditions should be deferred until adulthood or until the
mature adolescent is able to consent (139,140).
MAKING DECISIONS REGARDING THE

APPROPRIATENESS OF INTER-INSTITUTIONAL
TRANSFER
The transfer of neonates from referring to referral hospitals potentially adds an
additional layer of ambiguity to the process of ethical decision making, particularly if
there is uncertainty on the part of the NPM team at the referral hospital regarding
the level of treatment received by the infant since birth. For example, a 25-week
gestation premature infant who has been set aside as nonviable but is found
gasping 1 hour later and resuscitated is unlikely to benefit from transfer to another
institution for aggressive support.
A regional network of hospitals can reduce this problem by having a common set of
criteria for delivery room practices and inter-institutional referral. Fletcher and Paris
(141) suggested the following mnemonics to assist in formulating these policies:
ACUTE (Acute, Critical, Unexpected, Treatable, and Easily diagnosed)

Infants included in this group are those with prematurity and respiratory
distress syndrome; ELBW infants of gestational age known to be 25 weeks or
greater; term or preterm infants with sepsis, pneumonia, or meningitis; and
infants with surgically correctable malformations.
The authors place these infants in a definite category for transfer to an NICU
capable of providing the level of care required.
UNSURE (UNknown disease, SUspected REsponse). This group includes
preterms of 23 to 24 weeks gestational age or with very low-birth-weight and

uncertain gestational age, infants with severe birth asphyxia, and any infant
with an unexplained disease or syndrome that requires further diagnostic
efforts. Within this group there will be a significant number of infants for whom
the response to treatment will be unpredictable. These infants should be given
full medical care until the diagnosis has been made or the response to
treatment is clear; decisions to omit care or not to transfer should not be
made precipitously.
KNOT (Known, NOt Treatable). Although only a small number of infants fit
into this category, treatment decisions for this group frequently take a
disproportionate amount of time. This group includes neonates with
anencephaly and those with lethal genetic disorders such as trisomy 13 and
trisomy 18.
Transfer of neonates with anencephaly for aggressive support is not indicated;
transfer of those with lethal genetic defects is not indicated if there are
facilities for accurate diagnosis and appropriate care and counseling at the
hospital of birth. When diagnostic facilities are not available at the birth
hospital, onsite consultation by a specialist from the referral hospital is an
appropriate alternative to transferring the infant.
REFERENCES
1. Carter BS, Stahlman M. Reflections on neonatal intensive care in the US: limited
success or success with limits. J Clin Ethics 2001;12:215222.
2. Brenner S. Humanity as the model system. Science 2003;302:533.
3. Martin JA, Hamilton BE, Ventura SJ, et al. Births: final data for 2001. National
Center for Vital Statistics Reports 2002;51:1.
4. National Foundation March of Dimes. Available online at http://www.

marchofdimes.com
5. Olivennes F, Fanchin R, Ledee N, et al. Perinatal outcome and developmental

studies on children born after IVF. Human Repro Update 2002;8:117128.
6. Keith LG, Oleszczuk JJ, Keith DM. Multiple gestation: reflections on

epidemiology, causes and consequences. Int J Fert Womens Med 2000;45:206
209.
7. Schieve LA, Meikle SF, Ferre C, et al. Low and very low-birth-weight infants
conceived with use of assisted reproductive technology. N Engl J Med
2002;346:731737.
8. The President's Council on Bioethics. Reproduction and responsibility: the

regulation of new biotechnologies. 2004. Available online at http://bioethics.gov/
reports/productionandresponsibility
9. Hansen M, Kurinczuk JJ, Bower C, et al. The risk of major birth defects after
intracytoplasmic sperm injection and in vitro fertilization. N Engl J Med
2002;346:725730.
10. Beauchamp TL, Childress JF. Principles of biomedical ethics. 5th ed. New York:
Oxford University Press, 2001:231.
11. Buchanan AE. Is there a medical profession in the house? In: Spece RG,
Shimm DS, Buchanan AE, eds. Conflict of interest in clinical practice and research.
New York: Oxford University Press, 1996:105.
12. Emanuel EJ. The ends of human life: medical ethics in a liberal polity.
Cambridge, MA: Harvard University Press, 1991.
13. Thompson LA, Goodman DC, Little GA. Is more neonatal intensive care always
better? Insights from a cross-national comparison of reproductive care. Pediatrics
2002;109:10361043.
14. Committee on Perinatal Health. Toward improving the outcome of pregnancy:

recommendations for the regional development of maternal and perinatal health
services. White Plains, NY: National Foundation March of Dimes, 1976.
15. Benjamin M. Philosophy and this actual world. Lanham, MD: Rowman &
Littlefield, 2003:112.
16. Dewey J. Human nature and conduct. Carbondale, IL: Southern Illinois
University Press, 1988:164.
17. Paris JJ, Ferranti J, Reardon F. From the Johns Hopkins Baby to Baby Miller:
what have we learned from four decades of reflection on neonatal cases. J Clin
Ethics 2001;12:207.
18. Howe EG. Helping infants by seeing the invisible. J Clin Ethics 2001;12:191
204.
19. Cuttini M and the Euronic Study Group. The European Union collaborative
project on ethical decision making in neonatal intensive care (Euronic): findings
from 11 countries. J Clin Ethics 2001;12:290296.
20. Stinson R, Stinson P. The long dying of Baby Andrew. Boston, MA: Little
Brown, 1983.
21. Buber M. Between man and man. New York: Macmillan, 1965:164. Smith RG,
translator.
22. In re Baby K, 832 F. Supp. 1022 (E.D. Va. 1993); In re Baby K, 16 F. 3d 5900
(4th Cir.).
P.21
25. Brown-Ballard J. Consistency, common morality, and reflective equilibrium.
Kennedy Inst Ethics J. 2003;13:231258.
26. State ex rel. Infant Doe v Baker, No. 482 S 140 (Ind. May 27, 1982).
27. U.S. Department of Health and Human Services. Nondiscrimination on the

basis of handicaps: procedures and guidelines relating to health care for
handicapped infants. Federal Register 1984 Jan 12:49:622654.
28. Leikin S, Moreno JD. Pediatrics ethics committees. In: Cassidy RC, Fleischman
AR, eds. Pediatric ethicsfrom principles to practice. Amsterdam: Harwood
Academic Publishers, 1996:51.
29. Lo B. Behind closed doors: promises and pitfalls of ethics committees. N Engl J
Med 1987;317:4650.
30. Weir RF. Pediatric ethics committees: ethical advisers or legal watchdogs? Law
Med & Hlth Care 1987;15:99109.
31. Duff RS, Campbell AGM. Moral and ethical dilemmas in the special care
nursery. N Engl J Med 1973;289:890894.
32. Paris JJ, Schreiber MD, Reardon F. The emergent circumstances exception to
the need for consent: the Texas Supreme Court ruling in Miller v HCA. J Perinatol
2004;24:337342.
33. Tooley M: Abortion and infanticide. Philosophy and Public Affairs 1972;2:37
65.
34. Engelhardt HT: The foundations of bioethics. New York, Oxford University
Press, 1986:116119.
35. May WF. Parenting, bonding, and valuing the retarded. In: Kopelman LM,
Moskop JC, eds. Ethics and mental retardation. Dordrecht, The Netherlands: D
Reidel, 1984;141160.
36. Blustein J. The rights approach and the intimacy approach: family suffering
and care of defective newborns. Mount Sinai J Med 1989;56:164167.
37. Ross LF: Children, families and health care decision-making. Oxford, England:
Clarendon Press, 1998:47.
38. Rhoden NK. Treating Baby Doe: the ethics of uncertainty. Hastings Cent Rep
1986;16:3442.
39. Fischer AF, Stevenson DK. The consequences of uncertainty: an empirical

approach to medical decision making in neonatal intensive care. JAMA
1987;258:19291931.
40. Kraybill EN. Ethical issues in the care of extremely low-birth-weight infants.
Semin Perinatol 1998;22:207215.
41. Committee on Fetus and Newborn, American Academy of Pediatrics. The

initiation or withdrawal of treatment for high-risk newborns. Pediatrics
1995;96:362363.
42. Beauchamp TL, Childress JF. Principles of medical ethics, 5th ed. New York:
43. Fletcher JC: The decision to forgo life-sustaining treatment when the patient is
incapacitated. In: Fletcher JC, Lombardo PA, Marshall MF, et al., eds. Introduction
to clinical ethics. Frederick, MD: University Publishing Group, 1997:158.
44. Doran MW, Vaness-Meehan KA, Margolis LH, et al. Delivery room resuscitation
decisions for extremely premature infants. Pediatrics 1998;102:574582.
45. Volpe JJ. Guidelines for the determination of brain death in children. Pediatrics
1987;80:293297.
46. Farrell MM, Levin DL. Brain death in the pediatric patient: historical, medical,
religious, cultural, legal and ethical considerations. Crit Care Med 1993;21:1951
1965.
47. Ashwal S. Brain death in the newborn. Clin Perinatol 1997; 24:859882.
48. Council on Ethical and Judicial Affairs, America Medical Association.

Euthanasia. In: Code of medical ethics: current opinions with annotations. Chicago
American Medical Association, 2002:83.
49. Committee on Bioethics, Committee on Hospital Care, American Academy of

Pediatrics. Palliative care for children. Pediatrics 2000;106:351357.
50. Cuttini M, Casotto V, Kaminski M, et al. Should euthanasia be legal? An

international survey of neonatal intensive care units staff. Arch Dis Child Fetal
Neonatal Ed 2004;89:F19-F24.
51. Sklansky M. Neonatal euthanasia: moral considerations and criminal liability. J

Med Ethics 2001;27:511.
52. Edwards SJ. The distinction between withdrawing life-sustaining treatment

under the influence of paralyzing agents and euthanasia. The doctrine of double
effect is difficult but not impossible to apply. BMJ 2001;323:390391.
53. Harrison H. Making lemonade: a parent's view of quality of life studies. J Clin
Ethics 2001;12:239250.
54. Saigal S. Perception of health status and quality of life of extremely low-birth
weight survivors. The consumer, the provider, and the child. Clin Perinatol
2000;27:403419.
55. Saigal S, Burrows E, Stoskopf BL, et al. Impact of extreme prematurity on

families of adolescent children. J Pediatr 2000;137: 701706.
56. Saigal S, Pinelli J, Hoult L, et al. Psychopathology and social competencies of

adolescents who were extremely low-birth-weight. Pediatrics 2003;111:969975.
57. Paris JJ, Schreiber MD. Parental discretion in refusal of treatment for
newborns: a real but limited right. Clin Perinatol 1996;23: 573581.
58. Bartholome WG, The Child-Patient: Do Parents Have the Right to Decide. In
Spicker SF, Healey JM, and Engelhardt HT, eds. The Law-Medicine Relation: A
Philosophical Exploration. Boston, MA: Reidel, 1981:271.
59. Weir R: Selective treatment of handicapped newborns: moral dilemmas in

neonatal medicine. New York: Oxford University Press, 1984.
60. Haywood JL, Goldenberg RL, Bronstein J, et al. Comparison of perceived and
actual rates of survival and freedom from handicap in premature infants. Am J
Obstet Gynecol 1994;171:432439.
61. Hardart MKM, Truog RD. Spinal muscular atrophytype 1. Arch Dis Child
2003;88:848850.
62. Weir RF, Bale JF Jr. Selective nontreatment of neurologically impaired

neonates. Neurol Clin 1989;7:807821.
63. Silverman WA, Medical decisions: an appeal for reasonableness. Pediatrics

1996;98:11821184.
64. Veatch RM. Abandoning informed consent. Hastings Cent Rep 1995;25:512.
65. McCormick RA. To save or let die: the dilemma of modern medicine. JAMA
1974;229:172176.
66. Coulter DL, Murray TH, Cerreto MC. Practical ethics in pediatrics. Curr Probl
Pediatr 1988;18:168169.
67. Fost N. Parents as decision makers for children. Prim Care 1986;13:285293.
68. Dellinger AM, Kuszler PC. Infants: public-policy and legal issues. In: Reich WT,
ed. Encyclopedia of bioethics. New York: Simon & Schuster and MacMillan,
1995:1214.
69. Beauchamp TL, Childress JF. Principles of biomedical ethics. 5th ed. New York,
70. Fleischman AR, Nolan K, Dubler NN, et al. Caring for gravely ill children.
Pediatrics 1992;94:422439.
71. Silverman W. Overtreatment of neonates? A personal perspective. Pediatrics

1992;90:971976.
72. Ross LF. Children, families and health care decision-making. Oxford, England:
Clarendon Press, 1998:51.
73. Fleischman AR, Nolan K, Dubler NN, et al. Caring for gravely ill children.
Pediatrics 1992;94:422439.
74. Bartholome WG. Withholding/withdrawing life-sustaining treatment. In:

Burgess MM, Woodrow BE, eds. Contemporary issues in pediatric ethics. Lewiston
NY: Edwin Mellen Press, 1991:17.
75. Committee on Bioethics, American Academy of Pediatrics. Informed consent,

parental permission, and assent in pediatric practice. Pediatrics 1995;95:314317.
76. Carter BS, Sandling J. Decision making in the NICU: the question of medical
futility. J Clin Ethics 1992;3:142143.
77. Truog RD. Futility in pediatrics: from case to policy. J Clin Ethics 2000;11
(2):136141.
78. Boyle RJ, Kattwinkel J. Ethical issues surrounding resuscitation. Clin Perinatol
1999;26:779792.
79. Meadow W, Lantos JD. Ethics at the limit of viability: a premie's progress.
NeoReviews 2003;4:e157.
80. Tyson JE, Stoll BJ. Evidence-based ethics and the care and outcome of
extremely premature infants. Clin Perinatol 2003;30: 363387.
81. Wood NS, Marlow N, Costeloe K, et al. Neurologic and developmental disability
after extremely preterm birth. N Eng J Med 2000;343:378384.
82. Costeloe K, Hennessy E, Gibson T, et al. The EPICure study: outcomes to

discharge from hospital for infants born at the threshold of viability. Pediatrics
2000;106(4):659671.
83. Lemons JA, Bauer CR, Oh W, et al. Very low-birth-weight outcomes of the
National Institute of Child Health and Human Development Neonatal Research
Network, January 1995 through December 1996. Pediatrics 2001;107(1):e1.
84. Horbar JD, Badger GJ, Carpenter JH, et al. Trends in mortality and morbidity
for very low-birth-weight infants, 19911999. Pediatrics 2002;110(1 pt 1):143
151.
P.22
85. Vohr BR, Wright LL, Dusick AM, et al. Neurodevelopmental and functional
outcomes of extremely low-birth-weight infants in the National Institute of Child
Health and Human Development Neonatal Research network, 19931994.
Pediatrics 2000;105: 12161226.
86. Yu VYH. Developmental outcome of extremely preterm infants. Amer J

Perinatol 2000;17:5761.
87. El-Metwally D, Vohr B, Tucker R. Survival and neonatal morbidity at the limits
of viability in the mid 1990s: 22 to 25 weeks. J Pediatr 2000;137:616622.
88. Fanaroff AA, Wright LL, Stevenson DK, et al. Very-low-birth-weight outcomes
of the National Institutue of Child Health and Human Development Neonatal
Research Network, May 1991 through December 1992. Am J Obstet Gynecol
1995;173: 14231431.
89. Hack M, Friedman H, Fanaroff AA. Outcomes of extremely low-birth-weight

infants. Pediatrics 1998;98:931937.
90. Kilpatrick SJ, Schlueter MA, Piecuch R, et al. Outcome of infants born at 2426
weeks' gestation: I. survival and cost. Obstet Gynecol 1997;90:803808.
91. O'Shea TM, Klinepeter KL, Goldstein DJ, et al. Survival and developmental
disability in infants with birth weights of 501 to 800 grams, born between 1979
and 1994. Pediatrics 1997;100: 982986.
92. Emsley HCA, Wardle SP, Sims DG, et al. Increased survival and deteriorating
developmental outcome in 23 to 25 week old gestation infants. 19904 compared
with 19849. Arch Dis Child Fetal Neonatal Ed 1998;78:F99-F104.
93. Larroque B, Breart G, Kaminski M, et al. On behalf of the EPIPAGE study

group. Survival of very preterm infants: Epipage, a population based cohort study.
Arch Dis Child Fetal Neonatal Ed 2004;89:F139.
94. Lefebvre F, Glorieux J, St-Laurent-Gagnon T. Neonatal survival and disability

rate at 18 months for infants born between 23 and 28 weeks of gestation. Am J
Obstet Gynecol 1996:174:833.
95. Levene M. Is intensive care for very immature babies justified? Acta Paediatr
2004;93:144149.
96. Haidet KR, Kurtz AB. Routine ultrasound evaluation of the uncomplicated
pregnancy. In: Spitzer AR, ed. Intensive care of the fetus and neonate. St. Louis,
MO: Mosby-Year Book, 1996:45.
97. American Academy of Pediatrics, Committee on Fetus and Newborn, American

College of Obstetricians and Gynecologists, Committee on Obstetric Practice.
Perinatal care at the threshold of viability. Pediatrics 1995;96:974976.
98. Donovan EF, Tyson JE, Ehrenkranz RA, et al. Inaccuracy of Ballard scores
before 28 weeks gestation. J Pediatr 1999;135: 147152.
99. Lucey JF, Rowan CA, Shiono P, et al. Fetal infants: the fate of 4172 infants
with birth weights of 401500 gramsthe Vermont Oxford Network experience
(19962000). Pediatrics 2004;113: 15591566.
100. Sauve RS, Robertson C, Etches P, et al. Before viability: a geographically

based outcome study of infants weighing 500 grams or less at birth. Pediatrics
1998;101:438445.
101. Hack M, Horbar JD, Malloy MH, et al. Very low-birth-weight outcomes of the
National Institute of Child Health and Human Development Neonatal Network.
Pediatrics 1991;87:587597.
102. Piecuch RE, Leonard CH, Cooper BA, et al. Outcome of extremely low-birthweight infants (500 to 999 grams) over a 12-year period. Pediatrics
1997;100:633639.
103. Tyson JE, Younes N, Verter J, et al. Viability, morbidity and resource use
among newborns of 501800-g birth weight. JAMA 1996;276:16451651.
104. Fetus and Newborn Committee, Canadian Paediatric Society; Maternal Fetal
Medicine Committee, Society of Obstetricians and Gynaecologists of Canada.
Management of the woman with threshold birth of an infant of extremely low
gestational age. CMAJ 1994;151:547553.
105. American Academy of Pediatrics. Special considerations. In: Braner D,

Kattwinkel J, Denson S, et al., eds. Textbook of neonatal resuscitation, 4th ed. Elk
Grove Village, IL: 2000:7.
106. Halamek LP. Prenatal consultation at the limits of viability. NeoReviews

2003;4:e153.
107. Munro M, Yu VYH, Partridge JC, et al. Antenatal counseling, resuscitation

practices and attitudes among Australian neonatologists towards life support in
extreme prematurity. Aust N Z J Obstet Gynaecol 2001;41:275280.
108. Hulac P. Creation and use of You are not alone, a video for parents facing
difficult decisions. J Clin Ethics 2001;12:251252.
109. Todres ID, Krane D, Howell MC, et al. Pediatricians' attitudes affecting
decision-making in defective newborns. Pediatrics 1977;60:197201.
110. Treating the defective newborn: a survey of physicians' attitudes. Hastings

Cent Rep 1976;6:2.
111. Rasmussen SA, Wong LY, Yang Q, et al. Population-based analyses of

mortality in trisomy 13 and trisomy 18. Pediatrics 2003;111: 777784.
112. Support Organization for Trisomy 18, 13 and Related Disorders. Available
online at http://www.trisomy.org
113. Lynn J, Childress JF. Must patients always be given food and water? Hastings
Cent Rep 1983;13:1721.
114. American Academy of Pediatrics, Committee on Bioethics. Guidelines for

forgoing life-sustaining medical treatment. Pediatrics 1994;93:532536.
115. Nelson LJ, Rushton CH, Cranford RE, et al. Forgoing medically provided
nutrition and hydration in pediatric patients. J Law Med Ethics 1995;23:3346.
116. Cranford RE. Withdrawing artificial feeding from children with brain damage
is not the same as euthanasia or assisted suicide. BMJ 1995;311:464465.
117. Johnson J, Mitchell C. Responding to parental requests to forego pediatric

nutrition and hydration. J Clin Ethics 2000;11: 128135.
118. Carter BS, Leuthner SR. The ethics of withholding/withdrawing nutrition in

the newborn. Semin Perinatol 2003;27:480487.
119. Miraie ED. Withholding nutrition from seriously ill newborn infants: a parent's
perspective. J Pediatr 1988;113:262265.
120. Field MJ, Behrman RE, eds. When children die: improving palliative and end-
of-life care for children and their families. Washington, DC: The National
Academies Press, 2003.
121. Catlin A, Carter B. Creation of a neonatal end-of-life protocol. J Perinatol

2002;22:184195.
122. Franck L, Lefrak L. For crying out loud: the ethical treatment of infants' pain.
J Clin Ethics 2001;12:275281.
123. American Academy of Pediatrics and Canadian Pediatric Society. Prevention

and management of pain and stress in the newborn. Pediatrics 2000;105:454461.
124. Anand KJ, International Evidence-Based Group for Neonatal Pain. Consensus
statement for the prevention and management of pain in the newborn. Arch
Pediatr Adolesc Med 2001;155: 173180.
125. Lantos JD, Kohrman AF. Ethical aspects of pediatric home care. Pediatrics
1992;89:920924.
126. Goldberg AI, Faure EAM, O'Callaghan JJ. High-technology home care: critical
issues and ethical choices. In: Monagle JF, Thomasma DC, eds. Health care ethics:
critical issues for the 21st century. Gaithersburg, MD: Aspen Publications,
1998:146.
127. Bach JR, Campagnolo D, Hoeman S. Life satisfaction of individuals with

Duchenne muscular dystrophy using long-term mechanical ventilatory support. Am
J Phys Med Rehabil 1991; 70:129135.
128. Institute of Medicine, National Research Council. Reducing the odds:

preventing perinatal transmission of HIV in the United States. Washington, DC:
National Academy Press, 1999.
129. American College of Obstetricians and Gynecologists. Human

immunodeficiency virus screening. Joint statement of the American Academy of
Pediatrics and the American College of Obstetricians and Gynecologists. Pediatrics
1999;104:128.
130. Centers for Disease Control and Prevention. Revised recommendations for
HIV screening of pregnant women. MMWR 2001;50(RR19):59.
131. Mofenson LM, Committee on Pediatric AIDS, AAP. Technical report: perinatal
Human Immunodeficiency Virus testing and the prevention of transmission.
Pediatrics 2000;106:e88.
132. Acuff K. Perinatal drug use: state interventions and the implications for HIVinfected women. In: Faden RR, Kass NE, eds. HIV, AIDS and childbearing: public
policy, private lives. New York: Oxford University Press, 1996:214.
133. American Academy of Pediatrics, Committee on Substance Abuse. Drugexposed infants. Pediatrics 1995;96:364.
134. DeVille KA, Kopelman LM. Substance abuse in pregnancy: moral and social
issues regarding pregnant women who use and abuse drugs. Obstet Gynecol Clin
1998;25:237.
135. Jos PH, Marshall MF, Perlmutter M. The Charleston policy on cocaine use
during pregnancy: a cautionary tale. J Law Med Ethics 1995;23:120128.
P.23
136. Frank DA, Augustyn M, Knight WG, et al. Growth, development, and behavior
in early childhood following cocaine exposure: a systematic review. JAMA
2001;285:1613.
137. Chavkin W. Cocaine and pregnancytime to look at the evidence. JAMA

2001;285:1626.
138. Ferguson v Charleston 532 U.S. 67 (2001).
139. The American Society of Human Genetics Board of Directors and The
American College of Medical Genetics Board of Directors, Points to Consider:
Ethical, Legal and Psychological Implications of Genetic Testing in Children and
Adolescents, American Journal of Human Genetics 1995;57:12331241.
140. American Academy of Pediatrics, Committee on Bioethics, Ethical Issues with

Genetic Testing in Pediatrics, Pediatrics 2001;107:14511455.
141. Fletcher AB, Paris JJ. Bioethical issues surrounding transport of neonates. In:
Mhairi G. MacDonald, eds., Miller MK, assoc. ed. Emergency transport of the
perinatal patient Boston: Little Brown & Company, 1989:173.
142. DELGustafson JM. Mongolism, parental desires, and the right to life. Perspect
in Bio Med 1973;16:524.
143. DELMaine Medical Center v Houle, No. 74145, 1974 (Superior Ct,
Cumberland Co, Me. Feb 14, 1974).
144. DELWeber v Stony Brook Hosp, 476 NY.S. 2d 685, 686 (App. Div.); Bowen v
American Hospital Association, 476 US. 610 at 611 (1986).
145. DELU.S. President's Commission for the Study of Ethical Problems in Medicine
and Biomedical and Behavioral Research. Seriously ill newborns, in deciding to
forego life-sustaining treatment: a report on the ethical, medical, and legal issues
in treatment decisions. Washington: US. Government Printing Office; 1983; 197.
146. Child Abuse Protection Act, 42 U.S.C. 5103 (1982).
147. Annas G. Asking the courts to set the standard of emergency carethe case
of Baby K. N Engl J Med 1994;330:1542.
148. Paris JJ, Crone RK, Reardon FE. Physician refusal of requested treatment: the
case of Baby K. N Engl J Med 1990;322:1012.
149. State v Messenger, file 9467694-FY, Clerk of the Cir. Ct. County of Ingram,
Mich.
150. Clark FI. Making sense of State v Messenger. Pediatrics 1996; 97:579.
151. Paris JJ. Manslaughter or a legitimate parental decision? The Messenger case.
J Perinatol 1996;16:60.
152. HCA v Miller, 2000 WL 1867775, Tex. App. Hous. (Dec. 28, 2000).
Chapter 3
Neonatology in the United States: Scope and
Organization
Richard L. Bucciarelli
The practice of neonatal and perinatal medicine has changed enormously since its recognition as a distinct
subspecialty in 1975. The development of technologies as complex as extracorporeal membrane oxygenation
and as simple as the administration of exogenous surfactant, has resulted in the survival of many infants who
would have succumbed to their illnesses just a few years ago. In addition, an ever-increasing awareness of the
benefits of interpregnancy and prenatal care has lead to significant behavior modification among many
pregnant women, resulting in a reduction in the occurrences of conditions such as fetal alcohol syndrome and
neural tube defects (1). Early diagnosis and peripartum therapy has decreased the risk of the vertical
transmission of diseases such as hepatitis B and HIV (2). Despite erosions in health insurance coverage for
women and the general population, fewer pregnant women are uninsured than nonpregnant women (13%
versus 28%), and recent expansions in eligibility for coverage under Medicaid may reduce this number even
further (1). In 2000, 83.2% of women received early prenatal care, up from 75.8% in 1990 (Fig. 3-1) (3). Teen
birth rate has reached an all time low at 27 per 1,000 females ages 15 to 17 years, compared to 39 per 1,000
in 1991 (Fig. 3-2) (4).
Notwithstanding these positive trends, the rate of preterm, very preterm, low-birth-weight (LBW), and very-lowbirth-weight (VLBW) deliveries in the United States has not changed significantly in more than a decade and
remains far below the Healthy People 2010 goals (1). In 1998 the United States reported an infant mortality
rate of 7.2 per 1,000 live births, resulting in a ranking of 28th in the world behind Cuba and just ahead of
Slovakia (Table 3-1) (5). Even more troubling is the existence of significant racial and ethnic disparities in every
indicator of perinatal health outcome, with neonatal and infant mortality for African Americans exceeding twice
that of whites and three times the year 2010 goal (1).
This poor performance of the United States is not because of an unwillingness to commit significant resources
to the delivery of perinatal care. In 1985, expenditures for obstetric and neonatal care approached $15 billion.
Today this figure exceeds $40 billion, with the incremental costs for hospitalization for preterm labor reaching
$820 million (6).
The stark contrast between the enormous economic investment in perinatal care and the relatively static
outcome statistics suggests that while the United States has concentrated on developing the world's most
sophisticated high-tech care, we have done little to understand the biological and societal issues responsible for
producing high-risk infants. Perhaps this is a statement of our health care system's superb ability to care for the
individual patient, in this case a high-risk infant after delivery, while failing to develop systems that integrate
resources to provide efficient, cost-effective preventive care that could reduce human suffering and conserve
vast resources.
In this chapter, the current size and scope of the practice of neonatology in the United States is reviewed,
emphasizing how resources are currently used to deliver neonatal care. Issues are presented in an attempt to
identify potential areas of focus for those involved in developing future perinatal public policy.
Figure 3-1 Early prenatal care among mothers: United States, 1970-2000. Early prenatal care begins during
the first trimester of pregnancy. This figure is related to the Healthy People 2010 leading health indicators on
access to health care. (From Chartbook on trends in the health of Americans 2002. Centers for Disease Control
and Prevention, National Center for Health Statistics, National Vital Statistics System. Redrawn from table 6,
with permission.)
P.25
THE CAPACITY TO DELIVER IN-HOSPITAL NEONATAL CARE

Neonatal Hospital Bed Capacity
Because there is no national database defining or inventorying special-care neonatal beds, the exact capacity
for providing neonatal intensive care unit (NICU) services in the United States is unknown. The estimates of
capacity are further complicated by the fact that the definition for levels of inpatient NICU services varies
greatly from state to state. Although clearly outlined by the Committee on Perinatal Health (COPH) of the March
of Dimes Birth Defects Foundation in 1970 (revised in 1980 and 1993), financial, marketing, medical-legal, and
community demands have encouraged many hospitals to elevate the level of select perinatal services, focusing
on the provision of services locally with little regard to regional needs (7,8). As a result, there has been a
significant increase in the number of hospitals providing NICU care, as well as blending neonatal services
available primarily at basic care (level I) and specialty care (level II) centers (8).
Figure 3-2 Teen birth rate (births per 1,000 females ages 15 to 17 years), 1975-2000. Note: Teenage
childbearing has declined steadily since rising to 39 births per 1,000 teen girls in 1991. At 27 births per 1,000
in 2000, the teen birth rate has reached its lowest level ever. (From The Annie E. Casey Foundation, Kids
count data book, 2003 apocket guide. Baltimore, MD: The Annie E. Casey Foundation, 2003:7, with
permission.)
TABLE 3-1 INFANT MORTALITY RATES, BY COUNTRY, 1998

Rank
Country
Rate
Hong Kong
3.2
Sweden
3.5
Japan
3.6
4
5
Norway
Finland
4.0
4.1
Singapore
4.2
France
4.6
Germany
4.6
Denmark
4.7
10
Switzerland
4.8
11
Austria
4.9
12
13
13
15
15
17
17
Australia
Czech Republic
Netherlands
Canada
Italy
New Zealand
Scotland
5.0
5.2
5.2
5.3
5.3
5.5
5.5
19
19
21
21
21
21
25
26
27
28
Belgium
Northern Ireland
England and Wales
Greece
Israel
Spain
Portugal
Ireland
Cuba
United States
5.6
5.6
5.7
5.7
5.7
5.7
5.9
6.2
7.1
7.2
29 Slovakia
8.8
30 Kuwait
9.4
Note: Differences in data and definitions limit international compareisons. For more information, see
National Center for Health Statistics, 2002. Table 26, p. 114.
From Pastor PN, Makuc DM, Reuben C, et al. Chartbook on trends in the health of Americans 2002.
Hyattsville, MD: National Center for Health Statistics, 2002, with permission.
March of Dimes. March of Dimes data book for policy makers, Washington, DC: March of Dimes Office
of Government Affairs 2003, with permission.
In 1996, the National Perinatal Information Center reported an extensive survey of the supply and distribution
of obstetric and neonatal services. While the number of hospitals offering obstetric services had declined, the
numbers offering NICU services had increased dramatically by 64% such that one third of all hospitals with
obstetric services were also operating an NICU. When the investigators adjusted for the number of live births,
they discovered that the number of neonatal special care beds had increased significantly from 3.0 per 1,000
live births to 4.3 per 1,000 live births over the 10-year period of the study (9). More recent information
suggests that this ratio has climbed
P.26
even further to 5.1 per 1,000 live births, exceeding the calculated demand by a factor of between 2 and 3
(10,11). In 2002, Thompson and associates (11) compared the capacity for in-hospital neonatal intensive care
in the United States with that of the United Kingdom, Australia, and Canada (countries with similar neonatal
practices). It is no surprise that the United States leads the way with an NICU bed capacity equal to five times
that of the United Kingdom and twice that of Australia and Canada. Further, this difference remains even if the
capacity is adjusted for the incidence of LBW infants (Table 3-2). The authors also point out that despite this
huge capital investment, the United States trails all three countries in reported neonatal and infant mortality
rates.
TABLE 3-2 NEONATAL CARE: PHYSICIAN AND HOSPITAL CAPACITY
United States
Australia
Canada
United Kingdom
3942
256
349
636
144.7
34.2
59.7
20.0
6.1
8.0
3.7
5.7
3.3
5.5
2.7
3.7
100.2
42.2
45.3
24.3
3.2
0.4
NA
NA
169.7
813.9
816.7
597.7
1.21
0.90
0.72
2.92
Annual deliveries (in thousands)

Physicians* (per 10,000 live births):
Pediatricians
Neonatologists
Neonatologists per 1,000 low-birthweight infants
Obstetricians/gynecologists
Maternal/fetal specialists
Family practitioners
Number level III units* (per 10,000
live births)
Neonatal hospital beds* (per 1,000

3.3
NA
NA
0.67
live births) Intensive care beds
Intensive beds per 1,000 low birth
43.4
NA
NA
9.3
weight
Intensive + intermediate care beds
5.1
2.6
2.6
NA
Intensive + intermediate care beds
67.1
39.4
44.8
NA
per 1,000 low-birth-weight infants
From Thompson LA, Goodman DC, Little GA. Is more neonatal intensive care always better? Insights
from across-national comparison of reproductive care. Pediatrics 2002; 109:10361043, with
permission.
Although the United States has an in-hospital neonatal bed capacity that exceeds that of any other nation, its
distribution across our nation is not homogenous and does not seem to follow any identifiable pattern, ranging
from a low of 14 per 10,000 live births to a high of 59 per 10,000 live births (from 71 to 1,319 total births per
NICU bed). Once again there was no correlation between the number of beds in a region with the incidence of
LBW or neonatal mortality (12).
Size and Effectiveness of the Neonatologist Workforce

In an attempt to address the capacity issue more accurately, investigation recently has focused on the size and
distribution of the neonatal work force as a potentially more accurate proxy for our capacity to deliver inhospital neonatal care in the United States. Clearly, neonatologists are the catalysts that create a hospital
environment, resulting in the creation of NICU beds and ultimately in the desire to provide higher risk
obstetrical and neonatal care within that institution. Because a neonatologist can utilize a broad array of
neonatal health care specialists to provide services at several institutions simultaneously, the ratio of
neonatologists to both high- and low-risk deliveries and their distribution within the region may more accurately
define the capacity for in-hospital neonatal care.
In 1975, the American Board of Pediatrics' sub-Board of Neonatal and Perinatal Medicine certified 355
physicians as the first neonatologists in the United States. Since then, the American Board of Pediatrics has
certified an additional 3,531 candidates. Currently there are 3,886 neonatologists certified by the Board,
including the results of the latest certifying examination in 2001 (Fig. 3-3) (13). It is of interest to note that
between 1987 and 1999, an average of 300 candidates was certified with each examination. In 2001, the
number dropped to 182, the lowest number ever certified during a single examination. The average age at
certification is 32 years, while the average age of all certified neonatologists is 51 years. During the early years
of certification, males dominated by almost 3:1. In more recent years, the gender distribution approached 1:1,
with
P.27
more females certified during the 2001 certification (97 females, 85 males). Although the attrition rate for
neonatologists was once believed to be quite high, it has remained low at 1% to 2% over a 10-year period.
Additionally, it appears that almost 1,000 physicians actively engaged in the practice of neonatology are
physicians who have not completed sub-Board certification or are pediatricians with special training in
neonatology who concentrate their efforts on caring for neonates with special needs (14,15). Thus, the total
neonatal physician workforce approaches 5,500, having doubled in the last decade and more than tripled since
1985 (13,14 and 15).
Figure 3-3 Cumulative total of specialists in neonatal and perinatal medicine certified by the American Board
of Pediatrics, 1991-2001. (From American Board of Pediatrics. American Board of Pediatrics Neonatal-Perinatal
Medicine qualifying examination statistics. Chapel Hill, NC: American Board of Pediatrics. Available online at
http://www.abp.org/stats/WRKFRC/neostat.htm.)
TABLE 3-3 BIRTHS PER EACH CLINICALLY ACTIVE NEONATOLOGIST AND PEDITRICIAN IN UNITED
STATES, 19811996
Number of
Birth (millions)
*
1981
1985
1989
1992
1996
* Birth from
Number of
Neonatologists
3.63
3.76
4.04
4.07
3.90
U.S. Vital Statistics;
All Births
Births per Neonatologist

Very-LowLow-BirthBirthExtremely LowWeight
Weight
Birth-Weight
(<2500 g)
(<1500 g)
(1000 g)
504
7201
490
952
3951
266
1507
2681
189
1866
2178
154
2311
1687
123
1995 births substituted for 1996.
83
48
34
28
23
39
24
17
14
11
Physician numbers used for calculation exclude residents and fellows (N = 304), and those
predominantly engaged in teaching (N = 106), administration (N = 97), and research (N = 232).
From AMA Physician Masterfile (December 31 of calendar year). For comparison, the United States
Directory of Neonatologists 1996 listed 2,635 board-certified, 652 board admissible, and 230 others
practicing in neonatology; the professional activities (clinical care, teaching, etc) of the neonatologists
are not available.
From Goodman DC, Little GA. General pediatrics, neonatology and the law of diminishing returns.
Pediatrics 1998;102:396399, with permission.
Concerns about a physician workforce devoted to neonatology, its distribution, its affect on regionalization, and,
most importantly, its affect on outcome measures have been raised several times (16). In 1981, there were 83
very-low-birth-weight (less than 1,500 g) and 39 additional extremely low-birth-weight (ELBW [less than 1,000
g]) infants for every neonatologist. By 1996 these numbers dropped to 23 VLBW infants and 11 ELBW infants
for every active neonatologist (Table 3-3) (17). In its statement on the neonatal workforce, the American
Academy of Pediatrics Committee on the Fetus and Newborn (COFN) recommended that the neonatologist:live
birth ratio be 1:2,569 live birthswhich is 3.9 neonatologists per 10,000 live births (16). In 1996, the ratio was
calculated to be 1:1,687 live birthswhich is 5.9 neonatologists per 10,000 live births, 1.5 times the original
target. In addition, 50% of the neonatologists surveyed in this study reported that a very significant portion of
their practice was devoted to the care of normal newborn infants (18).
Similar to the in-hospital bed capacity analysis, Goodman and associates (18) report wide variations in the
availability of neonatologists, ranging from 1.2 per 10,000 live births to a high of 25.6 per 10,000 live births (a
range of 390-8,197 births per neonatologist) (Figs. 3-4a,b). Regional variations were unexplained by
differences in LBW, number of high-risk pregnancies, or use of mid-level providers. Regions with an average of
2.7 neonatologists per 10,000 live births or less experienced a slight, but significant increase in mortality only
for infants with birth weights of 500 to 999 g. Once birth weight exceeds 999 g or the number of neonatologists
exceeds 4.3 per 10,000 live births, there are no additional improvements seen in neonatal mortality (19).
Lessons Learned
The attempt to correlate outcome with either NICU in-hospital bed capacity or the number and distribution of
neonatologists presents several dilemmas. First, gross outcome of neonatal care, such as the incidence of LBW
or neonatal and infant mortality, may be insensitive to regional changes in resource availability. It is entirely
possible that there exists a yet-to-be-defined threshold above which the impact of additional available
resources is no longer obvious. Alternatively, it is well established that birth weight is a reflection of
socioeconomic status and quality of care before delivery and is not directly influenced by postnatal medical
interventions (12,20). As the COPH stated in 1993, appropriate preconception and prenatal care may contribute
significantly to increasing birth weight and decreasing neonatal mortality (7).
COSTS OF DELIVERING NEONATAL CARE

Acute Care In-Hospital Costs
Estimates suggest that the total expenditures for perinatal care in the United States exceed $6,850 per motherinfant pair (7). This represents 4.5% of personal health care
P.28
spending by all Americans and 7% of health care spending by the population younger than 65 years of age.
Most of these resources are spent on acute, high-tech care of the critically ill neonate, with daily charges easily
reaching $1,000 to $2,500 for hospital services and an additional $300 to $450 each day for physician services
(21). Of the four million births predicted in 2003, 280,000 will be premature, with an estimated cost of $12
billion for acute in-hospital neonatal care. Approximately one third of these expenditures, $4 billion, would be
used to provide care for infants weighing between 500 and 700 g (22). An additional $1 billion would be spent
on hospitalizations for preterm labor in an attempt to delay delivery as long as possible. Because the costs for
complicated births (ranging from $20,000 to $400,000) are many fold greater than the costs of an
uncomplicated delivery (averaging $6,400), the investment in delaying delivery as long as possible is entirely
justified (6).
Figure 3-4 Total birth figures. A: Total births per intensive care bed. B: Total births per neonatologist.
(From Goodman DC Fisher ES, Little GA. Are neonatal intensive care resources located according to
need? Regional variations in neonatologists, beds, and low birth weight newborns. Pediatrics:
2001;108:426, with permission.)
Charges associated with specific perinatal diagnoses rank among the highest of all
hospital diagnoses (Table 3-4). A 1999 report by the Health Care Cost and
Utilization Project's Nationwide Inpatient Sample listed neonatal respiratory
distress syndrome as the most expensive diagnosis, with a mean charge of
$82,648 and a mean length of stay of 27.8 days. Other diagnoses included in the
top 10 categories were:
Low-birth-weight (mean charge of $56,942)
Cardiac anomalies (mean charge of $49,858)
Intrauterine hypoxia and birth asphyxia (mean charge of $36,954) (23)
In analyzing data from 3,288 VLBW infants from 25 NICUs, Rogowski (24)
reported that treatment costs vary inversely with birth weight, with a median
treatment cost of $89,546 for infants weighing between 501 g and 750 g and
$32,531 for infants weighing between 1,251 g and 1500 g.
P.29
In this analysis, she included costs of accommodation (room and personnel costs)
and ancillary services (respiratory therapy, laboratory, radiology, pharmacy, and
medical supplies). While accommodation costs varied by a factor of two when
comparing the largest and smallest VLBW infants, ancillary costs varied by almost
five fold, with laboratory and respiratory services showing the widest variation.
When costs were corrected for length of stay, the costs per day were surprisingly
similar, varying by only $552 per day between infants weighing 1,251 to 1,500 g
at birth and those weighing 501 to 750 g at birth (Table 3-5).
Because physician services are billed separately from the hospital, most
studies do not include them in the analysis and, as a result, the true
picture of the cost of neonatal care is somewhat distorted. This distortion
can be significant when one considers that the cost of physician services
in the NICU approaches 15% of the total hospital bill (21).
Although these data are valuable and serve as a rough estimate of
expenditures, caution must be used when attempting to apply them to
actual expenditures for neonatal care at the local or state level. By and
large, these data are derived from in-hospital data as reported on state
and federal cost reports, which reflect charges and not the true costs of
delivering care. Further, many studies report individual hospital, state, or
regional experiences that vary greatly based on the accounting principles
used by that facility.
TABLE 3-4 MOST COMMON HOSPITAL DISCHARGE DIAGNOSES

WITH MEAN CHARGES AND MEAN LENGTH OF STAY
Mean Length
Rank
Principal Diagnosis
Mean Charges*
of Stay (days)
Respiratory distress
82,648
27.8
syndrome
2
Spinal cord injury
58,690
12.4
3
Short gestation, low56,942
22.7
birth-weight, and fetal
growth retardation
4
Leukemias
53,252
13.8
5
Heart valve disorders
51,292
8.8
6
Cardiac and circulatory
49,858
7.8
congenital anomalies
7
Other CNS infection
44,060
12.6
and poliomyelitis
aneurysms
8
Aortic, peripheral, and
42,461
8.7
visceral artery
9
Intrauterine hypoxia
36,954
12.1
and birth asphyxia
10 Cancer of stomach
35,426
10.8
* Charges are for acute hospital care and do not include physician
and other professional fees, rehabilitation expenses, or costs
associated with follow-up or home care.
CNS, central nervous system.
From March of Dimes Perinatal Data Centers. Economic costs of
prenatal careHealth Care Utilization Project nationwide inpatient
sample. 1999, White Plains, NY: March of Dimes, 2002, with
permission.
TABLE 3-5a MEDIAN TREATMENT COSTS AND LENGTH OF STAY FOR VERY-LOW-BIRTH-WEIGHT INFANTS
N
Total Cost
Accommodation Cost
Ancillary Cost
Length (days) of Stay
Cost per Day
3288
$49,457
$35,521
$13,872
49
$1115
501750 g
7511000 g
10011250 g
601
811
861
$89,546
$78,455
$49,097
$59,318
$54,259
$35,460
$28,094
$23,288
$13,376
79
72
49
$1483
$1200
$1059
12511500 g
1015
$31,531
$24,609
$ 6,224
35
$ 932
All infants
Birth weight
TABLE 3-5b MEDIAN ANCILLARY COSTS FOR VERY-LOW-BIRTH-WEIGHT INFANTS

Cost Per Day
All infants
Birth weight
501750 g
7511000 g
Total Ancillary
Cost
Respiratory Therapy
Laboratory
Radiology
Pharmacy
Other
Ancillary
3288
$13,872
$3112
$3308
$ 942
$2258
2474
$323
601
811
$28,094
$23,288
$8678
$7421
$6550
$5494
$1671
$1475
$3717
$3668
4054
3726
$546
$382
10011250 g
861
$13,376
$2884
$3205
$ 900
$2195
2554
$301
12511500 g
1015
$ 6,224
$1044
$1720
$ 473
$1101
1382
$194
From Rogowski J. Measuring the cost of neonatal and perinatal care. Pediatrics 1999;103:329335, with permission.
In many instances, hospital estimates of disease and patient-specific expenditures

are based on small samples, resulting in wide variations within and between
studies. Expenditure comparisons are complicated further by regional variations in
costs, charges, standards of care, differences in patient population demographics,
and intensity of illness. Finally, much of the available information is derived from
urban level III NICUs or state-funded perinatal programs, which include a high
percentage of publicly funded patients. Because these patients often represent the
highest perinatal risks, disease-specific expenditures may be skewed by
overrepresentation of the most complex patients in the database, resulting in
higher cost estimates when compared to private-sector spending.
In an attempt to control expenditures for neonatal services, both public and
private payers have developed
P.30
prospective payment systems, which place providers of neonatal care at risk for
managing costs generated by their patients. Payment systems such as the Florida
Neonatal Care Groups and the APR-DRGs (all patient refined-diagnostic care
groups) have proved to be acceptable ways to provide appropriate reimbursement
for services while controlling costs for public and private payers (21,25). With the
recent Medicaid expansions and many states enrolling more and more patients in
Medicaid managed care plans, fixed rather than cost-based reimbursement is
becoming more common, leading many institutions to explore new ways to control
NICU costs.
Without a doubt, NICU care is one of the most expensive services available because of the high technology and
the prolonged length of stay. If we are going to be able to continue to provide every patient the full range of
services, neonatal health care professionals will need to ensure that the available resources are used in the most
cost-effective manner to provide the highest quality of care. To achieve this goal, neonatal providers should
become aware of the costs and hospital cost structures associated with the care they are providing. Richardson
and associates (26) published a detailed analysis of the types of costs generated by an inpatient (Fig. 3-5).
Understanding how fixed, variable, direct, and indirect costs are derived will help the neonatal team identify
items that can be considered part of cost reduction and those that are unaffected by census or acuity. Although
we must strive to be sure that every neonate has access to a full range of therapies, resources used
inappropriately on one or a group of patients may deny full access to care to others (27).
Acute Medical Care Post Discharge

Graduation from the NICU does not herald the end of increased capital investment in the surviving infant. In a
series from the United Kingdom, Stevenson and associates (28) determined that the mean cost per LBW infant,
born at less than 1,500 g and surviving without major disability, was 13 times greater than that of an infant
weighing between 1,501 and 2,000 g; and the cost per infant weighing less than 1,000 g at birth was 55 times
control. Further, LBW infants continued to utilize hospital and practioner services at a greater rate than control
until age 8 to 9. Similarly, Rogowski (29) reported in 1998 that, on average, the VLBW who is an NICU graduate
will require almost $1,400 in outpatient services and be hospitalized twice with an average length of stay of 11
days at a cost of almost $12,000. In many instance these outpatient visits and pharmaceuticals are either not
covered by traditional payers or they have significant copayments and deductibles, resulting in significant
financial hardships for families. Thus, additional acute medical costs after discharge from the NICU are of major
concern to private and public payers as well as parents.
Long-Term Costs
Because LBW infants are at high risk for long-term sequelae, the resources dedicated to providing acute care
must be supplemented with the continued investment of resources for long-term followup. With the realization
that the long-term outcome of LBW infants is enhanced by developmental followup, including multidisciplinary
intervention programs, postdischarge costs per NICU graduate have become substantial (30). In a study
reported from Helsinki, Finland, the costs for 71 ELBW infants during
P.31
the first 2 years after discharge from an NICU averaged $34,529 compared to $1,034 for term, control infants
(31). Although this study attempted to identify total postdischarge costs, family out-of-pocket expenses were
not estimated. A recent survey indicates that these expenses for children with special health care needs can
exceed 15% of annual family income (32).
Figure 3-5 Total hospital costs for newborns greater than 1500 g birth weight, by birth-weight group. (From
Richardson DK, Zupancic JA, Escobar GJ, et al. A critical review of cost reduction in the neonatal intensive care
unit. I. The structure of costs. J Perinatol 2001;21:107, with permission.)
The increased survival of ELBW infants presents several challenges to our educational system. Stevenson and
colleagues (28) report that 52% of the long-term expenditures related to LBW are related to special education
needs. Lewit and colleagues (33) report that health care, education, and child care for the more than 4 million
children ages 0 to 15 born at LBW and VLBW cost between $5.5 and $6.0 billion more than they would have if
these children had been born at a normal birth weight.
Beyond the Costs of Acute Neonatal Care

When long-term cost-benefit ratios were calculated in each of the past 3 decades, the provision of neonatal care
appeared extremely cost-effective. This is still true, because high-quality neonatal care decreases the incidence
of significant morbidity in almost all weight groups. However, the increased number of surviving ELBW infants
with their financial, social, and educational impacts has led several researchers to raise significant ethical and
moral questions about future therapeutic directions as we approach the limits of viability for some of our
smallest patients. While the cost-benefit ratio for ELBW infants is of obvious concern, the more fundamental
ethical and moral issues deal with whether or not high-tech care should be provided to infants whose very
survival is extremely unlikely, and the long-term consequences for those who survive are significant. How should
we approach these infants? Who should decide what therapies should and should not be given? What is the role
of the neonatal team, the family, the payers in the decision-making process? There appear to be few, if any,
consensus answers to these critical questions because the ethical, legal, and moral considerations involving the
extremely premature infant have not kept pace with the advancement of technology. Our nation has only
recently been willing to admit that even the richest country in the world cannot afford to buy all that science has
to offer. Prioritizing therapies and making very difficult decisions about the use of resources must be discussed
more in years to come (34).
Missed Opportunities
One cannot consider the magnitude of these costs and the amount of human suffering involved in the delivery of
a high-risk infant without wondering why the United States continues to expend enormous resources on
postnatal care when half of the ELBW deliveries could be avoided with early prenatal care (1). The benefit of
$1,000 of prenatal care instead of $150,000 for each surviving LBW infant is obvious. Almost 2 decades ago, the
Institute of Medicine reported that for every $1 spent on prenatal care, $3 were saved in the first year of life,
and $10 more were saved over a lifetime (35). This relationship between prenatal care and downstream costs
appears to be true today as well. A recent report by Lu and associates (36) shows that for every $1 reduction in
prenatal care provided by public programs there was an increase of $3.33 in postnatal care costs and a $4.63
increase in long-term morbidity costs. The average cost of long-term care (medical, child care, and special
education) for women without prenatal care is $4,839 compared to $1,592 for women with prenatal care.
Fortunately, today 83% of women receive prenatal care and begin that care within the first trimester (3).
Financial access to care combined with comprehensive prepregnancy and prenatal programs, including good
nutrition and avoidance of high-risk lifestyles, have been identified as key in reducing LBW with its attendant
costs and human loss (7). Of major concern, however, is the remaining economic and noneconomic barriers
confronting minority populations, resulting in late or no prenatal care rates, which are three to four times that of
whites (1).
Major Payers of Perinatal Care

There are four major payers for perinatal care in the United States:
Private health insurance, purchased through employers or directly by individuals
Public programs including Medicaid and other federal, state, and local programs
Special perinatal demonstration projects, both public and private
Self-pay or uninsured
None of the major payers completely cover the costs of high-risk obstetric and neonatal care, especially for the
complicated ELBW infant. A significant portion of the uncompensated care created by underpayment by public
payers and the uninsured is often shifted to other subscribers in the form of increased charges. One study
suggests that as much as 27% of total costs of NICU care were shifted to paying patients who are responsible
for generating 60% of total revenues while accounting for only 33% of costs (37). In recent years, payments
under fee-for-service plans have been replaced by capitated managed care and discounted plans, reducing the
ability of NICUs to shift cost and placing many in financial jeopardy.
Private Health Insurance

Only 72% of women of childbearing age are covered by employer or privately purchased insurance (Fig. 3-6A).
Women are less likely than men to have employer-based coverage because they are more likely to be employed
less than full time, to work for small employers who do not offer insurance, or have spouses who work for
employers who do not offer dependent coverage (1). Even for those with health insurance, 7% of plans offered
in 2000 by
P.32
small employers (10 to 24 workers) did not cover prenatal care or delivery expenses. Plans providing maternity
benefits often impose substantial cost-sharing arrangements, which can become a significant barrier to care,
resulting in limited prenatal care and increasing the chance of a high-risk delivery. Similar limitations are often
seen in coverage of high-risk neonates after delivery (38). Cost-sharing arrangements with lifetime caps on
policy benefits can be financially devastating to families. For example, a 20% copayment responsibility on a
combined hospital and physician bill of $250,000 for an LBW infant results in a family out-of-pocket expense of
$50,000, an amount that would be impossible for most families to pay. Many policies carry a $250,000 to
$500,000 lifetime cap. In this case the infant would use all or half of the entire lifetime benefit at discharge from
the NICU.
Figure 3-6 Insurance status of (A) women ages 15 to 44 years and (B) pregnant women in the United States,
2002. (From U.S. Department of Health and Human Services. Retrieved July 1, 2003, from http://www.hhs.gov/
budget/docbudget.htm.)
Medicaid
Medicaid is the largest public program for the financing of prenatal and neonatal care. Authorized as Title XIX of
the Social Security Act in 1965, Medicaid is a federal-state partnership that finances care for 51 million lowincome pregnant women, children, the elderly, the blind, and the disabled. In 2002, it surpassed Medicare to
become the nation's largest health insurance program in numbers of people served and in expenditures, totaling
$259 billion and outspending Medicare by $2 billion (39). During the past several years, the federal-state match
has remained relatively constant, with federal funds accounting for an average of 56.9% of total fiscal-year
program expenditures. The exact level of federal funding is determined from a formula reflecting the state's per
capita income and varies between 50% and 80% of the state's total Medicaid expenditures.
Today, 51% of the targeted population consists of pregnant women and children; however, only 15% of total
program expenditures are directed toward them (40). With the exception of the matching requirements for
federal funds and the required compliance with specific federal mandates, each state has the responsibility for
developing and administering its own Medicaid program. This includes setting eligibility and coverage standards
within broad federal guidelines. As a result, there is considerable variation among states in eligibility, range of
services offered, limitations on services, and reimbursement policies.
Over the past several years, Congress has been successful in extending Medicaid eligibility to more pregnant
women and children. Beginning in 1989, all states were required to cover pregnant women and children younger
than 6 years of age with family incomes below 133% of the federal poverty level (FPL) and had the option of
extending benefits to pregnant women and infants younger than 1 year of age with family incomes below 185%
of the FPL. The State Children's Health Insurance Program, under Title XXI of the Social Security Act, enacted in
1997, gave states new latitude in identifying eligible patients and allows states to expand coverage for children
and pregnant women up to 200% of the FPL, with some states using additional funds to reach 300% of the FPL
(37,40).
In 2001, Medicaid was the source of payment for 1.6 million births, 37% of the nation's deliveries, ranging from
a low of 9% in Hawaii (the only state with universal health coverage) to a high of 56% of all births in New
Mexico (1,40,41). As a result of this expanded eligibility, many women drop more expensive employer-based
and private coverage and enroll in Medicaid. Thus, employer-based and private coverage among pregnant
women has dropped to 50% with 13.0% remaining uninsured (Fig. 3-6B) (37).
Title V Maternal and Child Health Block Grant Program

Located within the Health Resources and Services Administration (HRSA) of the Department of Health and
P.33
Human Services (HHS), the Maternal and Child Health (MCH) Block Grant Program was authorized by Title V of
the Social Security Act to provide grants to states for a variety of preventive and primary care services to
women and children, including prenatal care, immunizations, and rehabilitative services for children with special
needs. Although fiscal year 2002 funding for this program exceeded $930 million, it accounts for less than 1% of
the HHS budget (42). Current funding levels allow less than half of eligible women access to prenatal care under
the program.
Community Health Centers, Migrant Health Centers, and the

National Health Service Corps
As part of HRSA, $1.4 billion for community health centers, migrant health centers, and programs of the
National Health Service Corps are directed toward providing primary preventive care, including prenatal and wellchild care, for those served in urban community health centers and in rural areas of the United States (42).
Special Demonstration Projects

Because prevention of low-birth-weight infants can positively affect perinatal outcome, many public (e.g., state,
county, and municipal governments) and private (e.g., March of Dimes, Robert Wood Johnson, Pew Foundation)
organizations have dedicated significant resources to study how to reach high-risk populations more effectively
to reduce the mortality and morbidity related to premature deliveries. Although most of these projects are small
and contribute little to covering the total costs of perinatal care, many are highly successful, improve outcomes
for individual patients, and add significantly to our understanding of the perinatal health care system and the
delivery of health care to special populations. A major problem with these projects is that targeted patients are
highly selected, making it difficult to apply program results to larger populations. Many programs are funded for
a limited time, creating the problem of how to continue to provide services after funding has expired.
Uninsured, Self-Paying Patients

Many of the uninsured are in young families working for small business or are in the service industry and are not
provided employer-based health insurance. Sixty percent of young families of childbearing age are two-parent
families, and 85% of them have at least one member employed full- or part-time. Most uninsured families have
limited resources, with incomes between $30,000 and $32,000 annually. Sixty-one percent of uninsured women
have family incomes below 200% of the FPL, making it very difficult to pay out of pocket for medical expenses
or even for prenatal care (37).
More than a decade ago, the National Commission on Children found that being uninsured during pregnancy is
the greatest barrier to receiving prenatal care, and it results in the highest risk for delivering an LBW infant,
approximately three to five times the rate for women with even minimal prenatal care (43). Fortunately, this
barrier to access to prenatal care seems to have been significantly reduced. As a result of enhanced eligibility
under Medicaid, only 13% of pregnant women (422,000) are uninsured during pregnancy and up to 1 year
following delivery (Fig. 3-6B) (1).
Although enhanced coverage for pregnant women ensures access to prenatal care, the loss of eligibility 1 year
after delivery presents major problems for continued care between pregnancies and makes preparation for
subsequent healthy pregnancies problematic. As many as 20 million women have their insurance interrupted for
at least 2 months over a 24-month period, and data indicate that 26% of women are uninsured when they
conceive (44).
MAJOR MEASURES OF OUTCOME

Infant Mortality Rate
The infant mortality rate (IMR) is defined by the National Center for Health Statistics and by the World Health
Organization as the number of deaths occurring within the first year of life per 1,000 live births. Infant mortality
can be divided further into neonatal mortality (i.e., death before 29 days of age) and postneonatal mortality (i.
e., death between 29 days and 1 year of age). Neonatal mortality generally is the result of factors related to
pregnancy and birth, while postneonatal mortality generally is the result of environmental factors (e.g., trauma,
infection, nutrition, sudden infant death syndrome) (41).
Infant mortality in the United States has declined substantially since 1975. Initially, the largest declines were
seen in the neonatal mortality rate (NMR), but more recently these advances have been matched by similar
declines in the postneonatal mortality rate (PNMR). Between 1980 and 2001, the IMR dropped from 12.5 deaths
per 1,000 live births to 6.9 deaths per 1,000 live births as a result of a 45% decline in the NMR (Fig. 3-7) (41).
Despite these achievements the current IMR exceeds the Healthy People 2010 goal of 4.5 by 35% (1,3).
Of major concern is the persistent disparity in the IMR between white and African American infants. Although all
race groups have experienced a decrease in IMR, the rate for African American infants fell by only 36.5%, which
is 12.1% less than that of whites. Recently published 2000 rates show an IMR for white infants at 5.7 per 1,000
live births, whereas that of African American infants is reported to be 14.1 deaths per 1,000 live births, a ratio of
2.5:1 (Fig. 3-8). The infant mortality rate for Hispanics is slightly better than that for whites at 5.6 per 1,000
live births (41,44).
Neonatal Mortality Rate

Although the IMR reflects the general health of a community, separate examination of early and late infant
deaths
P.34
can further pinpoint problems. The NMR equals the number of deaths occurring at less than 29 days after birth
per 1,000 live births (39). These account for approximately 67% of all infant deaths. Half of all neonatal deaths
can be attributed to four leading causes:
Figure 3-7 Infant, neonatal, and postneonatal mortality for low-birth-weight (LBW), very-low-birth-weight
(VLBW), and preterm (PT) delivery in the United States from 1980-2001. Infant mortality rate (IMR) indicates
infant deaths per 1,000 live births; neonatal mortality rate (NMR) indicates neonatal deaths per 1,000 live
births; postneonatal mortality rate (PNMR) indicates postneonatal deaths per 1,000 live births; LBW indicates
percent low-birth-weight (<2500 g); VLBW indicates percent very-low-birth-weight (<1500 g); PT indicates
percent preterm (<37 weeks of gestation). (From MacDorman MF, Minino AM, Strobino DM, et al. Annual
summary of vital statistics 2001. Pediatrics 2002;110:1037, with permission.)
Figure 3-8 Infant mortality in the United States from 1990 to 2000 by race. Infant mortality constitutes deaths
that occur before 1 year of age. (From U.S. Department of Health and Human Services. Retrieved July 1, 2003,
from http://www.hhs.gov/budget/dccbudget.htm)
Figure 3-9 Neonatal mortality in the United States from 1990 to 2000 by race. Neonatal mortality constitutes
deaths that occur before 28 days of age. (From U.S. Department of Health and Human Services. Retrieved July
1, 2003, from http://www.hhs.gov/budget/dccbudget.htm)
LBW
Acute perinatal asphyxia
Congenital anomalies
Perinatal infections
Between 1980 and 2001, the NMR declined by 45%. The 2000 rate of 4.6 deaths per 1,000 live births is thought
to be a reflection of the marked improvement in the treatment of respiratory illnesses in term and preterm
infants. Once again the rate for African Americans (9.4 deaths per 1,000 live births) parallels the IMR and is
considerably higher than the white NMR, with an NMR ratio of African Americans to whites of 2.5:1 (Fig. 3-9).
These differences appear to be primarily because of the higher incidence of LBW (less than 2,500 g) and VLBW
(less than 1,500 g) infants born to African American women and a higher NMR for African American infants
weighing more than 2,500 g (1.2 deaths per 1,000 live births for African Americans compared to 0.9 deaths per
1,000 live births for whites). NMR for Hispanics is slightly better than for whites at 3.7 deaths per 1,000 live
births (41).
Perinatal Mortality Rate

The perinatal mortality rate (PMR) is defined by the National Center for Health Statistics as the number of late
fetal deaths (i.e., fetal deaths of 28 weeks or more gestation) plus early neonatal deaths (i.e., deaths of infants
0 to 6 days of age) per 1,000 live births (41).
Most fetal deaths are a result of chronic asphyxia (60% to 70%); congenital malformations (20% to 25%);
superimposed complications of pregnancy, such as placental abruption, diabetes mellitus, intrauterine infection
(5% to 10%); and unexplained deaths (5% to 10%) (43). Although the incidence of congenital malformations
has remained relatively constant over the last few years, improvements in the management of pregnancyrelated complications, combined with improved early neonatal survival, has resulted in
P.35
a significant decrease in PMR to 7.0 deaths per 1,000 live births for 2000. Race variations in PMR are slightly
better than those seen in IMR and NMR (Fig. 3-10). PMR in African American women is less than twice that of
whites (12.7 deaths per 1,000 live births compared to 6.7 deaths per 1,000 live births). The reported PMR for
Hispanics for 2000 is 6.0 deaths per 1,000 live births and is considerably less than the 6.7 deaths per 1,000 live
births reported for whites (41).
Figure 3-10 Perinatal mortality in the United States from 1990 to 2000 by race. Perinatal mortality constitutes
fetal death before 28 weeks of gestation plus infant deaths at 0 to 6 days of age. (From U.S. Department of
Health and Human Services. Retrieved July 1, 2003, from http://www.hhs.gov/budget/dccbudget.htm)
MAJOR MORBIDITIES
Preterm Birth
The preterm birth rate is defined as births from pregnancies that have completed less than 37 weeks of
gestation. In 2000 there were 467,201 (11.6% of live births) preterm births recorded, only 40% of which were
also low-birth-weight (Fig. 3-11). This is the first time there has been a decline in the rate of preterm births in
the United States since 1992, declining from 11.8% of live births in 1999 to 11.6% of live births in 2000 (1).
Although recent trends are encouraging, the 2000 preterm rate still exceeds that of the Healthy People 2010
goal of 7.6% of live births by 50%. The decrease in preterm births form 10.5% to 10.4% of live births among
non-Hispanic whites is the first decrease in a decade, during which rates steadily rose from 8.5% of live births.
The preterm birth rate in African Americans remains high at 17.3% of live births, but has trended downward
since peaking at 18.9% of live births in 1991. The very preterm rate (less than 32 completed weeks of
gestation) remains unchanged from the 1981 rate of 1.93% of live births, and is more than twice the Healthy
People 2010 goal of 0.9 % of live births (3).
Figure 3-11 Incidence of preterm and low-birth-weight births in 2000. (From March of Dimes. March of Dimes
data book for policy makers, Washington, DC: March of Dimes Office of Government Affairs, 2003, with
permission.)
Figure 3-12 Incidence of low-birth-weight and very-low-birth-weight births as a percent of all live births in the
United States from 1990 to 2000. (From March of Dimes. March of Dimes data book for policy makers,
Washington, DC: March of Dimes Office of Government Affairs, 2003, with permission.)
Low-Birth-Weight
LBW is responsible for 66% of the U.S. IMR and carries a six-fold increased risk of death for infants weighing
between 1,500 and 2,499 g, a 98-fold increased risk of death for infants weighing less than 1,500 g, and a twoto three-fold increase in the chance of long-term disability (41). Considering these realities, the prevention of
LBW has been one of the nation's top priorities in its effort to reduce infant mortality and morbidity. Despite this
focus, no progress has been made in reducing the incidence of LBW over the last decade. In 2000, more than
300,000 infants (7.6% of live births) were born at LBW, which is the highest recorded in the previous 20 years
and exceeds the Healthy People 2010 goal by 52% (Fig. 3-12) (1). Currently, the LBW rate among African
Americans, at 13.0 % of live births, is 1.7 times that of whites. Examination of annual LBW statistics reveals a
slow but continued increase in the overall number of infants born at LBW since 1990 (Fig. 3-13) (41,45).
However, the greatest rise in LBW occurred between 1990 and 1995, coincident to a dramatic increase in the
number of births to teens (Fig. 3-2). This association is not surprising when one realizes that teen pregnancy
carries a very significant increase in the risk of having an LBW infant.
Figure 3-13 Trends in low-birth-weight births (less than 2500 g) by race as a percentage of all births, United
States 1990-2000. (From U.S. Department of Health and Human Services. Retrieved July 1, 2003, from http://
www.hhs.gov/budget/dccbudget.htm)
P.36
Multiple Births
The twin birth rate has risen by 55% since 1980. This is of major health significance because half of all twin
gestations and the great majority of triplet births are high risk. In fact, twins are 5 times as likely to be
premature and 9.5 times as likely to be of VLBW than singleton deliveries. As such, they are responsible for
some of the continued trends in U.S. perinatal statistics and consume significant resources (41). However, this
increase in use of perinatal resources is directly related to prematurity and low-birth-weight rather than the twin
gestation per se (46). It appears that this increase in multiple births is related to two trends in American society.
First, more and more women are delivering later in life. In 2000, birth rates for women in their thirties reached
their highest levels in 30 years. Birth rates for women 40 to 44 years of age also rose in 2000 and have more
than doubled since 1981 (41). These increases are related to the growing tendency for many women to
postpone childbearing. Secondly, the use of assisted reproductive technology, ovulation-inducing drugs, and in
vitro fertilization, have long been associated with a higher risk of multiple births (47,48). The real potential for
multiple births and their attendant increased risk for LBW (with its increased mortality and morbidity) should be
considered when either delay in childbearing and/or assisted reproduction are being considered.
Lifestyle Choices
Some of the perinatal risk factors associated with LBW and poor perinatal outcome are not within a pregnant
woman's immediate control. However, many of the factors associated with lifestyle choices are (49).
Cigarette Smoking
Cigarette smoking during pregnancy is the single largest modifiable risk factor for LBW (up to 20%) and infant
mortality. Babies born to mothers who smoke are, on average, 200 g lighter than those born to mothers who do
not smoke. In a recent review, Eyler and Behnke (48) reported that infants exposed to tobacco in utero
demonstrated abnormal behavior up to 14 days after delivery. Abnormalities included poor auditory habituation,
and disorders of orientation and autonomic regulation. Although implications of these findings are unclear, they
do suggest that smoking during pregnancy carries significant long-term behavioral risks in addition to those
associated with a low-birth-weight.
Despite these and other well-publicized risks of smoking during pregnancy, a March of Dimes survey reports that
20% of pregnant women 15 to 44 years of age stated that they had been smoking cigarettes regularly, and
smoking during pregnancy was common with women under 20 years of age. Of the 10% to 15% of smoking
women who stop smoking during pregnancy, two thirds resume smoking within 1 year of delivery, exposing the
infant to second-hand smoke (1). Because most of these infants are term and near-term LBW infants, their
prognosis is good and their medical costs are only moderate. Nonetheless, reduction of smoking during
pregnancy can decrease the incidence of LBW and reduce the risk of long-term development delays (1,49,50).
Alcohol
The use of alcohol during pregnancy has been associated with both short- and long-term morbidity. Fetal alcohol
syndrome is a well-recognized consequence of excessive consumption of alcohol during pregnancy, occurring in
1 per 1,000 live births. However, several studies report an increase in LBW babies born to women who consume
between one and three drinks per day, resulting in an average decrease in birth weight of between 28 and 141
g. Almost 5% of pregnant women report binge alcohol use (five or more drinks on the same occasion or the
same day) (1,49). Recent studies performed in older alcohol-exposed infants show an increase in the number of
speech and language delays, which seem to be directly related to the amount of alcohol consumed during
pregnancy. In this study, the decreased mental developmental index in these infants also appears to be dose
dependent (50). These data and that of others strongly suggest that the avoidance of alcohol during pregnancy
can significantly improve perinatal outcome.
Illicit Drug Use

The exact incidence of illicit drug use during pregnancy is unknown, but several researchers report perinatal
cocaine exposure rates of between 0.4% and 27%, depending on the region of the country and the population
demographics (49,50 and 51). A myriad of neonatal complications have been reported as a result of cocaine
exposure, ranging from none to antenatal cerebral infarction. The estimated short-term costs of perinatal
cocaine exposure are between $33 million and $650 million (51). The long-term neurodevelopmental effects of
perinatal illicit drug exposure are just now
P.37
being studied carefully. Many studies in newborns have shown irritability and lability of state, decreases in
alertness and orientation, and abnormal muscle tone and reflexes. However, one third of newborn behavioral
studies have found no demonstrable effect of perinatal exposure. Further, studies performed in older infants
have failed to show significant effects of perinatal cocaine exposure on either the mental developmental index or
the physical developmental index (50).
Although perinatal exposure to any drug cannot accurately predict long-term outcome, such exposure often
exposes a child to multiple risksphysical, developmental, and psychosocial. If we can eliminate these risks, we
have every reason to believe that perinatal outcome will be enhanced.
Neurodevelopmental Sequelae
With survival rates approaching 95% for premature infants weighing between 1,200 and 2,500 g and 60% for
those between 500 and 750 g, the incidence and magnitude of the neurodevelopmental sequelae related to the
treatment of premature and critically ill neonates has become increasingly important (52). Long-term outcome
of infants treated in the NICU has an enormous impact on many aspects of society. Even though the vast
majority of children treated in the NICU survive intact, those who experience significant sequelae will need
considerable additional resources from the family, state, and federal government in order for them to achieve
their full potential. This impact is felt in the form of financial pressures placed on the family and public programs
to provide necessary medical services and on schools as they attempt to provide an appropriate learning
environment (53). The greatest risks of sequelae occur in the ELBW infants (less than 1,000 g birth weight) and
in large infants experiencing perinatal hypoxic-ischemic encephalopathy. These issues are discussed in detail in
other chapters; however, it is important to realize that after discharge from the NICU, infants with significant
neurodevelopmental sequelae need a system that not only provides long-term support for specialty medical
care, but also provides a lifetime of financial, social, and educational support for such children and their families
(54). Clearly, the best way to deal with these problems is to prevent them from occurring in the future.
REGIONALIZATION TODAY
Regionalization of perinatal care has been lauded by many as the single most important factor influencing the
birth-weight-specific neonatal mortality. Outlined in 1977 by the Committee on Perinatal Health of the National
Foundation of the March of Dimes in its landmark report, Toward Improving the Outcome of Pregnancy:
Recommendations for the Regional Development of Maternal and Perinatal Health Services, this concept served
as the guide for the development of perinatal services for the last 2 decades (7). The Committee's concept of
regionalization was based on the geographic concentration of neonatal intensive care services supported by
cooperative arrangements among hospitals within a region to provide, as a network, the necessary levels of care
defined in the document as level I, II, and III services. Although quite intuitive and logical, this concept of
cooperation is rather foreign to the American health care system, which is built on the philosophy of free
market, free enterprise, and competition. The level of planning and cooperation necessary to make
regionalization work is often resisted by all levels of health care providers. How then did regionalization take
hold? Many observers think that it came at the right time, when knowledge of high-risk mothers and neonates
was advancing rapidly. It came when the transfer of technology was confined to larger urban teaching hospitals
and academic centers. It came at a time when other hospital services were running at near capacity.
Under the plan, regions were not defined strictly by geography, but by tradition and the organizational skills of a
small number of highly respected early leaders in neonatology. This informal approach to the organization of
perinatal care worked well as long as cooperation was seen as mutually beneficial to all involved. In the late
1970s and early 1980s, the health care environment began to change significantly. Driven by dramatic changes
in the public financing of health care, the fragile alliance built on cooperation was being replaced by a drive for
competition (55).
Medicare's replacement of fee-for-service reimbursement with a prospective payment system caused hospital
inpatient census to drop as adult care was shifted to the outpatient clinics and ambulatory surgical centers.
Declining hospital margins with excess capacity in the form of empty beds began to drive competition. The
diffusion of technology into the community in conjunction with increased numbers of available neonatologists
made competition for neonatal patients a possibility. Managed care plans captured an increasing share of the
traditional indemnity plan markets and required hospitals to be full-service providers, making competition even
more of a reality. Obstetric and newborn services, once seen as avoidable losses, became a requirement of
participation in health plans. In some parts of the country, other factors hastened the move to competition. The
medical liability crisis of the 1980s and 1990s, with the huge legal awards for poor neonatal outcome,
particularly in the southern United States, increased the desire to have a neonatologist present at almost every
delivery. At the same time, the presence of a hospital-based neonatologist presented the opportunity and the
necessity to expand neonatal services beyond the delivery room into newly acquired NICU beds in an attempt to
cover the costs of the neonatologist and to present appropriate clinical challenges to him or her and the nursery
staff. In other regions, the overcrowding of level III units and the need to back transport convalescing neonates
provided additional opportunity for community hospitals to begin neonatal programs. The differences in the
levels of care, as defined by the COPH and by the American
P.38
Academy of Pediatrics, became blurred. A comprehensive study of regionalization revealed a general weakening
of structures and relationships, with many hospitals and physicians working outside the designed networks (55).
As a result, the COPH was reconvened in 1990 to respond to the changing health care environment and to make
recommendations for the regionalization of perinatal care in the 1990s and beyond. The COPH's report
reinforced the concept of regionalization of perinatal services, emphasizing the need for developing systems that
integrate all levels of care within the region into a matrix with specific mechanisms for quality review and
accountability between and among the various components. To facilitate this integration, the COPH suggested
the formation of state and regional perinatal boards, with the authority and responsibility for providing or
coordinating regional planning, monitoring access and data collection, and providing education (7). The plan's
foundation appeared to be very regulatory in nature at a time when competition and the market were still very
strong. As a result, very little progress has been made on the implementation of its recommendations.
So where are we today? It appears that regionalization of perinatal care still exists; however, its focus has
evolved from level III units to the level II units, which have expanded their roles. Many neonatologists at level II
units are providing labor room resuscitation and stabilization, as well as normal newborn and low-risk neonatal
care at level I hospitals in their community. If these infants need more acute care, they are often transported to
the local level II facility instead of the regional level III one. The availability of exogenous surfactant therapy,
safer mechanical ventilators, and expanded roles for members of the neonatal provider team have allowed more
of the LBW and VLBW infants to remain in level II facilities. Although this is troubling to some, the limited
outcome information on LBW and VLBW neonates treated at level II rather than level III facilities fail to show
any adverse outcomes as a result (56,57,58). It appears that the level III units have become the regional
referral centers for ELBW infants, for larger infants in need of high frequency ventilationextracorporeal
membrane oxygenation (ECMO)and for those in need of medical and surgical subspecialists. This redefinition
of regionalization is not too far from the 1993 recommendations. What's missing is the development of specific
relationships outlined in the report and the requirement that the various levels of care enter into formal
agreements for referral, data transfer, outcomes review, and followup.
What are the inherent concerns about the evolved system? There is a danger that smaller and smaller infants
will receive care at level II units as technology advances further. The guiding principle here should be the need
for, and availability of, pediatric specialists as well as the availability of postdischarge specialty care and
followup. Few level II units have sufficient numbers of critically ill neonates to justify on-site subspecialty
coverage, and very few have enough ELBW or VLBW infants to justify investment in posthospital care; however,
both are required to provide quality care for this population. Another concern is the effect that changes in
referral patterns have on level III regional referral centers. LBW and VLBW infants tend to be less expensive to
care for than ELBW infants, those with birth defects, and those in need of specialty care. As a result, the level III
units are experiencing more and more financial pressures as they receive the smallest, most complex patients
with the smallest reimbursement-to-cost ratio. Because public programs, which routinely are among the poorest
payers, cover so many premature infants, shifting the intensity of the level III units to even higher levels of
acuity could result in more financial stress for many. Although the primary goal of regionalization is to provide
the best, most cost effective care to today's neonates, we must remember that neonatology has evolved to
where we are today because of the discovery of new therapies and the training of the next generation of
neonatologists and neonatal specialists. A rational system of regionalized care must recognize that institutions
dedicated to the development of future therapies and the future providers of neonatal care cannot always
survive in a competitive environment. Our few truly educational and research institutions must continue to have
a sufficient resources and a patient population representative of all levels of care in order to develop the
workforce and therapies of tomorrow, today.
REFERENCES
1. March of Dimes. March of Dimes data book for policy makers, Washington, DC: March of Dimes Office of
Government Affairs, 2003.
2. Pickering LK, ed. Red book: 2003 report of the Committee on Infectious Diseases, 26th ed. Elk Grove
Village, IL: American Academy of Pediatrics, 2003:320321.
3. Pastor PN, Makuc DM, Reuben C, et al. Chartbook on trends in the health of Americans 2002. Hyattsville,
MD: National Center for Health Statistics, 2002.
4. The Annie E. Casey Foundation, Kids count data book, 2003 a pocket guide. Baltimore, MD: The Annie E.
Casey Foundation, 2003:7.
5. U.S. Department of Health and Human Services, Center for Disease Control and Prevention. Infant
mortality rate by country Hyattsville, MD: National Center for Health Statistics, 1998:114.
6. Nicholson WK, Frick KD, Powe NR. Economic burden of hospitalization for preterm labor in the United
States. Obstet. Gynecol 2000;96:95.
7. Committee on Perinatal Health. Toward improving the outcome of pregnancy: the 90's and beyond, White
Plains, NY: The National FoundationMarch of Dimes, 1993.
8. Gilstrap LC, ed. Guidelines for perinatal care, 5th ed. Elk Grove, IL and Washington, DC: American
Academy of Pediatrics, American College of Obstetrics and Gynecology, March of Dimes, 2003:17.
9. National Perinatal Information Center. American Hospital Association survey data tapes 19851988.
National Perinatal Information Center Newsletter 1990;Fall:1.
10. Schartz RM. Supply and demand for neonatal intensive care beds: trends and implications. J. Perinatol
1996;16:483.
11. Thompson LA, Goodman DC, Little GA. Is more neonatal intensive care always better? Insights from a
cross-national comparison of reproductive care. Pediatrics 2002;109:1036.
12. Goodman DC, Fisher ES, Little GA, et al. The relation between the availability of neonatal intensive care
and neonatal mortality. NEJM 2002;346:1538.
13. American Board of Pediatrics Neonatal-Perinatal Medicine qualifying examination statistics Chapel Hill, NC:
American Board of Pediatrics. Retrieved June 28, 2003, from http://www.abp.org/stats/WRKFRC/neostat.htm
P.39
14. Bhatt DR, Escobeda M, Kattwinkel J, et al. 1998 US neonatologist directory, perinatal pediatrics section.
Elk Grove, IL: American Academy of Pediatrics, 1996.
15. Pollak LD, Ratner IM, Lund GC. United States neonatology practice survey: personnel, practice, hospital
and neonatal intensive care unit characteristics. Pediatrics 1998;101:398.
16. American Academy of Pediatrics Committee on Fetus and Newborns. Manpower needs in neonatal
pediatrics. Pediatrics 1985;76:312.
17. Goodman DC, Little GA. General pediatrics, neonatology and the law of diminishing returns. Pediatrics
1998;102:396.
18. Goodman DC, Fisher ES, Little GA. Are neonatal intensive care resources located according to need?
Regional variations in neonatologists, beds, and low birth weight newborns. Pediatrics: 2001;108:426.
19. Goodman DC, Fisher ES, Little GA, et al. The uneven landscape of newborn intensive care services:
variations in the neonatology workforce. Eff. Clin. Pract. 2001;4:143.
20. Shiono PH, Behman RE. Low birth weight: analysis and recommendations. Future Child. 1995;5:4.
21. Resnick MB, Eitzman DV, Dickman H, et al. Data base management for Children's Medical Services
Regional Perinatal Intensive Care Centers program. JFMA 1983;70:718.
22. Sills J. Understanding catastrophic health care exposures. neonatal intensive careHow did we get here
and where are we going? Retrieved June 28, 2003, from http://www.amre.com/hc2003/summaries/sills.htm
23. March of Dimes Perinatal Data Centers. Economic costs of prenatal careHealth Care Utilization Project
nationwide inpatient sample, 1999. White Plains, NY: March of Dimes, 2002.
24. Rogowski J. Measuring the cost of neonatal and perinatal care. Pediatrics 1999;103:329.
25. Muldoon J. Florida profiles its hospitals. In: Today. National Association of Children's Hospitals and Related
Institutions, Alexandria, VA: 1996;Summer:10.
26. Richardson DK, Zupancic JA, Escobar GJ, et al. A critical review of cost reduction in the neonatal intensive
care unit. I. The structure of costs. J. Perinatol 2001;21:107.
27. Mugford M, Richardson DK, Zupancic, JA, et al. Response to Hugh MacDonald, letter to the editor. J
28. Stevenson RC, Pharoal PO, Stevenson CJ, et al. Cost of care for a geographically determined population of
low birth weight infants to age 89 years. II: Children with disability. Arch Dis Child 1996; 74:F118.
29. Rogowski J. Cost-effectiveness of care for very low birth weight infants. Pediatrics 1998;102:35.
30. Resnick MD, Eyler FD, Nelson RM, et al. Developmental intervention for low birth weight infants: Improved
early developmental outcomes. Pediatrics 1987;80:68.
31. Tommiska V, Trominen R, Fellman V. Economic costs of care in extremely low birth weight infants during
the first 2 years of life. Pediatr Crit Care Med 2003;4:157.
32. Committee on Children, Health Insurance, and Access to Care. Health insurance and access to care for
children. Washington, DC: The Institute of Medicine, National Academies Press, 1998.
33. Lewit EM, Baker, LS, Corman H, et al. The direct cost of low birth weight. Future Child 1995;5:35.
34. MacDonald H, American Academy of Pediatrics, Committee on Fetus and Newborn. Perinatal care at the
threshold of viability. Pediatrics 2002;110:1024.
35. Committee to Study the Prevention of Low Birth Weight. Preventing low birth weight. Washington, DC:
36. Lu MC, Lin YG, Prietto NM, et al. Elimination of public funding of prenatal care for undocumented
immigrants in California: a cost/benefit analysis. Am J. Obstet Gynecol 2000;181:233.
37. Imershein AW, Turner C, Wells JC, et al. Covering the costs of care in the neonatal intensive care units.
Pediatrics 1992;89:56.
38. Committee on the Consequences of Uninsurance. Health insurance is a family matter. Washington, DC:
The Institute of Medicine, National Academies Press, 2002:47.
39. Iglehart JK. The dilemma of Medicaid. NEJM 2003:2140.
40. Rosenbaum S. Medicaid. NEJM 2002;346:635.
41. MacDorman MF, Minino AM, Strobino DM, et al. Annual summary of vital statistics 2001. Pediatrics
2002;110:1037.
42. U.S. Department of Health and Human Services. Retrieved July 1, 2003, from http://www.hhs.gov/budget/
dccbudget.htm
43. National Commission on Children. Beyond rhetoric: a new American agenda for children and families.
Washington, DC: National Commission on Children, 1991:127.
44. U.S. Department of Health and Human Services, Center for Disease Control and Prevention. National Vital
Statistics Reports 2002;50(12):3.
45. Luke B, Bigger HR, Leugans S, et al. The cost of prematurity: a case-control study of twins vs. singletons.
Am J Public Health 1996;86:809.
46. Centers for Disease Control and Prevention. Contribution of assisted reproduction technology and
ovulation-inducing drugs to triplet and higher-order multiple birthsUS 19801997. MMRW 2000;49:535.
47. Chromitz VR, Cherry L, Lieberman E. The roll of lifestyle in preventing low birth weight. Future Child
1995;5:121.
48. Eyler FD, Behnke ML. Early development of infants exposed to drugs perinatally. Clin Perinatol
1999;26:107.
49. Phibbs C. The economic implications of perinatal substance exposure. Future Child 1991;1:113.
50. Sweet MP, Hodgman JE, Pena I, et al. Two-year outcome of infants weighing 600 grams or less at birth
and born 1994 through 1998. Obstet Gynocol 2003;101:18.
51. Hanke C, Lohaus A, Gawrilow C, et al. Preschool development of very low birth weight children born 1994
1995. Eur J Pediatr 2003;162:159.
52. Resnick MB, Gomatam S, Carter RL, et al. Educational disabilities of neonatal intensive care graduates.
53. Gagnon DE, Allison-Cook S, Schwartz RM. Perinatal care: the threat of de-regionalization. Pediatr Ann
1988;17:447.
54. LeFevre M, Sanner L, Anderson MA, et al. The relationship between neonatal mortality and hospital level. J
Fam Pract 1992;35:259.
55. Phibbs CS, Branstein JM, Buxtan E. The effects of patient volume and level of care that the hospital of
birth on neonatal mortality. JAMA 1996;276:1054.
56. Yeast JD, Poskin M, Stockbauer JW, et al. Changing patterns in regionalization of perinatal care and the
impact on neonatal mortality. Am J Obstet Gynecol 1998;178:131.
57. Mathews TJ, Menacher F, MacDorman MF. Infant mortality statistics from the 2000 linked birth/infant
death data set. National Vital Statistics Reports 2002;50:10.
58. Martin JA, Park NM. Trends in twins and triplet births 19801997. National Vital Statistics Reports
1999;47:27.
Chapter 4
Neonatal Transport
Karen S. Wood
Carl L. Bose
HISTORY
Neonatal transport began in 1900 with the development of the first mobile incubator for premature infants by
Dr. Joseph DeLee of the Chicago Lying-In Hospital (1). This hand ambulance provided warmth while
transporting premature infants to the hospital following home birth. The development acknowledged the need to
create a controlled environment for the transport of infants that simulated the inpatient setting. In 1934, the
first dedicated neonatal transport vehicle in the United States was donated to the Chicago Department of Health
by Dr. Martin Couney (2), following the closure of the Chicago World's Fair where the vehicle was used to
transport premature babies to the exhibit. The first organized transport program in the United States began in
1948 with the development of the New York Premature Infant Transport Service by the New York Department of
Health in conjunction with area hospitals (3,4). This remarkable system, created more than a decade before the
evolution of neonatal intensive care units (NICUs), incorporated many of the features of modern neonatal
transport programs, including around-the-clock staffing by specially trained nurses, dedicated vehicles, a clerk
to receive referral calls, and equipment designed specifically for neonatal transport. During a two-year period,
this program transported 1,209 patients, of whom 194 weighed less than 1,000 g (4).
Neonatal transport took to the air in 1958 with the first fixed-wing transport of a newborn infant by the Colorado
Air National Guard (2). The 1967 flight of a premature baby to St. Francis Hospital in Peoria, Illinois using the
Peoria Journal Star helicopter marked the first rotor-wing neonatal transport (2). Routine use of air
transportation for neonatal patients began in 1972 with Flight for Life of Denver's St. Anthony Hospital (5).
Proliferation of organized transport programs occurred in the late 1970s, in conjunction with regionalization of
perinatal care. Regionalization initially minimized the number of infants requiring transport by promoting
maternal-fetal transport. Regionalization also shifted the responsibility for transporting infants born in outside
centers to the tertiary care center. Subsequently, the next decade saw improvements in perinatal mortality (6)
and neonatal morbidity (7) as the percentage of very-low-birth-weight (VLBW) infants delivered in level III
hospitals increased.
Since the late 1980s patterns of referral dictated by schemes of regionalization deteriorated in many areas (8),
coincident with an increase in level II hospitals capable of providing some degree of neonatal intensive care. As
a result, increasing numbers of infants deliver at centers without subspecialists or the necessary support
services demanded by some VLBW infants. Community-based neonatal intensive care creates a need to
transport infants at a critical time in their illness, occasionally while receiving therapies such as high-frequency
ventilation or inhaled nitric oxide, which are not easily portable. Even in areas where regionalized perinatal care
persists and prenatal risk assessment is routine, unpredictable, emergent events may precipitate the delivery of
an infant in an unsuitable hospital. Collectively these situations mandate increasingly sophisticated neonatal
transport systems.
ORGANIZATION AND ADMINISTRATION

Neonatal transport can be performed by either the community hospital referring the patient (one-way transport)
or by the tertiary center receiving the patient (two-way transport). Two-way transport offers an economic
advantage, generally more highly skilled and experienced transport personnel (9), and may result in improved
survival (10,11). The American Academy of Pediatrics supports two-way transport (12), and in most perinatal
regions tertiary centers have assumed this responsibility. Of note the major disadvantage of two-way transport
is the time delay in getting the transport crew to the referring hospital. The remainder of this chapter discusses
two-way transport exclusively.
Administrative Personnel
The components of a transport program include those related to medical care and the nonmedical components
such as transportation, communications, finances, and

P.41
marketing. The medical components must fall under the direction of a physician who is credentialed to supervise
the patients served by the program. Direction of the nonmedical components and ultimate responsibility for the
program often rests with a hospital administration staff person (Fig. 4-1). The following is a brief discussion of
each of the potential contributors to the administration of a transport program (13).
Figure 4-1 The administrative structure of a typical neonatal transport program.
Hospital Administrator
Generally, a hospital administrator manages aspects of the program that are not directly related to patient care.
Many decisions regarding program operation require a cost-benefit analysis. While medical personnel are relied
on to provide an estimate of benefit, the hospital administrator must assess financial impact. Therefore, the
hospital administrator should be prepared to receive advice from medical personnel and develop the nonmedical
components of the program in consideration of the financial resources of the institution.
Medical Director
The medical director of a neonatal transport program is usually a neonatologist with expertise or a special
interest in transport. The medical director is ultimately responsible for the quality of care provided by the
transport team; this is particularly true if physicians do not participate directly in transport. The medical director
assumes responsibility for developing and updating training programs, equipment procurement, and treatment
protocols. The medical director, in conjunction with the coordinator of nonphysician personnel, must ensure that
all personnel have completed training requirements successfully and have satisfied the regulations of the
agencies that govern the various professional groups. The director also must develop and maintain a system for
reviewing the quality of care provided during transport.
Coordinator of Nonphysician Personnel

Each group of professionals (e.g., nurses, respiratory therapists, paramedics) on the transport team should have
a designated coordinator. The coordinator supervises the selection and training of personnel and develops
systems of peer review. Additional coordinator responsibilities include scheduling personnel, organizing
continuing medical education, ordering supplies and equipment, monitoring documentation standards, promoting
effective internal dynamics, and identifying the needs of team members. It is advisable to designate a single
person to coordinate team activities who will interface closely with the medical director and, in some programs,
the hospital administrator as well.
Consulting Neonatologists and Other Subspecialists

During the transport of a patient, it is important, and often mandated by state law, that a physician provides
consultation to the transport team. This physician, known as the medical control officer (MCO), is typically the
physician receiving the patient and often has already discussed the patient's care with the referring physician
and made recommendations about interim management. Given this broad consultative role to both the referring
physician and to the transport team, the MCO should be a person with extensive training, at a level in excess of
that available in the community hospital, such as a neonatologist, a trained pediatric subspecialist, or a
postdoctoral fellow. In addition, the MCO must be aware of the handicaps and hazards imposed by the transport
environment and must be familiar with the operational aspects of the program.
The Advisory Board

A neonatal transport program should be considered an extension of the inpatient unit to which it delivers
patients.
P.42
Therefore, the operation of the program should be reviewed periodically by representatives of all services
interfacing with the inpatient unit. These representatives, comprising the advisory board, might include the
following:
Medical Director of the NICU
Director of the Neonatal Division
Respiratory Therapy Administrator
Nursing Administrator
Outreach Education Coordinator
Director of Public Relations
Representatives of community hospitals involved in transport
Advice should be solicited from the advisory board about all major program changes because of the impact these
changes may have on their respective services.
The Transport Team

A variety of personnel participate in the inpatient care of infants, and all should be considered candidates for
caretakers during neonatal transport. These personnel include the following:
Neonatologists
Neonatal fellows
Pediatric house staff
Nurse practitioners
Transport nurses
Respiratory therapists
NICU staff nurses
Paramedics
The selection of the type of personnel used by each program is based on the unique aspects of that program;
however, some general principles apply that determine the relative desirability of various professionals. As the
number of transports increases, it becomes less practical to send physicians on transport. Neonatologists rarely
have the time to devote to frequent transports, and reimbursement is not sufficient to justify their presence.
Although participation in transport can be very educational, in high-volume programs, time spent on transport
by house staff and fellows potentially competes with other aspects of training. In addition, the interest in
participation and expertise may vary considerably among trainees. This is a particular problem if participation is
mandated. Pediatric residents who participate in transport should be senior-level trainees under close
supervision.
Most high-volume programs choose to use nonphysician personnel as attendants during transport. The use of
neonatal nurse practitioners offers an attractive alternative to physician attendance (14,15). Nurse practitioners
are highly skilled in neonatal stabilization and care and provide a consistency of expertise not usually
encountered in other professional groups. They are licensed in most states to perform all the diagnostic and
therapeutic procedures required during transport. The greatest disadvantages to the use of neonatal nurse
practitioners in some regions are their scarcity and high cost. They are also rarely trained, or willing, to
transport patients other than neonates.
As a cost-effective alternative to nurse practitioners, many centers train readily available NICU staff nurses to
participate in transport. In addition, most states permit them to perform invasive procedures as an extension of
their inpatient nursing role under guidelines and protocols approved by the Boards of Nursing. Therefore, NICU
staff nurses can be trained to provide all the care required by a critically ill neonate during transport. This
training often is extensive, however, because the cognitive knowledge necessary to diagnose disorders and the
experience to perform invasive procedures must be mastered. This extensive training must be considered when
estimating the cost of using staff nurses as compared to nurse practitioners. The requirement of training is
particularly burdensome when there is a high personnel turnover rate.
Most patients transported to the NICU have either respiratory failure requiring mechanical ventilation or are
receiving supplemental oxygen. Respiratory therapists should be considered when selecting transport personnel
because of their expertise in the use and maintenance of respiratory care equipment. The therapists' ability to
adapt this equipment to the unique environment of transport can be lifesaving, particularly in circumstances
when unexpected events occur. The only disadvantage of using therapists is the narrow focus of their usual
training. Further education and cross-training allows their scope of practice to be expanded.
Eliminating physicians from attendance during transport can create problems that must be anticipated. For
example, leadership of the team is not defined by the usual medical model in which a physician assumes this
role. Designating one member of the transport team as the leaderwho is accountable for communication,
decision making, and documentationsolves this problem.
Advisory personnel at the tertiary center, particularly physicians, often are unwilling to endorse a patient care
program that does not mandate initial evaluation by a physician. This resistance usually stems from a concern
for the well-being of the patient and can be overcome by the selection and training of competent nonphysician
personnel. The support and endorsement of an involved medical director may also be critical. A similar attitude
may prevail in community hospitals. Referring physicians may find it unacceptable to relinquish care of a
critically ill patient to nonphysician personnel. In an environment in which tertiary centers compete for patients,
this may be a motivation for maintaining physician attendance during transport. Most referring physicians,
however, are concerned only with transferring their patients in a safe and timely fashion. Anecdotal experience,
as well as retrospective and prospective studies, suggests that properly selected and trained nurses provide a
level of care during transport that approximates the level provided by physicians (5,16,17 and 18). Once a
nonphysician team demonstrates competence and efficiency, the concerns of most referring physicians vanish.
P.43
Because the use of specially trained nonphysician personnel represents both a safe and economic alternative to
physician participation in neonatal transport, most programs now rely on nonphysician personnel for patient care.
Transport personnel must be proficient in cognitive knowledge of neonatal diseases, management principles of
acute problems, and technical skills. The method and extent of training necessary to reach proficiency will
depend on the type of personnel; however, the pattern of preparation will be similar for all professionals (19).
Cognitive knowledge is best provided in didactic sessions in conjunction with self-study exercises. Management
principles may also be taught in a didactic setting, but refinement of these skills requires repeated experiences
in the inpatient setting. Laboratory simulation of technical skills, such as intubation, umbilical vessel
catheterization, and thoracostomy tube placement, provides a good introduction to these procedures. These
skills can then be refined in the inpatient setting under supervision. Demonstration of proficiency in these areas
should be ensured by examination or observation by a qualified supervisor. After initial preparation, a period of
training should be provided, during which the trainee accompanies a more experienced team member on
transport. Final certification of competence should be awarded by both the medical director and the coordinator
for the trainee's professional group.
Communication
The quality of the communication system that supports a transport program may be the key determinant for its
success. The communication system serves two basic functions: to provide a point of access for referring
physicians, and to coordinate the activities of the transport team (20). A single call by the referring physician
should provide access to all of the neonatal services of the tertiary center. The use of a toll-free hot line, often
associated with a memorable acronym, is favored by some centers (21). Alternatively, referring physicians can
call the NICU directly. If consultation is requested, the referring physician should be connected in a timely
fashion with a consultant of appropriate training. If transfer is requested and deemed appropriate, an available
bed in the NICU of the tertiary center, or an alternate center if necessary, should be identified. Bed procurement
and all subsequent details of the transport should occur without additional calls by the referring physician.
Locating an available and appropriate site of care may be difficult because of a shortage of NICU beds or the
lack of availability of subspecialty support in some areas. These regions often benefit from an organized system
of identifying available resources. Several such programs exist and are of two varieties. In some areas,
sophisticated computerized communication networks link neighboring centers (22). An alternative is an operatorassisted central referral or bed locator system. These systems speed the referral of patients and relieve both the
referring physician and the physician at the tertiary center of the burden of placing numerous calls to locate a
bed.
Once the decision is made to transport the patient and an admission bed is located, the role of the
communications system shifts to dispatching the team and disseminating information about the transport. In
this role, the system is best served by a communication center that is staffed and equipped for emergency
medical service functions. The referring hospital should be informed of the estimated time of arrival and of any
necessary preparations for the arrival of the vehicle. The receiving unit should be notified and be provided with
medical information necessary for admission of the patient.
During the conduct of the transport, periodic communication between the dispatch center and the vehicle
operator is advisable. Unexpected delays or mishaps are identified promptly and appropriate action taken. Some
high-volume transport programs use satellite-tracking systems to monitor the movement of their transport
vehicles, which can be exceedingly useful if diversion is necessary. When the transport team does not include a
physician, the team should have the capability of communicating directly with the consulting physician at all
times. This level of communication is mandated by some states' nurse practice acts. Communication capabilities
are usually a trivial problem while in the referring hospital, however they can present a challenge during transit.
The gravity of this problem declines each year with the improvement of telecommunications equipment. Cellular
phones are typically used during ground transport because of the broad coverage in most areas and the general
familiarity of medical personnel with this type of communication. The use of VHF and UHF radios with patching
devices to phone lines is an alternative during flight. Typically, air traffic control, medical control, and general
communications have separate frequencies.
Many communication centers are equipped with automated devices that record all communications. Although not
essential, the recorded transmissions may be valuable educational tools and aids in identifying system errors; in
addition, they are often critical if a medicolegal question arises.
Communication should not end with the conclusion of the transport. The transport team should contact both the
patient's family and the referring facility to relate the events of the transport. The receiving physician should
update the referring physician following admission and give further followup information at regular intervals,
including at the patient's discharge. This update should be expedited if an acute event occurs and should occur
immediately in the event of death. Even with the ease of communication systems, failure to effectively
communicate followup information remains one of the most common criticisms of tertiary care centers.
Financial Considerations
Subjecting a transport program to periodic cost-benefit analyses is a critical aspect of the program's operation.
The following elements should be included in the cost of operation:
P.44
Medical components
Nonphysician personnel salaries/benefits
Salary support of the medical director
Equipment and supplies
Medication
Expenses related to education of personnel
Nonmedical components
Administrative overhead
Vehicle operation, maintenance, insurance
Communications
Educational and marketing material
Identifying the costs associated with the program may be difficult if its operation is financially integrated into the
operation of the NICU. For example, personnel costs often are difficult to quantify because, except in very highvolume programs, transport personnel usually contribute to inpatient services during transport duty time.
Therefore, the cost allocated to the transport program should be discounted based on this contribution. The
proportion of time devoted by the medical director is even more difficult to quantify and often is ignored in the
financial analysis. The cost of equipment most easily separates from the cost of inpatient services as transport
equipment rarely is used for other purposes. Included in estimates of equipment costs should be allowances for
depreciation and maintenance.
The nonmedical components of a program are often more costly than the medical components because of
expenses related to transportation. This is particularly true when air transportation is used. Sharing resources
with other hospitals or agencies can minimize these expenses. Ground ambulances can be shared with local
emergency medical service agencies or be used for convalescent transport. Aircraft can be used by a consortium
of hospitals. The major disadvantage of the shared approach is the possibility of a vehicle being unavailable at
the time of a request for transport; however, the potential for this occasional conflict may be far outweighed by
the cost reductions.
The revenues of a transport program come from three general sources: reimbursement, support from
governmental agencies, and support from other extramural organizations (23). Support from government and
charitable organizations is unusual in the United States, and hospitals are increasingly dependent on
reimbursement to support transport programs. Most third-party payers will reimburse the majority of the initial
transport as long as the care rendered at the receiving hospital was unavailable at the referring hospital.
Reimbursement for back transports is less consistent. In general, the cost of a transport program exceeds its
revenues. Subsistence of the program, therefore, depends on financial assistance from the sponsor hospital.
The decision to fund a transport program usually is based on a favorable cost-benefit analysis. Providing a
service that is unavailable at the outside hospital is clearly of benefit to the patient. Benefit can be quantified by
reductions in mortality, morbidity, and length of hospital stay. Among low-birth-weight (LBW) infants with
respiratory disease, one study has demonstrated that the services of a hospital-based neonatal transport team
reduce hypothermia and acidosis, the greatest prognostic indicators of mortality (10). However, beyond this
study little evidence exists to support the benefits of neonatal transport teams. In an attempt to quantify the
benefits of a neonatal transport program, the most prudent approach may be to scrutinize carefully the type of
patients being transported to ensure potential benefit from transport. These benefits should be combined with
nonmedical benefits to the institution, such as improved public relations and the recruitment of new patients.
Ultimately, many institutions in the United States elect to support a neonatal transport program, despite its
financial disincentives, in order to increase occupancy of NICU beds.
A potential economy for transport programs may be to combine services, either within a program or between
programs. An example of the former would be to cross-train members of specialty transport teams (e.g.,
pediatric, neonatal, and adult) such that the total number of personnel can be reduced. This strategy invariably
results in some loss of expertise but may be necessary to ensure financial viability. Collaboration between
programs may include sharing vehicles or teams. Smaller institutions may benefit from outsourcing entirely by
contracting with larger medical centers for the provision of all transport services.
TECHNICAL ASPECTS
The Transport Environment
The principles of care provided during transport resemble the principles of inpatient care. Any differences in
practice arise from the unique features of the transport environment (24). Many featuresincluding excessive
noise, vibration, improper lighting, variable ambient temperature and humidity, changes in barometric pressure,
confined space and limited support servicescan create problems during transport. The impact of these
environmental factors relative to the mode of transportation is summarized in Fig. 4-2.
Figure 4-2 Environmental factors and their impact relative to different modes of transportation.
P.45
Noise
High sound levels, in the range of 60 to 70 decibels, are inherent to the NICU (25,26 and 27); but levels
recorded on transport are significantly higher, on the order of 90 to 110 decibels (28,29). The effects of
exposure to high sound levels on the neonate are not known, but the possibility of physiologic changes is
suggested by studies of hospitalized infants (30,31). Brief exposure to high sound levels probably has little longterm effect on transport personnel; however, repeated exposure over time may result in hearing loss. Personnel
should protect themselves from exposure by using sound-attenuating devices. Probably the most significant
problem resulting from high sound levels is the inability to use auscultation to assess the patient. This handicap
must be recognized before transport, and alternative methods for assessing heart rate and respiratory
sufficiency must be available during transport.
Vibration
Vibration exposure is a problem unique to the transport environment (29,32,33). The physiologic consequences
in patients of this exposure are not known. Animal studies and investigations using healthy adults suggest that
negative effects on the autonomic and central nervous system may occur (34,35 and 36). The vibrational effects
on transport personnel are potentially important. For example, a typical helicopter transport results in vibration
exposure associated with reduced personnel efficiency (37). The overt symptoms of motion sickness resulting
from low-frequency vibration may be incapacitating. A more subtle manifestation of motion sickness, termed the
sopite syndrome, also may affect transport team members (38,39). The symptoms associated with this
syndrome include drowsiness, inability to concentrate, and disinclination to communicate with others. The sopite
syndrome is common among crew members during transport, regardless of the mode of transportation (40).
The impact on patient care is poorly understood but may be significant.
The effect of vibration on equipment also poses a major problem. Monitor artifact is a common phenomenon.
Personnel should be familiar with monitor artifact and with the use of alternative monitoring techniques. The
selection of equipment should be made in consideration of resistance to the effects of vibration. Premature
failure of equipment secondary to vibrational damage should be anticipated, and preventive maintenance should
be on an accelerated schedule.
Poor Lighting
Improper lighting in transport vehicles is a common problem. The patient care compartment should have
illumination to 400 lux (41). In addition, high-intensity directional lighting (1,000-1,500 lux) should be available
for procedures. The eyes of the patient, as well as those of the driver or pilot, should be screened from these
light sources.
Heat Loss
The difficulties in maintaining a neutral thermal environment are accentuated during transport because of the
increased opportunities for heat loss. Hypothermia can be a significant problem during transport and has been
linked to increased mortality (10). Heat loss in the transport environment usually occurs by two mechanisms:
convection and radiation. Heat loss can be minimized by use of a double-walled isolette, avoiding opening of the
isolette, heating the transport vehicle, creating barriers between the isolette and cold surfaces and limiting time
in transit.
Variable Humidity
Transport teams often elect not to humidify respiratory gases both for simplicity and to eliminate the negative
effects of water vapor pressure on infants in respiratory failure. This is a reasonable approach assuming short
transport times; however, long-term effects of poor humidification include dehydration and increased tenacity of
secretions. Therefore, gas humidification and close attention to hydration are desirable for longer transports.
Variable Altitude
Changes in altitude that occur during air transport present a potential hazard to an acutely ill neonate because
of the phenomena that occur during ascent. As altitude increases the following occur:
Air temperature decreases
Partial pressure of gases decreases
Total atmospheric pressure decreases
For the change in altitude to be clinically important, ascent must be of significant magnitude (in excess of 5,000
feet). At constant temperature, gas volume expands as atmospheric pressure decreases:
Therefore, gases contained in spaces not in continuity with the atmospheresuch as those in cuffed
endotracheal tubes, sinuses, middle ear canals, pneumothoraces, pneumatoses, intrapulmonary cysts,
pulmonary interstitial emphysema, intracranial and intraglobal air, and air spaces distal to obstructed bronchi
can all expand as atmospheric pressure declines. Attempts should be made to ventilate closed-space gas to the
atmosphere when significant changes in altitude are expected. Also, the impact of gas expansion during ascent
can be minimized by the use of pressurized aircraft if a significant change in altitude during the conduct of a
transport is anticipated.
P.46
Confined Space
Space limitations in transport vehicles may impact care. The recommended floor space for the care of a critically
ill neonate in an NICU setting is 150 square feet (42); however, the standard ambulance has approximately 47
square feet and aeromedical helicopters have 22 to 36 square feet of workspace. Personnel must remain seated
and restrained while the vehicle is in motion; therefore, typically only one provider has any access to the patient
in transit.
An appreciation of the problems created by the transport environment and strategies to minimize their impact
are essential for safe transport. Some general principles include the following:
Prepare the transport vehicle. The vehicle should be retrofitted to simulate the inpatient environment as
much as is possible and practical. This generally requires the addition of supplemental lighting, sound
insulation, and a regulated heating-cooling system.
Assess and stabilize the patient extensively before transport. Other than surgical emergencies, neonates
have problems that can be managed adequately by the transport team. There is rarely urgency in
returning to the tertiary center, and time spent in the community hospital preparing the patient for
transport is not time wasted. Stabilization will prepare the patient for the highest risk period, the time in
transit between hospitals.
Monitor electronically all possible physiologic parameters. Because of the dynamic nature of the diseases
in most transported patients and the inability to assess patients by physical examination in transit,
electronic monitoring is critical to the identification of significant changes in physiology.
Anticipate deterioration. All possible forms of deterioration should be anticipated before transport, and
strategies to support the patient in the event of deterioration should be planned. Application of this
principle can result in the performance of procedures or therapies that may not be necessary in the
inpatient setting.
Equipment
Considerable effort has been devoted to the development of devices specifically for neonatal transport, resulting
in greater safety and efficacy. The following is a list of the major pieces of equipment used during transport:
Essential equipment
Portable incubator
Mechanical ventilator
Cardiorespiratory monitor
Blood pressure transducer
Transcutaneous O2 monitor or pulse oximeter
Intravascular infusion pumps
Air-oxygen blender
Suction apparatus
Desirable equipment
Body temperature monitor
Transcutaneous CO2 or end-tidal CO2 monitor
Noninvasive blood pressure monitor
Airway humidification system
Blood gas/glucose analyzer
Although these devices can be purchased individually and either carried separately or attached to the incubator,
it usually is advisable, and often more economical, to purchase a modular incubator that includes many of the
devices listed. Modular transport incubators have been designed to minimize space and weight. They also use a
common battery power supply for most devices. Several transport incubators are commercially available. The
logical choice for each program often depends on the size, weight, and heating capability of the unit.
With the increased use of inhaled nitric oxide, many transport teams have acquired the ability to provide this
gas during transport. The inhaled nitric oxide delivery system can be cumbersome. Some transport teams use a
small, portable, commercially available device, while other teams have constructed their own system for
transport (43).
Small accessory equipment and supplies can be divided into respiratory care supplies and nursing supplies.
These supplies can be carried in packs or equipment bags (Tables 4-1,4-2,4-3 and 4-4). They should be
organized in a recognized and reproducible fashion. This technique will aid in rapidly locating an item during
transport and assist in restocking after use.
Transport Vehicles
An essential component of neonatal transport is rapid, safe transportation. The types of vehicles in use include
standard ambulances, specially prepared ground ambulances, rotor-wing aircraft, and fixed-wing aircraft. The
selection of one or more of these vehicles to support a neonatal transport program usually is based on patient
population, resources, geography, and practical issues, such as the use of the vehicle by other hospital-based
services (44,45).
Ambulances are economical, available, and least affected by weather; however, they generally require
retrofitting to make them acceptable for neonatal transport. Extensive retrofitting, including the addition of
radiant heat and a blood gas analyzer, improves patient care capabilities but dramatically increases costs and
decreases usefulness for other services. The major disadvantage of ground ambulance transport is time
consumption, which can be prohibitive if frequent long transports are anticipated.
Rotor-wing aircraft minimize transit time and, within a 150-mile radius, usually provide the fastest service. The
major disadvantages of helicopter transportation are the constraints of the patient care environment, the high
cost of operation, and the inherent safety risks (46,47) of helicopter flight. The cost of rotor-wing transportation
usually cannot be justified unless the vehicle can be shared by other emergency medical services.
Fixed-wing aircraft are more economical, spacious, quiet, and efficient as compared to rotor-wing aircraft;
however, they must travel between airports and therefore
P.47
additional transfers are required. These shuttles between the hospital and airport often are troublesome and
increase the likelihood of mishap. For these reasons, transportation by fixed-wing aircraft usually is
advantageous only for distances between hospitals in excess of 150 miles.
TABLE 4-1 NEONATAL NURSING PACK
Equipment
Amount
Equipment
Amount
Chest tube insertion kit

Omphalocele bag, sterile
Sterile lancets
1
1
4
Disposable transducer
Scissors
Hemostat
2
1
1
Blood culture bottle

Angiocath
18 gauge
Tape measure
Lubricant
Disposable blood pressure cuffs
sizes 2, 3, 4, and 5
1
2
1 each
22 gauge
2
2
24 gauge
Intraosseous needles
IV limb board
9
2
2
Pacifier
Bulb syringe
Sterile gauze
1
1
2
Rubber bands
Safety pins
Tape
Silk
Dermaclear
Stethoscope
IV fluids
6
6
Stopcocks
Extension tubing
Thoracostomy tubes
10 French
12 French
Digital thermometer
Umbilical catheters
2
1
1 roll
1 roll
1
2
2
1
D10W
1 500-mL bag
3.5 French
D5W
1 500-mL bag
5.0 French
LR
1 1000-ml bag
Heimlich valves
1 100-mL bag
2
Alcohol and Betadine swabs

IV extension T-connectors
Butterfly needles
19 gauge
23 gauge
25 gauge
Syringes
10 each
2
NS
Masks
Syringes
20-mL Luer Lok
60-mL Luer Lok
Transilluminator
Gloves, sterile
Size 6
Size 7
Suction catheters, sterile
Size 6 French
2
4
1
2 pairs
2 pairs
2
10 mL
3 mL
1 mL
60-cc catheter tip syringe
2
2
3
3
9
9
1
Size 8 French
Stockinette for caps
Cotton balls
Feeding tubes
8 French
5 French
2
2
4
60-cc luer lock syringe

4
Needles, 19 gauge
10
Tubing adapters (blunt needles) 2
Replogle, 10 French
2
Tubing
Mini-volume extension
1
2
2
Gowns
2
Low-volume extension
4
Blood gas syringe
2
Blood component and filter set
1
Tegaderm
1
IV ext. double T-connector
1
Protocol Manual
1
cc, cubic centimeter; D5W, 5% dextrose in water; D10W, 10% aqueous dextrose solution; IV,
intravenous; LR, lactated ringer; NS, normal saline.
DOCUMENTATION
Transport programs typically maintain record-keeping systems that are distinct from the inpatient record. An
accurate, thorough record of each transport is essential to provide permanent documentation of the care
rendered. The transport record should adhere to the standards of documentation of the sponsor institution. The
record also is a valuable tool for quality assurance and education. The critical components of a typical transport
record include the following:
Medical necessity documentation/referral form
Transport medical record
Parental consent form
Billing form
The trend in transport documentation, just as with inpatient charting, has been toward computerized medical
records using laptops or personal digital assistants in transport.
TABLE 4-2 NEONATAL RESPIRATORY THERAPY PACK
Equipment
Amount
EXTERIOR POCKETS
Oxygen tubing
Infant nasal cannula

1
Complete ventilation setup with 1 valve (plus one
exhalation
in isolette)
Equipment
Amount
Silicone adapter
O2 flowmeter nipple
2
2
One-way valve
Set of EKG lead wires
1
2
23G, 25G, -inch butterfly

Breath Tracker
3 each
1
Spare exhalation valve

Face tent
1
1
Albuterol
Racemic epinephrine
1
1
Treatment setup
Space blanket
1
1
Tape measure
Infant MVB bag, O2 tubing,
1
1 (plus one in
isolette)
PEEP valve
Thermal hats
Pulse oximeter sensors (N-25
2
2 each
and I-20)
Suction catheters 6 Fr, 8 Fr, 10 3 each
Fr
Infant oral airway
Normal saline
Silk tape
4 vials
1 roll
NCPAP prongs (#1,#2)

Luekens trap
1 each
1
INTERIOR OF BAG
Airway supplies
Laerdal masks
No. 0
No. 1
No. 2
Infant McGill forceps
Laryngoscope handle
Blades, Miller/Shaw #0,#1
Other equipment
Benzoin applicators
Alcohol preps
Adjustable wrench
E-tank wrench
Cable ties
Scissors
2
2
1
1
1
1 each
6
4
1
1
10
1
Oxygen connectors
Hemostat
Briggs T-adapters
15-mm adapter
O2 connectors (NCG, OES, PB)
Air connectors (NCG, P-B)
EKG lead pads
Three-way stopcock
E-Z Heat hot packs
Stethoscope
2
1
2
2
1 each
1 each
3
2
4
1
1-mL and 3-mL syringe

3 each
ABG kit
3
Low-volume extension tubing 4
Full set Mini-Med Tubing
1
Endotracheal tubes
2.5 mm
3
3.0 mm
3
3.5 mm
4.0 mm
3
3
9-volt battery
1
4.5 mm
3
Assorted laryngoscope bulbs 4
Pedi-cap detector
2
Adjustable venturi
2
Istat Pack
1
ABG, arterial blood gas; EKG, electrocardiogram; Fr, French; MVB, manual ventilation bag; NCPAP,
nasal continuous positive airway pressure; PEEP, positive end-expiratory pressure.
P.48
QUALITY ASSURANCE
Performance review of the transport program should be a continual process. All activities of the program should
be reviewed periodically to ensure that standard operating procedures are being observed. These reviews are
best conducted by people directly related to the program activities. Also, the medical care provided by the team
should be scrutinized for adherence to protocols and quality assurance. For nonphysician teams, the medical
director or a physician designate should conduct this level of review. In addition, to assure that every chart is
reviewed and problems discovered in a timely fashion, peer review in the immediate post-transport period is
valuable.
Quality assurance activities should be closely linked to education and research. Review of individual transport
records can be an extremely valuable method of identifying transport personnel in need of further education and
training. The compilation of reviews and the monitoring of patient outcomes provide an assessment of the
efficacy of existing protocols and procedures and may identify a need to alter program activities. In addition,
new therapies and equipment can be evaluated using existing quality assurance techniques.
General guidelines for developing quality assurance programs have been published (48,49). Guidelines, specific
for air transport, have been produced by the Association of Air Medical Services (50). The Commission on
Accreditation of Medical Transport Systems (CAMTS) provides external quality assurance reviews for transport
programs and allows programs to benchmark themselves against measurable standards (51). A number of
states require CAMTS accreditation, and some states use CAMTS certification in lieu of state regulations (52).
TABLE 4-3 NEONATAL MEDICINE PACK
Drug
Isotonic saline vials, 20 mL

Heparin
Sodium chloride (2%)
Phenytoin
Naloxone
Epinephrine 1:1,000, 1 mL
Epinephrine 1:1,000, 30 mL
Ampicillin, 500 mg
Ampicillin, 250 mg
Gentamicin, 20 mg
Acyclovir
Cefotaxime, 1 g
Flumazenil
Calcium gluconate
Dobutamine
D50W
Calcium chloride Bristoject
Atropine Bristoject
Survanta
Sterile water vials, 20 mL
Heparin lock flush
Dexamethasone
Furosemide
Artificial tears
Lidocaine 1%
Digoxin
Adenosine, 6 mg
Hydralazine, 20 mg
Clindamycin
Aminophylline
Dopamine
Pancuronium
Vecuronium
Amiodarone
Sodium bicarbonate
Lidocaine 2% Bristoject
Vitamin K, 1 mg/mL
Neonatal Narcotic Pack
Morphine 2 mg/mL
Midazolam 5 mg/mL
Phenobarbital 130 mg/mL
Fentanyl 50 g/mL
Lorazepam 2 mg/mL
D50W, 50% dextrose water.
Amount
4
1
1
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
2
4
2
1
1
1
1
1
2
1
1
1
1
1
1
2
1
1
1
2
1
1
1
1
P.49
PSYCHOSOCIAL CONSIDERATIONS
Psychological Impact on the Family
It is impossible to eliminate the parental anxiety associated with neonatal transport; however, a few techniques
exist that may help families in coping with this difficult situation. The transport team should provide the family
with as much information as possible about the nature of their child's illness, the therapies and equipment that
will be used, the NICU to which the infant will be transported, and the professionals who will provide care. A
member of the referring hospital staff should be in attendance during this discussion in preparation for dealing
with questions that may arise after the departure of the transport team (53). This information should be
provided both verbally and in written form. Many teams use brochures that describe their service, and provide
relevant phone numbers and directions to the tertiary care center.
Parents should see their infant before departure from the referring hospital. The benefit of this interaction
outweighs any delay in departure (54). This contact should be encouraged prior to the transport of even the
most critically ill infant or when parents are reluctant to view their child. When possible, a photograph of the
infant should be left with the family. On arrival in the receiving hospital, the transport team should call the
family immediately to reassure them that their child has arrived safely. The transport team should alert the
tertiary center staff to any unusual problems the parents might have in coping with their child's illness.
Relationships with Referring Hospital Personnel

Transporting a neonate from a community hospital to a tertiary center has the potential to either improve
dramatically the relationship between institutions or to cause irreparable damage. Each transport represents an
opportunity for success or failure. To ensure success, referring personnel must have easy access to and a quick
response time from the transport team. Rapidity of response often is critical from a public relations standpoint,
even when the infant's medical condition does not mandate speed.
Even the most responsive service will fail to satisfy personnel in the referring hospital if the team does not
conduct itself appropriately. The team must understand the psychological milieu surrounding a transport. The
event is often emotionally charged because of the acute nature of the infant's illness and feelings of inadequacy
on the part of the referring hospital personnel. These feelings seem to arise even when excellent,
comprehensive care is provided. Referring hospital personnel may be very sensitive to criticism, and any critique
of care, unless requested, should be deferred until a later time. The team should seek information about the
history and condition of the infant before their arrival, and they should ask for assistance from referral hospital
personnel when practical. Members of the transport team should clearly state appreciation for the contribution
made by the referring staff. The need for performing all procedures should be explained, this being particularly
important when referring personnel have made the decision not to perform a procedure because of their lack of
understanding of the transport environment. Nonphysician teams must avoid conflicts with referring physicians
over the need for therapies or procedures. Any disagreements should be resolved through discussion between
the referring physician and the consulting physician in the tertiary center.
TABLE 4-4 BACK TRANSPORT PACK
Equipment or Drug
Amount
EQUIPMENT
Equipment or Drug
Amount
1
1
2
6
Minimed full set

Minimed half set
Oximeter probes
EKG electrodes
Hot packs
Small
3
1
Large
Silver thermal hats
1
3
10
1
1
1
Small
Large
IV limb board
Intraosseous needle
BP cuffs (2,3,4,5)
Scissors
Hemostat
1
1
1
1
1 each
1
1
Butterfly needles
23 gauge
T-connector
Angiocath, 24 gauge
2
4
25 gauge
ET tubes, 2.5 through 5.0
2
2 each
Safety pins
Tape
6
2 rolls
Istat Pack
Lancets, sterile
Stethoscope
Syringes
1 mL
3 mL
10 mL
20 mL
60 mL
Needles, 19 gauge
Bulb syringe
Feeding tubes
5 French
8 French
1
5
1
Stylets
2
Laryngoscope and blades (Miller 0, 1) 1 each
Face masks, assorted sizes
1 each
Face tent
1
Manual ventilation bag
1
Nebulizer setup
1
Benzoin
Sterile gloves
Size 6
Size 7
Diapers
Pacifiers
Venturi tubing
Nasal cannula
Alcohol and Betadine swabs
Thermometer
Suction catheters
6 French
8 French
2
2
10 each
1
Oxygen tubing
2
2
3
1
MEDICATIONS
2
2
Isotonic saline
Heparin flush
Sodium bicarbonate
Epinephrine 1:10,000 Bristoject
2
2
2
1
10 French
Yankauer suction
2
1
Atropine Bristoject
Sterile water
1
2
Three-way stopcock
IV tubing
Low-volume extension
D10W, 500cc
D5W, 100 cc
D5/0.2 NaCl, 250 cc
Mini-volume extension
1
Normal saline, 250 cc
1
BP, blood pressure; cc, cubic centimeters; D5/0.2 NaCl dextrose 5% injection in 2% sodium chloride;
D5W, 5% dextrose in water; D10W, 10% aqueous dextrose solution; EKG, electrocardiogram; ET,
endotracheal; IV, intravenous.
P.50
LEGAL CONSIDERATIONS
Emergency medical services personnel may be held to a different standard of care while providing care at the
scene of an accident or in a transport vehicle compared to performing the same tasks in the inpatient setting
(55). Neonatal transport, however, is more closely associated with inpatient intensive care and less likely to be
considered an emergency service (providing extraordinary care under adverse conditions). Therefore, neonatal
transport personnel should assume that they have the same high risk of litigation as other perinatal caretakers.
Although there are few regulations and little case law defining the legal obligations of transport services,
understanding the principles that are likely to govern legal decision making will help guide programs in
establishing sound practices and limit risk of litigation (56). The principles of respondeat superior define the
hospital as the party responsible for governing the protocols and procedures followed by its personnel (57).
These principles appear to apply to mobile services as well as inpatient care. Therefore, the hospital that
sponsors a transport program is responsible for selecting and training the personnel and defining their scope of
practice. Logically, the medical director, as the medical professional delegated to ensure the quality of care, also
is liable for the governance of the team. Team members assume personal liability only if they perform outside
their enfranchised scope of practice.
Each transport program should construct a manual of operations clearly denoting its standard procedures. The
method used for selecting, training, and certifying personnel should be documented. Similarly, protocols and
procedures should be recorded and approved by the medical director. Activities of nonphysician personnel that
exceed
P.51
their usual scope of practice in the inpatient setting should be listed and approved by the respective governing
bodies (e.g., Board of Nursing). All documentation should be kept on permanent file.
During the conduct of a transport, the team should adhere to established protocols and procedures (58) unless
the patient's needs dictate an abridgement of usual standards. In this situation, advice from a consulting
physician should be sought, and the recommendations of this physician should be carefully noted in the patient
record.
Referring hospitals have both ethical and legal responsibilities to patients requiring interhospital transfer, with
the latter responsibilities outlined primarily in the Consolidated Omnibus Budget Reconciliation Act (COBRA) of
1985 (59). This federal legislation assigns to the referring hospital the responsibility to adequately stabilize the
patient prior to transport. The referring hospital also must establish an agreement with a hospital to receive the
patient and must ensure that the receiving hospital is capable of providing for the predicted needs of the patient.
Amendment of this act in 1989 added the requirement that referring hospitals make an effort to obtain written
consent from parents of a minor patient prior to transport. Failure to comply with these requirements is
considered medical abandonment.
Referring and receiving hospitals, and their personnel, have distinct responsibilities to patients at varying points
in time during the conduct of a transport. There is no single point at which the responsibility shifts absolutely
from referring to receiving hospital. From the time of a referral call to the arrival of the patient in the receiving
hospital, there is a stepwise decline in the responsibility of the referring hospital (Fig. 4-3). Critical events that
shift responsibility include:
Arrival of the transport team in the referring hospital
Assumption of direct patient care by the transport team
Departure of the team and patient from the referring hospital
Arrival in the receiving hospital (60)
Figure 4-3 Changing levels of legal responsibility for patient care. (From Brimhall, DC. The Hospital
Administrator's Perspective. In: MacDonald MG, ed., Miller MK, assoc. ed. Emergency transport of the perinatal
patient. Philadelphia: Little, Brown and Co, 1989: 148, with permission.)
The scope of nursing practice is usually established by two sets of regulations, the rules and regulations set forth
by their employing hospital or agency and the nurse practice acts in the state in which they practice. Transport
creates unique problems regarding nursing scope of practice because transport nurses often provide care in a
hospital other than their sponsoring institution, and sometimes they may practice in a state other than the one
in which they are licensed (60). Nurses in general are not permitted to practice under the supervision of a
physician not associated with their employing hospital. Therefore, while providing care in the referral hospital
and sharing responsibilities with the referring physician, a transport nurse must follow protocols and procedures
established by the sponsoring institution, or receive verbal orders from the receiving hospital's MCO. In
anticipation of problems during interstate transport, some adjacent states have established reciprocal
relationships for licensure; however, this is not routine. Many transport teams circumvent this problem by
admitting the patient to their home hospital at the time of first contact with the patient and thus team
members are considered to be practicing in their licensed state.
Vendors who provide transportation for medical personnel and patients are governed by either state or federal
legislation. The Emergency Medical Services Act of 1973 places the responsibility for the governance of ground
transportation under the guidelines of state emergency medical service regulations. Air transportation services
must comply with Part 135 of the Federal Aviation Administration regulations, which govern medical air
operations.
BACK TRANSPORT
The return transport of convalescing infants to community hospitals before discharge home is referred to as
back transport. Benefits of back transport include:
Reserves tertiary center resources for critically ill patients, decreasing overcrowding in these units (61,62)
Improves use of level I and level II center resources and helps prepare their personnel for the care of
acutely ill patients
Improves relationships between tertiary centers and their referral hospitals
Familiarizes primary care physicians with infants before discharge home
Improves family visitation and promotes family-infant bonding
Reduces the total cost of medical care (63,64)
There are also potential disadvantages associated with back transport, including the following:
Parental anxiety and loss of continuity of care caused by the change of caretakers
Occasional requirement for readmission to the level III center (65)
Lost opportunity by level III personnel to participate in convalescent care

P.52
Hazards and cost of transport
Lack of reimbursement by third party payers for the transport (64)
Back transport should be considered an option for all infants who no longer require the unique resources of the
level III center and for whom the tertiary center is not the site of subsequent primary care (65).
REFERENCES
1. DeLee JB. Infant incubation, with the presentation of a new incubator and a description of the system at
the Chicago Lying-In Hospital. Chic Med Rec 1902;22:22.
2. Butterfield LJ. Historical perspectives of neonatal transport. Pediatr Clin North Am 1993;40(2):221.
3. Losty MA, Orlofsky I, Wallace HM. A transport service for premature babies. Am J Nurs 1950;50:10.
4. Wallace HM, Losty MA, Baumgartner L. Report of two years experience in the transportation of premature
infants in New York City. Pediatrics 1952;22:439.
5. Pettett G, Merenstein GB, Battaglia FC, et al. An analysis of air transport results in the sick newborn infant:
Part I: the transport team. Pediatrics 1975;55(6):774.
6. Cifuentes J, Bronstein J, Phibbs CS, et al. Mortality in low birth weight infants according to level of neonatal
care at hospital of birth. Pediatrics 2002;109(5):745.
7. Hohlagschwandtner M, Husslein P, Klebermass K. Perinatal mortality and morbidity: comparison between

maternal transport, neonatal transport and inpatient antenatal treatment. Arch Gynecol Obstet 2001;265:113.
8. Richardson DK, Reed K, Cutler JC, et al. Perinatal regionalization versus hospital competition: the Hartford
example. Pediatrics 1995;96(3):417.
9. Bose CL. Organization and administration of a perinatal transport service. In: MacDonald MG, Miller MK,
eds. Emergency transport of the perinatal patient. Boston: Little, Brown and Co, 1989:43.
10. Hood JL, Cross A, Hulka B, et al. Effectiveness of the neonatal transport team. Crit Care Med 1983;11
(6):419.
11. Chance GW, Matthew JD, Gash J, et al. Neonatal transport: a controlled study of skilled assistance. J
Pediatr 1978;93(4):662.
12. American Academy of Pediatrics, Committee on Fetus and Newborn, and American College of Obstetricians
and Gyneco-logists, Committee on Obstetric Practice. Guidelines for perinatal care, 5th ed. Evanston, IL:
American Academy of Pediatrics and American College of Obstetricians and Gynecologists, 2002:58.
13. Brimhall D. Developing administrative support for the transport system. In: McCloskey K, Orr R, eds.
Pediatric transport medicine. St. Louis, MO: CV Mosby, 1995:56.
14. Mitchell A, Watts J, Whyte R, et al. Evaluation of graduating neonatal nurse practitioners. Pediatrics
1991;88(4):789.
15. Karlowicz MG, McMurray JL. Comparison of neonatal nurse practitioners' and pediatric residents' care of
the extremely low-birth-weight infants. Arch Peds Pediatr Adolesc Med 2000;154(11):1123.
16. Thompson TR. Neonatal transport nurses: an analysis of their role in the transport of newborn infants.
Pediatrics 1980;65(5):887.
17. Aylott M. Expanding the role of the neonatal transport nurse: nurse-led teams. Brit J Nurs 1997;6(14):800.
18. Cook LJ, Kattwinkel J. A prospective study of nurse-supervised versus physician-supervised neonatal
transports. JOGN Nurs 1983; 12(6):371.
19. American Academy of Pediatrics, Task Force on Interhospital Transport. Guidelines for air and ground
transport of neonatal and pediatric patients. Elk Grove Village, IL: American Academy of Pediatrics, 1999:37.
20. Conn AKT, Bowen CY. The communications network for perinatal transport. In: MacDonald MG, Miller MK,
eds. Emergency transport of the perinatal patient. Boston: Little, Brown and Co, 1989:93.
21. Perlstein PH, Edwards NK, Sutherland JM. Neonatal hot line telephone network. Pediatrics 1979;64(4):419.
22. Bostick JS, Hsiao HS, Lawson EE. A minicomputer-based perinatal/neonatal telecommunications network.
Pediatrics 1983;71(2): 272.
23. Risemberg HM. Financing a perinatal transport program in the United States. In: MacDonald MG, Miller
MK, eds. Emergency transport in the perinatal patient. Boston: Little, Brown and Co, 1989:85.
24. Bose CL. The transport environment. In: MacDonald MG, Miller MK, eds. Emergency transport of the
perinatal patient. Boston: Little, Brown and Co, 1989:195.
25. Philbin MK, Gray L. Changing levels of quiet in an intensive care nursery. J Perinatol 2002;22(6):455.
26. Kellman N. Noise in the intensive care nursery. Neonatal Netw 2002;21(1):35.
27. Robertson A, Cooper-Peel C, Vos P. Peak noise distribution in the neonatal intensive care nursery. J
Perinatol 1998;18(5):361.
28. Shenai JP. Sound levels for neonates in transit. J Pediatr 1977; 90(5):811.
29. Campbell AN, Lightstone AD, Smith JM, et al. Mechanical vibration and sound levels experienced in
neonatal transport. Am J Dis Child 1984;138:967.
30. Gadeke R, Doring B, Keller R, et al. The noise level in a children's hospital and the wake-up threshold in
infants. Acta Paediatr Scand 1969;58:164.
31. Blackburn S. Environmental impact of the NICU on developmental outcomes. J Pediatr Nurs 1998 13
(5):279.
32. Shenai JP, Johnson GE, Varney RV. Mechanical vibration in neonatal transport. Pediatrics 1981;68(1):55.
33. MacNab A, Chen Y, Gagnon F, et al. Vibration and noise in pediatric emergency transport vehicles: a
potential cause of morbidity? Aviat Space Environ Med 1995;66(3):212.
34. Floyd WN, Broderson AB, Goodno JF. Effect of whole-body vibration on peripheral nerve conduction time in
the rhesus monkey. Aerospace Med 1973;44(3):281.
35. Clark JG, Williams JD, Hood WB, et al. Initial cardiovascular response to low frequency whole body
vibration in humans and animals. Aerospace Med 1967;38(5):464.
36. Ando H, Ishitake T, Miyazaki Y, et al. The mechanism of a human reaction to vibration stress by palmar
sweating in relation to autonomic nerve tone. Internat Arch Occup Environ Health 2000; 73(1):41.
37. Adey WR, Winters WD, Kado RT, et al. EEG in simulated stresses of space flight with special reference to
problems of vibration. ElectroencephalogrElectroenceph Clin Neurophysiol 1963;15:305.
38. Graybiel A, Knepton J. Sopite syndrome: a sometimes sole manifestation of motion sickness. Aviationt
Space Environ Med 1976;47:873.
39. Lawson BD, Mead AM. The sopite syndrome revisited: drowsiness and mood changes during real or
apparent motion. Acta Astronauticat 1998;43(36):181.
40. Wright MS, Bose CL, Stiles AD. The incidence and effects of motion sickness among medical attendants
during transport. J Emerg Med 1995;13(1):15.
41. Patient compartment illumination. Federal specifications for ambulance KKK-A-1822B. Washington, DC:
National Automotive Center, General Services Administration, 1985. Para 3.8.5.1.
42. American Academy of Pediatrics, Committee on Fetus and Newborn, and American College of Obstetricians
and Gynecologists, Committee on Obstetric Practice. Guidelines for perinatal care, 5th ed. Evanston, IL:
American Academy of Pediatrics and American College of Obstetricians and Gynecologists, 2002:45.
43. Kinsella JP, Griebel J, Schmidt JM, et al. Use of inhaled nitric oxide during interhospital transport of
newborns with hypoxemic respiratory failure. Pediatrics 2002;109(1):158.
44. Schneider C, Gomez M, Lee R. Evaluation of ground ambulance, rotor-wing, and fixed-wing aircraft
services. Crit Care Clin 1992;8(3):533.
45. Brink LW, Neuman B, Wynn J. Air transport. Pediatr Clin North Am 1993;40(2):439.
46. King BR, Woodward GA. Pediatric critical care transportthe safety of the journey: a five-year review of
vehicular collisions involving pediatric and neonatal transport teams. Prehosp Emerg Care 2002;6(4):449.
47. DeLorenzo RA. Military and civilian emergency aeromedical services: common goals and different
approaches. Aviation Space Environ Med 1997;68(1):56.
48. Council on Medical Service. Guidelines for quality assurance. JAMA 1988;259(17):2572.
49. Joint Commission on the Accreditation of Hospitals and Health Organizations. Examples of monitoring and
evaluation in emergency services. Chicago: JCAHHO, 1988:13.
P.53
50. Eastes L, Jacobson J, eds. Quality assurance in air medical transport. Orem, UT: WordPerfect Publishers,
1990.
51. Accreditation standards of the commission on accreditation of medical transport systems, 3rd ed.
Anderson, SC: Commission on Accredition of Medical Transport Systems, 1997.
52. Frazier E. How many state EMS agencies require CAMTS accreditation for air ambulance services. Air Med
J 2001;20(1):8.
53. McBurney B. The role of the community hospital nurse in supporting parents of transported infants.
Neonatal Network 1988; 6:60.
54. MacNab AJ, Gagnon F, George S, et al. The cost of family-oriented communication before air medical
interfacility transport. Air Med J 2001;20(4):20.
55. Reimer-Brady JM. Legal issues related to stabilization and transport of the critically ill neonate. J Perinat
Neonatal Nurs 1996; 10(3):62.
56. Ginzburg HM. Legal issues in medical transport. In: MacDonald MG, Miller MK, eds. Emergency transport
of the perinatal patient. Boston: Little, Brown and Co, 1989:152.
57. Tonsic v Wagner, 458 Pa. 246; 329 A 2d 497;1974.
58. American Academy of Pediatrics, Task Force on Interhospital Transport. Guidelines for air and ground
transport of neonatal and pediatric patients. Elk Grove Village, IL: American Academy of Pediatrics, 1999:16.
59. Ross M, Hayes C. Consolidated Omnibus Budget Reconciliation Act of 1985. Soc Secur Bull 1986;49(8):22.
60. Brimhall DC. The hospital administrator's perspective. In: MacDonald MG, Miller MK, eds. Emergency
transport of the perinatal patient. Boston: Little, Brown and Co, 1989:147.
61. Jung AL, Bose CL. Back transport of neonates: improved efficiency of tertiary nursery bed utilization.
62. Zarif MA, Rest J, Vidyassagar D. Early retransfer: a method of optimal bed utilization of NICU beds. Crit
Care Med 1979;7:327.
63. Bose CL, LaPine TR, Jung AL. Neonatal back transport: cost effectiveness. Med Care 1985;23(1):14.
64. Phibbs CS, Mortensen L. Back transporting infants from neonatal intensive care units to community
hospitals for recovery care: effect on total hospital charges. Pediatrics 1992;90(1Pt1):22.
65. Lynch TM, Jung AL, Bose CL. Neonatal back transport: clinical outcomes. Pediatrics 1988;82(6):845.
Chapter 5
Telehealth in Neonatology
Sarah C. Muttitt
Mary M. K. Seshia
Liz Loewen
Health care is facing many challenges and changes with increasing pressure to improve access
and quality while reducing the administrative and financial burden of providing care. There is
growing expectation that technology has a key role to play in meeting these demands. Telehealth,
defined as the use of information and communications technology (ICT) to deliver health services,
expertise, and information over barriers of distance, geography, time, and culture, fulfills this role
(1). Telehealth can support services within the whole spectrum of health care including diagnosis,
treatment, and prevention of disease, continuing education of health professionals and
consumers, and research and evaluation. When well-integrated into routine clinical practice,
telehealth can improve the efficiency and cost effectiveness of the health care system by moving
people and information virtually rather than physically (1). Although not new, telehealth has
experienced rapid expansion over the past decade and is being utilized in a growing number of
medical specialties including dermatology, oncology, radiology, surgery, cardiology, mental
health, and home health care. While telehealth may not have reached the volume and maturity
required for large-scale randomized studies, the value of telehealth has been well accepted by
consumers and health care providers alike (1).
Telehealth implementation addresses three major issues: access to health care services;
retention, recruitment, and support of rural physicians and other health care providers; and
potential cost savings to the health care system and/or patients and their families. In order to
access specialty care, residents of rural areas are often forced to travel long distances at
significant cost, inconvenience and, in some cases, aggravation of underlying medical conditions.
Although some tertiary care centers provide itinerant specialty clinics, these services may not be
available where and when a patient requires specialist advice. Physician travel for itinerant clinics
also has associated risks and costs, including the loss of valuable time while traveling between
clinics. Telehealth has the potential to provide access to a broader range of comprehensive
primary, secondary, and tertiary health care services, more timely intervention, earlier
repatriation, and improved continuity of care for rural patients. Physicians and health care
providers in rural areas have limited direct access to peers, specialists, education, and
opportunities to participate in health care administration or professional association activities. This
sense of professional and social isolation often contributes to clinicians leaving positions
prematurely and the inability to recruit skilled practitioners, leaving rural communities largely
underserviced. Access to education and peer support through telehealth may impact retention and
recruitment and allow health care providers in rural settings to work to the full potential of their
scope of practice as well as provide more complex care closer to home.
Although telehealth is primarily driven by the demand for more equitable access to health care for
rural and remote residents, much attention has been given to the telehealth business case. Cost
avoidance and cost savings are constantly sought to offset the substantial costs of telehealth
implementation and operations. Although there are cost savings associated with a reduced
number of unnecessary medical transports, earlier patient discharge to community hospitals or
home, and decreased education and administrative travel, the longer-term savings associated
with more timely access to care resulting in less consumption of health care resources and
improved health outcomes have yet to be measured (1,2,3). The ability of telehealth to contribute
to greater efficiencies within the system may offer the opportunity to redirect any savings towards
improved patient services. Even if telehealth does not result in a reduction in total health
expenditures, improved access to quality health care services should be of high importance to
patients, providers, and health care funders (1).
P.55
THE HISTORY OF TELEHEALTH

Health care at a distance, or telehealth, has been practiced for decades using less sophisticated
communications technologies than we currently associate with telehealth. The National
Aeronautics and Space Agency (NASA) played an important role in the early development of
telehealth. The remote monitoring of crew, spacecraft, and environmental health has been an
integral part of NASA operations. Similarly, the United States (U.S.) military has been actively
involved in telehealth research and applications as a means of bringing medical expertise to those
injured in battle with less risk of injury to health care personnel. In the 1960s, the Nebraska
Psychiatric Institute became one of the first facilities in the United States to develop a two-way
link via microwave technology to provide education and consultations between specialists and
general practitioners. NASA's efforts at enhancing satellite communications provided the
opportunity to pilot telehealth in more remote locations such as Alaska in the early 1970s (4).
Satellite technology also facilitated the development of the Telemedicine Centre at Memorial
University of Newfoundland (MUN) in 1977. Using simple, low-cost audio-conferencing technology,
the MUN program linked hospitals, community colleges, university campuses, high schools, town
halls, and education agencies throughout the province for educational programs and transmission
of medical data (5). In the 1980s, there was a flurry of telehealth activity in North America as well
as around the world, with new projects in Australia, New Zealand, the United Kingdom, France,
and Norway. Early projects were largely focused on the technical feasibility of telehealth and often
were developed around a single application and single clinical champion. Because of the high
capital costs of these projects, they were also extremely dependent on grant funding. Many came
to end when funding expired or the clinical champion moved on to other areas of research interest.
In the late 1980s and early 1990s, telehealth technology became more robust and less expensive,
making telehealth a more viable alternative for health care service delivery. The past decade has
seen rapid expansion of telehealth sites and applications. The Telemedicine Information Exchange
currently lists more than 160 telehealth programs internationally (6). A better understanding of
the human factors associated with telehealth success has allowed improved planning, deployment,
and management of telehealth programs. However, the majority of telehealth activity worldwide
still remains grant funded or hospital supported with consequent vulnerability to annual funding
cycles. Today's projects focus on assessing sustainability. It is becoming apparent that telehealth
solutions have to be integrated within the traditional healthcare system to be sustainable. With
careful attention to cost-effectiveness and quality services, telehealth is destined to become a
standard component of health service delivery.
TECHNOLOGY
There are two main types of communications in telehealth. Asynchronous, or store-and-forward,
involves the capture and later transmission of data or images for dissemination or interpretation.
Teleradiology, the sending of radiographs, computed tomography (CT) scans, or other digital
scans, is the most common store-and-forward application of telehealth in use today, and is often
integrated into larger picture archiving and communication systems (PACS). Pathology and
dermatology are other specialties that typically use store-and-forward technology for remote
diagnosis. Synchronous, or real time, implies the transmission of information instantly and is
primarily associated with the use of videoconferencing to support face-to-face consultation
between a patient in one location and a provider in another. Almost all medical specialties have
found an application for the use of videoconferencing technology and with the addition of
appropriate peripheral medical devices, such as stethoscopes, otoscopes, and examination
cameras, a comprehensive examination can be conducted remotely. Some telehealth applications
use a combination of store-and-forward and videoconferencing technologies to allow both the
review of still images and interactive consultation with peers and patients. In all cases, the clinical
requirements must drive the technical solution. The price and performance of telehealth
technology has improved dramatically over recent years and in many cases, off-the-shelf
hardware now provides the necessary functionality at much lower cost than systems specifically
designed for telehealth. All equipment should comply with accepted technical standards to ensure
quality, flexibility, and compatibility between systems.
In addition to end points, telehealth requires a telecommunications network to facilitate the
exchange of information. Although telecommunications infrastructure in urban settings has
developed remarkably over the past decade, the primary focus of telehealth has been to serve
rural and remote populations where connectivity continues to be a considerable challenge.
Requirements for bandwidth (communication channel capacity) vary depending on the application.
The higher the bandwidth, the more information can be sent in a measured time period. POTS
(plain old telephone system) may be appropriate for transmitting low volumes of nonurgent X-ray
images between two destinations for a teleradiology service. Larger volumes of images or a need
for urgent interpretation would require a higher bandwidth solution. Similarly, higher bandwidth is
required to support quality interactive videoconferencing for clinical applications. Although urban
sites may be able to choose between a number of suitable solutions such as ISDN (Integrated
Services Digital Network), DSL (digital subscriber lines), or high-speed cable, geographically
remote communities may only have access to the necessary bandwidth through satellite or other
high-cost wireless solutions. Telecommunications costs for rural education and health care
networks in the United States
P.56
are heavily subsidized through the Universal Service Fund, a fund generated through contributions
from telecommunications companies. In Canada, the federal government continues to support the
deployment of broadband to many rural, northern, and isolated communities through initiatives
such as the Broadband for Rural and Northern Development and the National Satellite Initiatives.
In some cases, sharing infrastructure costs with other sectors, such as education, justice, or
industry, may improve the viability of telehealth in a small, remote community. The availability of
low-cost telecommunications solutions is critical to the expansion and sustainability of telehealth
in many of the neediest areas, including developing countries.
The advent of IP (Internet Protocol) videoconferencing is impacting the design and operations of
telehealth networks everywhere. Traditional copper-based networks required dedicated
connections so telehealth was often limited to a single site or suite within a health care facility.
With line installation, monthly line rentals, and long-distance charges associated with each
session, telecommunications costs often accounted for as much as 15% to 25% of total telehealth
costs (2). With advances in digital video compression, composite audio and video signals can now
be carried over typical IP network circuits either on a LAN (local area network) within a health
center, across a broader WAN (wide area network), or private network. With nearly ubiquitous
access, telehealth can be available on every physician's desktop, at the patient bedside, and
throughout every hospital and primary health care facilityproviding access wherever and
whenever healthcare services are delivered. Although there is a fixed cost associated with an IP
network, there is little or no additional cost associated with actual use. As a result, the actual per
session cost for telehealth declines with increasing utilization. In addition to long-term cost
savings, the convergence of voice, video, and data onto a single network will ultimately allow
telehealth to interface with other health information, including PACS and electronic health records.
Issues surrounding network quality, bandwidth requirements, and security continue to be refined
but telehealth over IP networks is becoming an attractive option for many programs.
Choosing the right technology for telehealth is complicated in the face of declining equipment and
telecommunications costs, inevitable capital depreciation, and rapid technical innovation. Clinical
users of the equipment must be involved in purchasing decisions, as clinical and operational
requirements will directly determine technical specifications; such involvement will also foster
acceptance by the users. In general, a telehealth program should purchase the highest
specification equipment available to meet user expectations at the lowest possible cost. Similarly,
decisions regarding telecommunications infrastructure should be based on file size, immediacy,
and volume of usage balanced by fiscal realities and whether or not funding or revenue streams
can offset the associated capital and operating costs. Telehealth programs must also plan for
equipment and network maintenance, support, and upgrading. Vendor relationships are crucial
to the success of telehealth programs. In addition to price and technical specifications,
appropriate service level agreements should be instituted to ensure high-quality, reliable, and
state-of-the art telehealth operations.
MEDICO-LEGAL AND REGULATORY ISSUES

The clinician's duty of care and clinical case responsibility in the telehealth setting adhere to the
same principles found in face-to-face encounters. The ethical and quality standards of practice
governing clinicians are not altered by the use of telehealth. Proper documentation, including
written consent when applicable, should be maintained for all telehealth encounters. The
advantage of telehealth is that it permits health care to be delivered anywhere, with no
recognition of borders; however, this inherent distinction from traditional in-person care also
raises new issues surrounding regulations and policy related to health care practice (7) (see also
Chapter 8).
Licensing of health professionals is typically a jurisdictional responsibility. If physicians and
patients are located in different provinces or states, it is important to determine whether the locus
of accountability will be the patient site or provider site. If the service is deemed to occur at the
patient site, the physician may be required to obtain appropriate licensing and credentialing as if
practicing in person at that site. A number of policy options to overcome these licensing issues
have been proposed including universal licensure, a special-purpose license for telehealth, and
mutual agreements, but there is still a significant need to harmonize standards to support crossjurisdictional telehealth activities (7).
The absence of policies regarding physician reimbursement for telehealth consultations has
historically been a significant barrier to the widespread adoption of telehealth. Early telehealth
initiatives were often pilot projects or clinical trials located at hospitals or universities, and
physician reimbursement was not a major concern as most physicians treated their participation
as a research initiative or were compensated through alternative payment arrangements (salary
or capitation). However, as more practitioners incorporate telehealth into routine practice,
compensation has become a pivotal issue. Although many insurance plans still require patients to
be seen in person by a physician in order to bill, others have developed specific fee codes for
telehealth services; however, these often have significant limitations related to geographic
location, specific facilities, number and types of services, and professional category. Canada's
publicly administered system is also inconsistent, with some provinces allowing direct
reimbursement for telehealth services and others not. There is slow progress toward expanding
telehealth reimbursement, but health care organizations should determine the policies in their
jurisdiction regarding payment prior to implementing telehealth services.
P.57
All physicians engaging in telehealth consultations should verify with their insurance providers
that telehealth is included within their malpractice insurance policies. To date, there has been
very little litigation associated with telehealth, but there are some specific issues to be
considered. Not all consultations are appropriate for telehealth. Providers must use their best
clinical judgment to determine whether services can be safely and effectively provided by
telehealth. In addition, a back-up process must be established to ensure that patients receive
appropriate and timely care in the event of technical failure. Practitioners may require specialized
training and expertise for telehealth and demonstrate acceptable technologic competence prior to
providing telehealth services. Specific clinical protocols and guidelines may be necessary to
ensure consistent, high-quality telehealth applications in some settings. At all times, telehealth
services must adhere to basic quality assurance and professional standards of care (7).
Privacy of personal information related to the use of ICTs in health has been a developing issue
over the last decade. Concerns over the use of technology to track everything from health care
services to spending habits have spurred the development of policies to regulate the protection of
individual privacy. Standards for maintaining the privacy of health information in a telehealth
context do not differ from those in a face-to-face encounter; however, the introduction of
technology adds privacy and security considerations (7). In addition to maintaining privacy
through more traditional measures such as a private physical environment and organizational
processes, delivery of telehealth requires attention to security of data during transmission and, in
some cases, storage. Ensuring security in an ever-changing technologic environment requires a
proactive and evolving approach (8). Ensuring confidentiality in a telehealth setting can be more
challenging given the potential risks for interception, potential for a permanent video record, and
additional people involved in each care session. This is compounded by the variety of equipment
and complexity of transmitting images between two settings (9). As telehealth moves from singleroom standalone applications to integration within direct patient care areas, such as the neonatal
intensive care unit (NICU), the complexity of ensuring privacy increases.
THE ECONOMICS AND EVALUATION OF TELEHEALTH

The body of scientific literature related to the application of telehealth has been growing steadily
over the last 30 years. A search for the keyword telemedicine on PubMed found only 24
references for the decade 1970 to 1979; this increased to 2,903 for 2000 to 2003. Several recent
literature reviews have examined the current state of research knowledge related to telehealth,
and most have determined that while telehealth shows promise, research and evaluation of
telehealth is still maturing with few scientifically conclusive studies completed to date. The
reviews completed to date have identified small numbers of studies which meet criteria for
inclusion ranging from 7 for those requiring a randomized controlled trial to 50 when inclusion
criteria were broadened to include any controlled design (10,11). The strongest studies examining
clinical impacts have demonstrated support for telehomecare initiatives, management of chronic
illness, psychiatry, dermatology, cardiology, teleradiology, transmission of digital images for
neurosurgery consultation, and echocardiographic image transmission (10,11,12,13,14).
Most telehealth studies include some element of economic analysis. Whitten and colleagues (3)
completed a systematic review of telemedicine cost effectiveness studies from 1966 to 2000 and
found 612 articles that included some economic analysis, most of which were recorded as smallscale or short-term evaluations with less than ideal economic analysis and limited transferability.
Only 55 included actual cost data, 24 of which met the requirements for a full review. Jennett and
associates (1) identify more than 40 telehealth studies that found economic indicators (many of
which are repeated from the Whitten study); however, again, most were rated as less than
optimal analyses. Economic indicators for telehealth include costs related to travel and travel time
for patients and providers; patient transport; equipment and telecommunications; shift of care
from larger centers; and recruitment and retention related to access to continuing education.
While the reviews done to date have demonstrated the feasibility of telehealth, telehealth
research has not yet included large-scale randomized studies (10). An additional challenge in the
evaluation of telehealth is whether face-to-face care is, in fact, the gold standard against which
any new service delivery method should be tested (10). As telehealth moves toward a more
integrated model, evaluation approaches must also use an integrated and systemic approach
rather than focusing on limited indicators to measure the impact of the technology. The potential
for increased access to healthcare resulting from telehealth applications may have long-term
systemic benefits not easily captured in a single study. A more recent review examining the
socioeconomic impact of a variety of telehealth applications found benefits to patients, clinicians,
and the health care system (1).
TELEHEALTH APPLICATIONS IN NEONATOLOGY

In 1970, regionalization of perinatal-neonatal care was advocated following the observation that
neonatal mortality was highest in those hospitals delivering small numbers of infants compared
with that of larger hospitals with neonatal referral units. Additionally since that time, the care of
the sick neonate has become increasingly technologically dependent, providing further
justification for regionalization of neonatal care. This has resulted in dislocation of families during
complicated pregnancies, limited
P.58
access to specialized resources for sick neonates, and difficulties in ensuring access to follow-up
resources.
In this same time period, technologic advances have resulted in the ability to deliver health care
effectively over distance. Those practicing neonatal-perinatal medicine need to harness this
technology, incorporating it into daily care, to provide consultation (both emergent and elective)
over distance, allow for televisitation by families, facilitate discharge from tertiary to secondary
levels of care, and provide follow-up of high-risk newborns in their home communities, ideally
within their home.
This section demonstrates how telehealth has been used effectively through the perinatalneonatal continuum, both for the delivery of care and for education of health care providers.
Antepartum
Telehealth has been used successfully for a number of antepartum applications including genetic
counseling and teleultrasound. In Queensland, Australia, a weekly connection, developed in 1997,
allows maternal fetal medicine specialists to provide direction to a sonographer at the patient site
while viewing the fetal ultrasound in real time. At the end of the consultation, the subspecialist
counsels the parents on diagnosis, prognosis, and management and prepares a report for the
referring physician. In addition to increasing patient access to the subspecialist, communication
between the two hospitals is improved and knowledge transfer between health care providers is
facilitated. A review of the program found only one missed fetal diagnosis out of 120 cases, and
patients were highly satisfied with the consultative process. The real-time interaction is
considered a key component of the success of this project (15,16).
Neonatal
Health care professionals caring for neonates know that even a low-risk pregnancy can result in a
high-risk situation for the neonate in 2% to 4% of deliveries. Additionally, newborn infants who
initially appear well can deteriorate rapidly, particularly from sepsis, respiratory problems, and
congenital heart disease. Timely access to specialists in newborn care can be a problem for
remote and isolated health facilities. In Manitoba, Canada, the provincial MBTelehealth Network
supports a telehealth link between a tertiary NICU and a general hospital 760 kilometers north.
Pediatricians at the remote site can consult the receiving NICU on an emergent basis for
assistance with management and stabilization. In this integrated system, both the remote and
receiving staff operate the equipment and only rely on technical support when problems are
encountered. The equipment at the remote end is positioned to allow the NICU staff to see the
newborn, cardiorespiratory monitor, and ventilator. The NICU health care professionals at the
receiving site take over camera control from the remote site to ensure that the remote staff
members are not distracted from the care of the newborn. Image resolution is higheven the
numbers on the side of the umbilical catheter can be read. The addition of the visual image of the
baby to the usual telephone verbal description allows for improved assessment and advice on
management (Fig. 5-1). The interaction between the neonatologist and referring physician
provides a conduit for ongoing education in neonatal care; further, the virtually present
neonatologist can talk the referring physician through procedures such as umbilical
catheterization, which the referring physician may not have recently performed, enhancing that
physician's confidence; and finally, but importantly, the parents are comforted by the ready
availability of a specialist for their newborn. This link has also allowed parents and extended
family to visit the neonate and participate in management decisions from the remote location in
situations when they are unable to travel to the NICU. While technology is not going to replace
the need for neonatal units, it does allow for improved access to care. In addition, if appropriate
education is provided to those smaller referring facilities, earlier retrotransfer of newborns should
occur, reducing family dislocation and costs. As telehealth networks expand and personnel realize
both its potential and ease of use, this application will surely become more widespread.
Telecardiology
Digital echocardiography has enabled the transmission of echocardiograms from remote sites to
pediatric cardiologists both by store-and-forward and real-time, synchronous transmission. Realtime transmission allows for continuous live contact between the cardiologist, sonographer, and
other health professionals, and also family members at the remote site. Although this technology
is used for all age groups, Finley (17) found that 51% of urgent examinations were for newborn
infants. In addition to diagnosis, the utilization of interactive videoconferencing can support case
conferencing and echocardiogram review between referring sites and remote surgical teams for
infants being referred to a cardiac surgical center. Families can participate in this process,
increasing their confidence in the care of their infant. With appropriate bandwidths, images can be
of clinical quality. Ideally the sonographer at the remote end should have pediatric cardiology
experience to ensure that more difficult diagnoses, such as total anomalous pulmonary venous
return and coarctation of the aorta, are not missed (17,18,19,20). Although in many situations
the neonate will still have to be transported to the tertiary center, with telecardiology, some
transports can be avoided or undertaken more electively and more appropriate management
decisions can be made.
Retinal Telephotoscreening
Retinopathy of prematurity (ROP) is a complication of surviving preterm low-birth-weight infants,
with
P.59
approximately 65% of infants <1,300g at birth and up to 80% of infants <1,000g developing
ROP. These infants require frequent retinal screening once the disease is diagnosed to prevent
blindness. The gold standard for screening is indirect ophthalmoscopy by a pediatric
ophthalmologist with expertise in ROP management. Such subspecialists are few in number and
with the geographical diversity of levels II and III NICUs, providing this service is challenging.
Telehealth technology, using a digital retinal camera equipped with an ROP lens, is currently being
investigated as an alternative method for retinal screening. The camera system provides an
immediate wide-angle view of 120 degrees and also produces a real-time image on a computer
monitor, which is then stored, uncompressed, on a digital videodisc (Figs. 5-2 and 5-3). Two
recent studies concluded that the sensitivity was insufficient for recommendation as a screening
tool for ROP (21,22). However Schwartz and associates (23) suggested that this telemedical
strategy might be highly accurate when assessing whether a child's eye required the urgent
attention of a physician capable of evaluating and managing threshold ROP. Ells (24)
subsequently adopted a pragmatic approach; rather than using this technology to differentiate
between stages 1 and 2 ROP, she focused on whether this technology could identify those eyes
that require treatment. In this approach, digital photography had a sensitivity of 100% and a
specificity of 96% in detecting referral-warranted ROP. As the technology improves and more
people become trained in its use, the use of digital retinal photography will become a useful
adjunct to hospitals providing level II care; and growing premature infants at risk of ROP will be
able to return to their home community rather than remaining in a center where there is access to
a trained pediatric ophthalmologist.
Figure 5-1 Videoconferencing between the neonatologist in the NICU and remote health care
professionals following the delivery of this preterm infant.
Televisitation
Telehealth can promote a connection to the neonate's home setting both before and after
neonatal discharge. Interactive videoconferencing can provide a link between the mother and her
family. Gray and associates (25) have used a web-based solution designed to reduce the costs of
care and at the same time provide enhanced medical, informational, and emotional support to
families of very
P.60
low-birth-weight infants in the NICU. Via the Internet, families can link to the NICU at any time to
obtain information about their infant, as well as obtaining educational information and information
on experiences of other families. Also incorporated, and perhaps the most exciting aspect for
families, is the interactive videoconferencing that can occur. Parents can see their infant and
receive information and support from staff at times when they cannot visit. In a small,
randomized study of this technology, hospital stay was shorter and families experienced greater
satisfaction with the care received. The system requires the installation of an easy-to-use
computer in the home. Given that access to the Internet has been adopted more rapidly than any
other technologic advance in history, the use of this technology as a means of delivering care can
only expand.
Figure 5-2 Retinal telephotoscreening taking place in the NICU using the RetCamII.
Telehomecare
Telehomecare is a growing application with considerable potential to support neonatology.
Telehomecare applications link patients from their home environment to hospital or communitybased providers with ongoing monitoring of vital signs, and when indicated, videoconferencing
contact on a scheduled or urgent basis. Reviews of these applications in the non-neonatal
population have demonstrated improved control of chronic conditions such as diabetes and
chronic heart failure, along with high levels of patient satisfaction (1,12). Data such as blood
glucose, blood pressure, spirometry, or weight can be collected in the home by digital equipment
and transmitted to a health care provider at a distance. Some applications include automated
reminders about when to monitor and about opportunities to respond to specific questions to
enhance the remote assessment and improve self-management. Access to assessment data on an
ongoing basis allows providers to readily identify trends requiring intervention, and patients are
motivated and reassured knowing their results are monitored routinely. Within the neonatal
setting, telehomecare can be a support to parents following discharge, a time when families may
feel inadequate and may be reluctant or have difficulty accessing care. Telehomecare may include
regular vital sign monitoring as well as videoconferencing to assess color or respiratory status in
conditions such as congenital heart disease or bronchopulmonary dysplasia.
Figure 5-3 This image of an infant with ROP demonstrates the wide-field imaging capability of
the RetCamII. (See color plate)
P.61
Postdischarge telehealth applications can include improved access to a wide range of disciplines
and support systems, including follow-up developmental assessments, feeding assessments,
speech and language therapy, genetic counseling, and many others. For neonates with ongoing
health issues, telehealth can also provide access to patient education and support for families in
rural locations, who may not have regular access to other families supporting a newborn with a
chronic condition. Telehealth can provide key support for neonates as they transition out of the
institutional setting, and it may result in earlier discharge and improved patient outcomes.
Continuing Education
The education of healthcare professionals involved in newborn care is critical to maximizing
telehealth's potential to improve neonatal care. Continuing education is challenging, particularly
for professionals working at a site removed from a tertiary center. It is now recommended that all
hospitals delivering newborns should have an individual certified in the Neonatal Resuscitation
Program present at the delivery. In addition to initial certification, recertification is required every
two years. Cronin and associates (26) have demonstrated that instruction via interactive
videoconferencing, including the testing of practical skills, can be provided satisfactorily. Similarly,
Loewen and associates (27) have demonstrated that the neonatal stabilization program, S.T.A.B.L.
E. (sugar, temperature, artificial breathing, blood pressure, lab work, and emotional support),
can be delivered effectively. In a study of 56 health professionals randomized to receive the
program either in person or by videoconferencing, both groups, with similar pretest scores, had a
similar, but significant increase in their posttest scores.
In Canada, trans-Canada rounds have been established to allow neonatal fellows from across the
country to interact and share information on unusual cases. These sessions are highly interactive
and in addition to improving knowledge, have resulted in an element of friendly competitiveness.
PRACTICAL STEPS IN GETTING STARTED

As telehealth moves away from pilot projects to operational telehealth networks, there have been
many lessons learned and some consensus on the critical factors for success. This background of
experience has provided a framework for developing new telehealth services or programs.
Needs Assessment
Experience has proven that time and resources spent on a needs assessment will result in the
ability to design systems that truly meet the user requirements. The needs assessment process
includes obtaining information and ideas from many groups involved in the delivery of health
services to determine the goals, objectives, and priorities for telehealth. The clinical needs
identified will drive the goals of the telehealth program and system design. Engaging clinical staff
early and promoting buy-in will ensure that clinical needs remain the primary focus of telehealth
programs and will begin the change management process necessary to integrate telehealth into
clinical workflow. A well-structured telehealth needs assessment will not only define the clinical
direction of a potential telehealth application or network, but will identify strategies to resolve
issues that could potentially add long delays and increased costs to telehealth implementation.
Readiness Assessment
Prior to telehealth investment, there is a clear need to determine the telehealth readiness of
communities and organizations to reduce the risk of failure and losses in time, money, and effort.
Although there is a fundamental requirement for sufficient bandwidth to support telehealth, nontechnical organizational factors are equally important to successful telehealth implementation. The
perception of need and the ability and willingness of users to adapt to the changes associated with
the introduction of telehealth can have enormous impact on adoption and utilization. Telehealth
success depends upon selecting communities, organizations, and programs that are aware of
telehealth and its benefits, have a genuine need for and commitment to telehealth, and can
provide or acquire the capacity to support and resource telehealth beyond implementation.
Telehealth Business Plan

To successfully deploy and evolve a telehealth program, it is important to develop a telehealth
business plan. The business plan serves as a guide for the organization and as a communications
tool for external stakeholders and funders. This plan should be generated in collaboration with a
range of staff and be revisited frequently in response to changes in the environment. The plan
should fully describe the telehealth opportunity, products, and services to be provided;
organizational structure; operational and
P.62
technical requirements; implementation plan; marketing and business development strategy; and
financial analysis. There is increased pressure for telehealth programs to become viable
commercial enterprises. Even in publicly funded health systems, telehealth is expected to show a
return on investment realized through cost savings and improved health outcomes. The goal of
the financial analysis is to define potential funding/revenue sources, expenses (capital and
operating), and cost recovery, as well as to identify and mitigate financial risk. Projections should
contain best-case, expected-case, and worst-case scenarios based upon a mix of alternative
assumptions related to growth and use of the telehealth network, cost of operations, cost
recovery, and sources of funding/revenue. Careful attention to longer-term sustainability is critical
from the outset. Guidelines for developing a telehealth business plan can be found in several of
the listed references (2,28).
Program Management
Successful telehealth initiatives are built upon a robust operational infrastructure that ensures
efficient delivery of telehealth services on a daily basis. In addition to the obvious need for
technical training and support for users, there are many other functions required for effective
telehealth program management including scheduling, policy and standards development,
workflow and process design, marketing and communications, research and evaluation, and
financial management. Strategic clinical leadership is also essential to ensure buy-in of key clinical
providers, validate clinical telehealth applications, and develop clinical policies and risk
management strategies. Understanding these functions and the related skills and competencies
will allow a telehealth program to implement an appropriate organizational structure and human
resource plan to support operations. The optimal organizational model will also be determined by
the degree of telehealth integration into existing organizational structures and functions.
Evaluation and Quality Assurance Plan

Evaluation and quality assurance are vital to ensuring quality telehealth activities. To date,
telehealth evaluation has focused on technical accuracy and reliability, diagnostic quality and
effectiveness, impact on clinical management, user satisfaction, and impact on service or program
costs (9,10). There is still great need for standardization of evaluation tools, measures and key
indicators to assess the impact of telehealth on cost, quality, and accessibility of care. An effective
evaluation plan will provide evidence to funding agencies, health care administrators, providers,
and patients and will significantly contribute to the strategic development and sustainability of a
telehealth program.
Healthcare organizations, including telehealth programs, are adopting a focus on quality
improvement in response to increased demands for accountability, heightened consumer
expectations, resource limitations, and fundamental changes in health service delivery models
(29). Management tools such as the balanced scorecard are increasingly being adapted to the
health care setting to guide ongoing evaluation and quality improvement initiatives and can be
tailored to include telehealth and other ICTs (30,31,32,33). Regardless of the approach chosen,
telehealth services should be continually monitored through a quality assurance process that
allows rapid and responsive improvements to performance and the maintenance of quality
practices. When possible, both evaluation and quality assurance processes should be aligned with
broader program or organizational initiatives.
IMPLICATIONS FOR THE FUTURE

The introduction of advanced information and communications technology is inherently changing
the way that medicine is practiced. With the introduction of health information systems, clinical
decision support tools, and online learning, medical professionals are being provided the tools to
support high-quality, evidence-based practice. Telehealth and the electronic health record allow
access to health care services and information across sites, providers, and the continuum of care,
removing the traditional organizational boundaries and building a foundation for integrated service
delivery. These technologies are transforming the way that health care professionals and patients
interact with the health care system. IP-based technologies will provide a convergence of data
with images, video, and voice in a media-rich environment. Wireless network solutions will enable
mobile access to the health care system and these supporting technologies.
Telehealth is about the relationships and processes that allow health care services to be delivered
virtuallyit is not just about technology. Limits to telehealth expansion have less to do with
access to bandwidth and equipment costs than the human factors. Telehealth initiatives must
have the necessary time and resources committed to address not only the technical solution but
also the process redesign, and change management necessary to encourage more widespread
adoption of telehealth. Unless those involved with telehealth programs develop their skill and
expertise in these areas, the programs will run the risk of limited user acceptance and poor
network use. With the relatively high cost of implementation and operations, low utilization
threatens the sustainability of any telehealth program.
Neonatology has been accused of focusing on the high tech. Telehealth is also viewed by some
as yet another application of technology. Ultimately the fundamental purpose of both remains
grounded in patient care and achieving the best possible outcomes for patients. Telehealth
improves access to health care, which means access to specialist consultation for the neonate;
improves access to education of health care professionals in the care of the neonate; and
facilitates family involvement with
P.63
their newborn infant over distance. Telehealth applications in neonatology can range from lowtech home care applications to high-end intensive care and wireless transport communications.
For both, the ultimate goal of improved patient care remains. To achieve this goal, regardless of
the technical solution chosen, the need for building user acceptance and confidence, as well as
simplifying the telehealth process to support integration into the workplace, is essential.
REFERENCES
1. Jennett PA, Scott R, Hailey D, et al. Socio-economic impact of telehealth: evidence now for
health care in the future. Volume one: state of the science report. Calgary, Alberta: Health
Telematics Unit, University of Calgary, 2003.
2. Darkins AW, Cary MA. Telemedicine and telehealth: principles, policies, performance and
pitfalls. New York: Springer Publishing Co., 2000.
3. Whitten PS, Mair FS, Haycox A, et al. Systematic review of cost-effectiveness studies of
telemedicine interventions. BMJ 2002;324:14341437.
4. Brown N. A brief history of telemedicine. Updated June 03, 2004. Available at http://tie.
telemed.org/articles/tmhistory.asp?www=t1&tree=telemed101/understand/. Accessed
September 29, 2004.
5. Elford R. Telemedicine activities at Memorial University 19751997. November 28, 2002.

Available at http://www.med.mun.ca/telemed/telehist/telemulti.htm. Accessed September 29,
2004.
6. TIE (telemedicine information exchange). Available at http://www.tie.telemed.org/. Accessed

September 29, 2004.
7. Hasham S, Akalu R, Rossos PG. Medico-legal implications of telehealth in Canada. Telehealth
Law 2003;4:920.
8. Blum JD. Telemedicine poses new challenges for the law. Hlth Law Canada 1999;20:115126.
9. National initiative for telehealth (NIFTE) framework of guidelines. Ottawa, Ontario: NIFTE;
2003.
10. Currell R, Urquhart C, Wainwright P, et al. Telemedicine versus face to face patient care:
effects on professional practice and health care outcomes. Cochrane Database Syst Rev
2000;2: CD002098.
11. Roine R, Ohinmaa A, Hailey D. Assessing telemedicine: a systematic review of the

literature. CMAJ 2001;165:765771.
12. Hersh WR, Helfand M, Wallace J, et al. Clinical outcomes resulting from telemedicine
interventions: a systematic review. BMC Med Inform Decis Mak 2001;1:5.
13. Ohinmaa A, Hailey D, Roine R. The assessment of telemedicine: general principals and a
systematic review. Alberta, Canada: Finish Office for Health Care Technology Assessment and
Alberta Heritage Foundation for Medical Research, 1999.
14. Hersch W, Helfand M, Wallace J, et al. A systematic review of the efficacy of telemedicine
for making diagnostic and management decisions. J Telemed Telecare 2002;8:197209.
15. Chan FY, Soong B, Watson D, et al. Realtime fetal ultrasound by telemedicine in
Queensland. A successful venture? J Telemed Telecare 2001;7 Suppl 2:711.
16. Soong B, Chan FY, Bloomfield S, et al. The fetal tele-ultrasound project in Queensland. Aust
Health Rev 2002;25:6773.
17. Finley JP, Sharratt GP, Nanton MA, et al. Paediatric echocardiography by telemedicinenine
years' experience. J Telemed Telecare 1997;3:200204.
18. Casey FA. Telemedicine in paediatric cardiology. Arch Dis Child 1999;80:497499.
19. Sable C. Digital echocardiography and telemedicine applications in pediatric cardiology.

Pediatr Cardiol 2002;23:358369.
20. Widmer S, Ghisla R, Ramelli GP, et al. Tele-echocardiography in paediatrics. Eur J Pediatr
2003;162:271275.
21. Roth DB, Morales D, Feuer WJ, et al. Screening for retinopathy of prematurity employing
the retcam 120: sensitivity and specificity. Arch Ophthalmol 2001;119:268272.
22. Yen KG, Hess D, Burke B, et al. Telephotoscreening to detect retinopathy of prematurity:
preliminary study of the optimum time to employ digital fundus camera imaging to detect ROP.
J AAPOS 2002;6:6470.
23. Schwartz SD, Harrison SA, Ferrone PJ, et al. Telemedical evaluation and management of
retinopathy of prematurity using a fiberoptic digital fundus camera. Ophthalmology 2000;107:
2528.
24. Ells AL, Holmes JM, Astle WF, et al. Telemedicine approach to screening for severe
retinopathy of prematurity: a pilot study. Ophthalmology 2003;110:21132117.
25. Gray JE, Safran C, Davis RB, et al. Baby CareLink: using the internet and telemedicine to
improve care for high-risk infants. Pediatrics 2000;106:13181324.
26. Cronin C, Cheang S, Hlynka D, et al. Videoconferencing can be used to assess neonatal
resuscitation skills. Med Educ 2001; 35:10131023.
27. Loewen L, Seshia MM, Fraser Askin D, et al. Effective delivery of neonatal stabilization
education using videoconferencing in Manitoba. J Telemed Telecare 2003;9:334338.
28. Office for the advancement of telehealth. Available at http://telehealth.hrsa.gov/welcome.

htm. Accessed September 29, 2004.
29. Harrigan M. Quest for quality in Canadian health care: continuous quality improvement. 2nd
ed. Ottawa, Ontario: Minister of Public Works and Government Services Canada, 2000.
30. Baker GR, Pink GH. A balanced scorecard for Canadian hospitals. Healthc Manage Forum
1995;8:721.
31. Kaplan RS, Norton DP. The balanced scorecard: translating strategy into action. Boston,
Mass: Harvard Business School Press, 1996.
32. Kaplan RS, Norton DP. The strategy-focused organization: how balanced scorecard
companies thrive in the new business environment. Boston, Mass: Harvard Business School
Press, 2001.
33. Castaneda-Mendez K, Mangan K, Lavery AM. The role and application of the balanced
scorecard in healthcare quality management. J Healthcare Qual 1998;20:1013.
Chapter 6
Newborn Intensive Care Unit Design:
Scientific and Practical Considerations
Robert D. White
Gilbert I. Martin
Judith Smith
Stanley N. Graven
DEVELOPING A MISSION STATEMENT

The first step in a successful newborn intensive care unit (NICU) design project is
the development of a shared vision among NICU, hospital, and community
leaders. The construction of a new unit or redesign of an existing facility should
successfully address key problems in the region's current system of neonatal care.
For an existing, stable unit wishing to modernize its facility, these problems may
be primarily internalthe need for more space, current technology, and family
support. For an entirely new facility, these issues will be predominantly external
who are the existing providers and what niche will this new unit most
appropriately fill? A careful appraisal of medical care trends anticipated within the
community over the next 20 years, especially with regard to obstetric and
intensive care services, should precede any decision on a mission statement
intended to guide NICU construction and operation. Whether a hospital is the only
provider of NICU services in its area or one of several, it is crucial that it
understand its niche and the responsibilities inherent to that role, so that
appropriate physical facilities can be constructed to fulfill its mission.
A mission statement that might provide a framework for many NICUs is as follows:
The Newborn Intensive Care Unit will provide a state-of-the-art, integrated, familycentered approach to neonatal care. This interdisciplinary program will strive to
give patients, families, and payers high quality and compassionate care through
the provision of:
Care that honors the racial, ethnic, cultural, religious, and socioeconomic
diversity of family and staff
Education, information, and emotional support
Access to the most current effective therapies
Integrated treatment plans emphasizing coordination throughout the

continuum of care
Encouragement of family support and involvement
When consensus is reached on a mission statement, specific goals and objectives

can be defined that apply this mission to the local realities of demographics, care
practices, and competition. Defining these goals will be the first step toward
decisions on bed capacity, types of equipment needed, and changes in care
practices. All of these can then be explored in depth by the teams described
below. The goals and objectives should be measurable (e.g., survival and
morbidity rates compared to regional and national standards, staff experience and
turnover, parental satisfaction ratings, cost per patient day) and realistic, so that
the overall value of the project can be ascertained at the time cost projections are
available, and on an ongoing basis after construction is complete.
CREATING THE TEAMS

Once a decision has been made to proceed with the development of either a new
or a renovated NICU, it is important to form specialized teams with well-defined
goals. In many hospitals, several persons will serve on two or more teams. Each
team should have access to the several excellent reference materials available to
assist in the design of an NICU (1,2,3,4,5,6).
The Strategic Planning Team

The strategic planning team will continue to develop the vision and goals that led
to the decision to pursue new
P.65
construction. It should include, at a minimum, an administrator, a neonatologist,
and a nursing director. This group will be responsible for reviewing utilization and
demographic information (available from state and local planning and health
agencies, the insurance industry, and the Census Bureau) in order to define the
NICU's service area and appropriate number of beds. Some general guidelines in
this area include estimates that 5% to 12% of all newborns will need intensive
care at the time of birth, and 1% to 2% will need ventilator care. Total NICU days
for a defined region will approximate 1.25 patient-days per live birthif a region
has 10,000 live births per year, it will generate approximately 13,000 NICU
patient days, or an average census of 35 babies, with an average length of stay of
20 to 25 days. These numbers will be influenced considerably by the level of care
provided by a particular NICU, referral and back transport patterns, admission and
discharge criteria, and competing NICUs serving the same region. Because of
fluctuations in census, a unit should have sufficient bed positions to care for 40%
to 50% more babies than the average census calculations, so a region with
10,000 deliveries per year will need 50 to 55 NICU beds. Many areas currently
exceed this number by a sizable proportion because of serving a particularly highrisk population, or because of inefficiencies inherent in multiple NICUs serving a
single region.
The strategic planning team also should make some basic calculations regarding
staffing patterns, if this is to be a new service for the hospital. Depending on
patient mix, overall staffing patterns may require 4 to 6 nurses and 2 support
staff (inclusive of nursing administration, respiratory therapy, developmental
therapy, social work, ward clerk, and housekeeping staff) for every 10 babies. A
unit with an average census of 20 babies, for example, would require 8 to 12
nurses on each shift, or 40 to 65 full-time equivalents (FTEs), with an additional
20 to 25 FTEs for support staff. One neonatologist is needed for each 6 to 8 babies
(average census); this is one area, among several, where units with an average
census of less than 20 infants encounter certain inefficiencies because of their
size. On the other hand, large units (more than 40 beds or so) can be difficult to
design and manage efficiently because of the space requirements. Some hospitals
have addressed this issue by dividing their NICU into intensive care and
continuing care, or step-down areas. Some units will also need to plan staffing for
transport team or nurse practitioner coverage. Six to 7 full-time positions are
required for 24-hour coverage of each position for these services. A transport
team with 2 nurses and a respiratory therapist, for example, would require at
least 12 nurses and 6 therapists to provide full-time, in-house coverage.
Next, the strategic planning team also will need to assess the impact of the new
or renovated NICU on other hospital departments, especially obstetric,
maintenance, and supply services. When these issues are clear, the team will
proceed to interview and engage an architect. The timing of this step is important
the architectural firm should be involved before any design decisions are made,
but after the strategic planning team has articulated a clear set of goals for the
process. Several architectural firms can be asked to make formal presentations of
their general concepts for the project, as well as present examples of other
projects that they have designed. The architectural firm chosen should have
complete engineering and interior design specialists on staff, as well as an
equipment representative and a neonatal nurse planner. The entire architectural
team should be familiar with the latest trends in NICU design and the scientific
principles behind the design process. Once chosen, the architectural group and the
strategic planning team can then develop a timetable for planning and
construction of the new facility.
The Financial Planning Team

This group is composed of the hospital's chief financial and operational officers,
nursing management, and any other individuals representing areas of the hospital
whose budgets will be significantly affected by the new construction. Consultants,
such as those helpful in the equipment selection process (see Equipment Selection
section), also may be asked to join the team. The one-time costs for the project
will include architectural fees, the physical plant, and new equipment. Ongoing
costs may include increased staffing for direct patient care, housekeeping, and
maintenance functions. These costs should be clearly identified early in the
planning sequence, with the assistance of the care practices team (see The Care
Practices Team section). This team should continue to meet intermittently
throughout the planning process, because some unanticipated expenses often
surface as the design takes shape.
The Care Practices Team

This is the largest group, and it represents caregivers in the NICU. At a minimum,
the following disciplines should participate: neonatology, nursing management
and staff, respiratory therapy, social work, pharmacy, laboratory, radiology,
infection control, nutrition, housekeeping, and parents. Depending on local care
practices, surgery, obstetrics, and clinical research may be appropriate additions
to this list. As each individual provides information on his or her own specialty to
enhance care for babies, families, and staff, the team should be encouraged to
look beyond existing practices and view the new construction as an opportunity to
change the way things are done. In particular, most NICUs will need to reevaluate
their care practices with respect to parental access and developmentally
supportive care, which may have been severely restricted under their previous
NICU design. They should consider environmental design features that will provide
the most optimal living conditions for babies, working conditions for staff, and
integration of families, and they should begin to address some of the conflicts
inherent in these ideals. As a team, this group can then identify general
P.66
space, equipment, staffing, and maintenance and cleaning requirements, which
can be reported to the financial and design teams for determination of feasibility.
The Design Team

When the strategic planning, financial planning, and care practice teams have
agreed on the general scope of the project, including goals, approximate size and
cost, major changes in care practices, and an architectural firm, the design team
can begin its work. Members of this team should include the architectural group,
the hospital's project manager, neonatology, nursing, respiratory therapy,
parents, and the equipment consultant. After creating a functional plan that
addresses all the needs identified by the care practices team, this group should
visit new or updated NICUs to get additional ideas, learn how to avoid pitfalls, and
solicit suggestions on their functional plan. This team then will meet regularly over
several months as blueprints are developed to reconcile all these concepts into a
working unit. Detailed minutes should be kept and reviewed at the beginning of
each meeting.
SITE VISITS
Touring other NICUs can be a valuable investment for many reasons, whether
planning for a renovation or new construction. Visits provide a firsthand look at
design features that are and are not working for others. The benefits can include a
better plan, fewer change orders, and a smoother transition because the planning
team gained new ideas and a clearer understanding of the pros and cons of the
proposed changes.
Careful planning will assure that site visits are productive. Start by identifying the
objectives of the tours, areas of greatest interest, potential units to visit, and the
best timing. The typical purpose of a site visit during the early stages of planning
is to trigger ideas. Many NICU teams find it useful to tour other units during the
stage when they have identified their vision, defined their care practices, and
prepared a preliminary functional program. Others stagger their tours and see
different units during various stages of planning. Waiting until the preparation of
construction documents is underway is not usually the best timing for initial site
visits, unless the objective is to clarify operational issues related to the selected
design approach.
Two other essential components to be addressed are selecting the members of the
tour group and funding the travel. Selecting tour team members should tie back
to the objectives. Most units request a budget for site visits as part of the project
funding. Some fortunate units have donors who help with the travel budget, and
even include interested donors on the tour team, when appropriate. If a travel
budget is not feasible, the next best strategy is virtual tours. In many cases, the
NICUs or their architects have photos in hard copy, on Web sites, videos or other
material that could be viewed, and several of these are also available on a single
Web site (6). A phone interview or written survey would be helpful to provide
descriptive information for a virtual tour.
Tips for Planning a Site Visit

Performing the tasks below will enable the planning team to get maximum benefit
from a site visit:
Interview key people by telephone or in person before touring their unit.
Obtain permission to photograph, videotape, or receive audiovisuals already

available from public relations, architects, and other sources. If assigning
photography to a team member, assure that he or she has the proper
equipment and competence.
Design a questionnaire that will be used consistently as a guide for all visits.
Assign a person or people to key topics to assure all pertinent questions are
answered before, during, or after the tour. Categorize questions by area of
interest, and create the expectation that the responsible person will report
on the findings.
Determine how the information from the tour will be used and by whom.
After the visit and in a timely manner, discuss and record general
impressions of people who toured the facility.
Sample Questionnaire
A questionnaire should be developed to ensure that all needed information in
relation to a site visit is captured and recorded. Segments of the questionnaire
could be contact information, general statistics, and specific questions.
Contact Information
The contact information should include:
Date of Visit
Organization Name/Location
Contact Names/Titles
Phone Numbers/E-mail Addresses/Web Site
Who attended from your organization
General Statistics
General statistics should include:
Annual NICU volume (and birth volume, if relevant)
Number of beds by level of care or other category
Annual number of infant transports in and out
Average length of stay (by level, if appropriate)
Economic and cultural characteristics of population served
Acuity of patients served (and information on how acuity is defined)
Payer mix
Date opened or renovated
P.67
Examples of Questions
The questions to ask at the site being visited should include, but not be limited to:
Describe the changes that resulted in the current services and facilities and
the reasons driving the changes.
Were the objectives met? (How is success measured?)
What types of NICU health professionals practice at this hospital (number of

neonatal nurse practitioners (NNPs), neonatologists, residents, etc.)?
What are the key staffing issues that were or need to be resolved?
What type of staffing education programss, orientation, budget, and amount

of time were required to make changes to the current model?
How is each type of patient care room or area used?
Where and how are most procedures handled?
Describe communication systems among staff, between staff and families

and others.
Describe special features of unit.
What was the biggest challenge related to this project?
What would be done differently based on the wisdom learned from the
experience?
What were the benchmarks or other resources that were most helpful
during the planning process?
Other Potential Topics for Questionnaire

Other topics that may be included on the site visit questionnaire are:
Family input
Visitation practices
Family space
Care practices
Admission process
Maintenance
Monitoring systems
Lighting
Sound control
Interior design
Storage
Communication Systems Security
Flow paths
Cleaning
Head walls
Staffing
Fundraising
Total project costs
Acceptance by staff/administration
Response by community/competitors
Ceiling/floor/wall/cabinet finishes patterns
Coordination with labor and delivery
One of the final planning steps is to determine how the information gained from
the site visit will be shared with the stakeholders who could not participate directly
in the tour. Good documentation of the visits and effective presentation of the
information will provide maximum benefit from the activities.
SPECIFIC DESIGN ISSUES

Floor Space Requirements
NICUs have historically been undersized, with insufficient room for storage,
parental access, and offices, even when accurate estimates are made for the
number of babies to be cared for. Change has been incremental, as
administrators, architects, and NICU staff gradually expand their horizons to
accommodate new practices, particularly increased parental access.
Given this historical perspective and extrapolating from floor space requirements
in adult and pediatric intensive care units (ICUs), it is likely that an NICU design
should allow 600 gross square feet per bed to meet the needs of the babies, staff,
and families. This figure includes office, storage, and support space as well as the
patient care areas; gross square footage estimates also include space required for
hallways, ductwork, restrooms, and the like. This figure is an average; teaching
hospitals with active clinical research programs, for example, might require more
than 600 gross square feet per bed, whereas a community NICU might require
somewhat less. Any design that provides less than 500 gross square feet per bed
will inevitably demand compromises and limit flexibility for the future, when
parental access desires and lengths of stay (because of increasing survival of
extremely low-birth-weight infants) may be even greater than they are today.
Previous estimates of floor space requirements often recommended square
footage per bed position (e.g., 120 to 150 square feet per intensive care bed
position), and most state codes still use this convention. If this method is used, an
additional 30 square feet of storage space must be planned outside of the patient
care area, as well as 24 cubic feet of cabinet storage at each bedside, at a
minimum.
Location Within the Hospital

The NICU should be immediately adjacent to the high-risk delivery rooms
whenever possible. There are few, if any, NICUs that are more than a few
hundred feet from a high-risk delivery service that do not consider their location
as the source of certain risks to optimal care. In some circumstances, especially
very large hospital complexes or freestanding children's hospitals, this ideal may
be impossible to achieve. In the former situation, a resuscitation area within the
delivery room complex or in each delivery suite, always valuable, becomes crucial;
in the latter case, all infants are accepted as transfers, and proximity to the
ambulance entrance and heliport becomes particularly significant. When the NICU
is on a different floor from the delivery suites, as often occurs in larger hospitals,
controlled elevator access between these two areas is essential.
Many NICUs, particularly in small or medium-sized hospitals, share staff and
responsibilities with the well baby nursery or pediatric ICU. When these areas are
P.68
contiguous, much of the support space (e.g., family lounge, staff lockers,
equipment storage) can be shared, and the opportunities for staff to assist one
another are enhanced.
Traffic patterns for infants who leave the NICU for procedures should be identified
and private hallways created wherever possible, so that ill infants and their
attendants do not need to use public areas.
Security Considerations
As family access to the NICU has become more prevalent, so have security
considerations. Most NICUs now permit continuous access to families, so traffic in
and out of the NICU is commonplace. Certain design features are necessary in this
setting to protect babies and staff. First, the NICU should be designed with only a
single public entrance. Other entrances and exits (e.g., to the delivery suite, the
clean utility room, and the staff lockers) may be desirable, but these should be
less apparent to the public (except as fire exits) and either within constant
observation of the staff or equipped with alarms.
The public entrance to the NICU should be designed with the security of all major
parties (babies, families, and staff) in mind. Families and staff need a secure
locker area to store personal belongings. This area should be large enough to
allow more than one family to enter or leave at a time and should be under
constant observation. Electronic detection systems increasingly are being used in
newborn nursery areas to prevent kidnapping; these are uncommon in NICU
settings, but worthy of consideration if constant visual monitoring of all exits from
the NICU is not feasible.
Fire exits should be carefully planned in the initial design, and clearly marked, as
should the location of fire extinguishers. The fire marshal should be given the first
draft of the design documents so that any problems can be corrected early.
Reception Area
In most NICUs, the reception area is part of or near the public entrance. This area
also may encompass the family lounge and the ward clerk's workspace. The size
and layout of this space is highly dependent on the size of the NICU and its
culture, or typical practices.
The family lounge should always be large enough to accommodate at least two
families (6-10 people) comfortably, with seating that is comfortable, but not
conducive to overnight sleeping. Lockers for valuables, a coffee pot, a television
set, reading material (including informational and educational literature for
families of NICU babies), and a toy box for children are items found in many such
areas; public restrooms and telephones should be nearby.
Families and the public should be able to immediately reach an NICU staff
member from the reception area. In some larger units, the contact person might
be a full-time ward clerk; in others, it might be a hot line or direct visual contact
into the patient care area. In any case, this function should be carefully thought
out so that those who should be encouraged to enter (i.e., families) find
themselves feeling welcomed, while others (e.g., curious individuals wandering
through the hospital) are discouraged.
Direct Patient Care

As might be expected, the greatest design effort is spent creating the patient care
area. The first and major decision to be made is the extent to which individualized
environments are desirable. The goal to provide each baby with a space that is
not impinged upon by noise, bright lights, and commotion from an adjacent bed
position, and that has enough space to accommodate staff, family, and bedside
equipment and storage seems intuitive and is certainly the norm for adult ICUs;
but this has not historically been the pattern for NICUs, for several reasons.
NICUs in the 1970s and 1980s were routinely designed as large, open wards.
These permitted care of the largest number of babies in a given space and allowed
direct visibility of each baby by the staff, and of staff members by one another.
Babies (and staff) were not thought to be adversely affected by the noise and
bright lighting that were the norm for such designs. New practices and equipment
(as long as they did not require much additional space) could be incorporated
easily into this design, and highly variable census patterns also could be
accommodated more easily than in the older premature nursery design of several
smaller rooms. Parents were allowed, or expected, to visit for only short periods,
so very little space at the bedside was allotted to them.
In recent years, many of these assumptions have changed. Direct, continuous
observation of most babies no longer is considered necessary, thanks to improved
monitoring systems (especially noninvasive oximetry and blood pressure
monitoring), and certainly is no longer practiced except for the sickest babies,
even in units where it is still physically possible. Babies clearly are affected by
excessive, noxious levels of light and noise, and their response to these
environmental hazards varies with their gestational age and medical condition.
Extensive adult studies in workplace settings also have established that staff
needs and responses can be quite different from those of babies (7). Perhaps,
most dramatically, family expectations have changed. Many families now want
continuous access to their babies and want to be able to stay at the bedside
without interfering with the ability of staff to care for their babyin any but the
most extraordinary circumstances.
Designing an NICU that takes these newer realities into consideration requires
establishing a local philosophy on individualized environments. In some units,
private or semiprivate rooms for at least some of the infants will be considered
desirable and feasible; in others, space, staffing, or philosophical considerations
will lead the design team in a different direction. Often, a combination of
pods (one, two, or four beds), with a larger room for recovering
P.69
neonates, has been the design of choice. In any NICU, however, the need for each
baby to be protected as much as possible from the noise, light, and activity
generated at other bed positions and the differing needs of babies from staff need
to be discussed and integrated into the design process at the outset (8).
Certain principles can be established for all direct patient care area plans,
regardless of whether a single large room, multiple smaller rooms, or private
rooms are chosen as the model. First, each bed position must have sufficient
space for families to stay for extended periods without interfering with staff
duties. Second, each bed position must have individualized lighting, data entry,
and communications systems. Third, traffic patterns must be well planned, with
sufficient aisle widths to accommodate diagnostic equipment and personnel;
generally, this requires aisles of at least 5 feet in width. Nursing functions should
be separated from the bedside whenever possible. For babies, the NICU is perhaps
ideally visualized as a bedroom for a sick infant; for staff, it is a workplace, where
both work-related and social communication occurs. These concepts are often in
conflict, yet the NICU design must accommodate both to the greatest extent
possible. Therefore, the design team should (again, within the local culture and
practices) separate those nursing functions that do not involve direct patient care
(e.g., most charting, giving reports, receiving and making most phone calls) from
the bedside to the greatest extent possible, especially if those activities require
different lighting levels or create noise that would disturb the baby. This is done
by providing adequate space for these functions away from the bedside, yet
sufficiently proximate to allow appropriate response to the unpredictable needs of
the baby. Finally, in any plan that uses modules or individual rooms, staffing
patterns should be considered. In general, enough babies should be clustered
together to justify staffing with at least three nurses. Designing smaller clusters
creates significant staffing issues during those times when at least one nurse must
leave the area.
After these considerations are adequately debated and decided, the general layout
of the patient care area can proceed. Each bed position can be identified, as well
as nursing work areas (charting, medication preparation, etc.), sinks, and traffic
flow patterns. At this point, an infectious disease specialist also should be
consulted, to assure that the design is conducive to good infection control
practices. This individual also will be interested in airflow considerations and in
floor, wall, and ceiling finishes (discussed below). Proximity to other important
support areas also will begin to take shape at this stage (specific design features
for each of these areas are discussed in the next section).
The design team then can spend considerable time planning the layout of a typical
bed position. Perhaps most important in this regard is the headwall, including
monitoring and communication systems. Construction of a full-size mockup of an
individual bed position, complete with headwall, monitors, and communications
systems is well worth the effort and expense to be sure the ergonomics of the
design have been adequately considered. Many members of the design team will
be unable to fully visualize the layout from a blueprint, and it would be rare that a
new NICU would resemble an existing design closely enough to bypass this step.
Most NICUs will want to include one or more special care areas for isolation,
treatment rooms, parent rooming in, and breast-feeding. In most units, some of
these functions can be combined, and many will choose to eliminate one or more
of these because of local practice patterns. The need for each should be carefully
considered in new designs, however, and if not served by a dedicated space, at
least clearly accommodated at some appropriate location in the NICU.
General Support Space

Clean utility, soiled utility, and general storage are functions that each require a
dedicated space; in some units, a satellite pharmacy and clinical laboratory are
also given designated areas. The layout of these rooms is straightforward; errors
in design usually are made by allowing too little space or placing these rooms in
an inconvenient position away from the direct patient care area. Ideally, these
areas also should be accessible from an outside corridor so that restocking can
take place without unnecessary traffic through the NICU, but this must be
designed in accordance with security precautions (discussed previously). Storage
areas should have a generous supply of electrical outlets and shelving so that
battery-powered devices can be recharged. Additional specific considerations are
outlined in each state's code documents and in several published guidelines
(1,2,3).
Staff Support Space

As noted previously, some charting space, especially for the nursing and
respiratory therapy staff, needs to be allocated within the direct patient care area.
Additional space, especially for physician and nurse practitioner charting and
discussion, should be provided adjacent to the patient care area, so that the noise
and activity generated there does not impinge on the babies' bedsides. The
communications systems (phone, computer terminal, printer) that link the NICU
with the hospital laboratory, pharmacy, and central supply generally will be
situated in this area as well, which will benefit from being centrally located.
The staff lounge and locker room should be adjacent to the NICU, with restroom
facilities integral or nearby. The lounge should be large enough to accommodate
at least one third of all NICU staff (nurses, therapists, physicians, and other
support staff) at one time and be accessible to the NICU by phone or intercom.
Several disciplines should have office space immediately adjacent to the NICU,
including social work, medical and nursing administration, and developmental and
respiratory therapy. When parent support or research staff
P.70
are actively involved in a unit's activities, they also will need nearby office space.
On-call rooms and a conference room should be situated within this complex, with
phone and computer links, including digital x-ray transmission, and restrooms
with showers.
Family Support Space

In addition to generous provision of space at the bedside and in the family lounge,
parents need space to stay overnight, to meet in private with staff to discuss their
baby or to grieve, and to breast-feed. Again, depending on the size of the NICU
and its local practices, some of these functions can be combined, but none can be
ignored.
Rooming in, even with critically ill patients, is provided (indeed, encouraged) in
most areas of a typical hospital, except for the NICU, where this practice usually is
limited to those infants close to discharge. We are unaware of any documented
factors, other than tradition, that justify the exclusion of parents from rooming in
with NICU patients from immediately after birth, however, and suggest that future
designs should anticipate that this practice will become more prevalent as families
come to request and expect this opportunity. To be sure, many families will not
choose or will be unable to room in with their ill newborn, so an ideal design need
not anticipate that all babies will need space for rooming in at all times, but it
should allow most parents who desire this privilege the opportunity to do so.
Experience in NICUs that permit unlimited rooming in indicates that up to 50% of
babies may have a family member staying overnight at a given time.
Breast-feeding of premature or ill infants is poorly accommodated in many
existing NICUs because of space and privacy considerations. Mothers should be
able to breast-feed babies at the bedside without compromising their privacy or
being asked to wait until the baby is stable enough to be transferred to an area
where these problems do not exist. Alternatives in this regard include private
rooms for those babies whose mothers plan to room in and/or breast-feed, or
moveable partitions. Breast-feeding rooms are also a good alternative for babies
who are healthy enough to be moved and can be utilized by mothers who need to
use a breast pump. These rooms should be designed with most of the capabilities
of a headwall (e.g., oxygen, suction, monitor, electrical outlets, and direct
communication to the NICU staff), yet be as much like home as possible.
Room for parent education and counseling should be available within the NICU
complex. This is particularly important when private discussions need to occur
with families of critically ill infants.
The location and content of signs are often overlooked when planning an NICU.
Some thought should be given to traffic patterns for families and the public from
the hospital entrance(s) and how signage will be used to direct them clearly to the
NICU. Information on signs should be phrased warmly, in a way that will make
families feel welcome, rather than sternly, in a way that could make them feel like
outsiders and intruders.
Lighting
Planning appropriate lighting for the NICU requires consideration of the disparate
needs of both the babies and staff. In general, babies need very little light, but
exposure to moderate levels of illumination during part of the day may help
establish circadian rhythmicity.
Staff need moderate levels of illumination at the bedside to evaluate babies and to
perform charting and manual tasks. At times, intense levels of illumination are
necessary to perform procedures and for phototherapy of hyperbilirubinemia. It is
doubtful that babies need natural lighting, but studies of adult office workers and
hospital patients document the benefit of windows for staff and families (9).
Consideration of these disparate needs should lead the design team to plan a
multilevel lighting scheme. At the bedside, a low level of indirect ambient lighting
(200 to 300 lux [20 to 30 foot-candles]) is desirable, so the baby is not exposed
to a continuous or direct bright light source. Task lighting, both for the bed
surface and the nursing work surface(s), should be highly focused (framed), and
rheostatically controlled, so that only the amount of light needed is provided, and
only to the specific location desired. This may require multiple light sources at the
bedside. The ideal task light would be adjustable to an infinite variety of positions,
yet be able to recess fully into the wall or ceiling. It would have a beam of
adjustable width and intensity and be free of shadowing effects.
At nursing work areas and traffic circulation areas elsewhere in the patient care
area, moderate lighting levels (300 to 1,000 lux) are suitable. The light sources
used should avoid glare, especially on work surfaces and computer terminals, and
should have multiple switches or rheostatic controls to allow flexibility in lighting

levels. A single master switch may be useful, however, if a darkened room is
desired for transillumination. All fixtures should have filters or shields that block
ultraviolet radiation and minimize the risk to babies and staff if a bulb should
shatter.
Provision of natural lighting is highly dependent on the NICU locale. Ideally, each
room where adults spend several hours at a time would have a window with an
attractive view, allowing the eye to escape and orient to external conditions. In
the northern hemisphere, windows that face north are optimal, because they will
not transmit glare or heat gain. Windows facing other directions should be
appropriately shaded, either with an external overhang or with internal louvers. If
the ability to completely darken the room is desired, internal (within a tripleglazed window unit) electrically operated louvers are essential. Because even wellinsulated external windows allow some heat gain or loss, these should be situated
several feet away from any baby's bed position.
P.71
Noise Abatement
The initial design issue in making an NICU as quiet as possible is to eliminate or
reduce as many sources of background noise as possible. Some NICUs are
situated in noisy communities, which requires extra insulation in the external walls
to minimize the impingement of outside sounds into the NICU. Airflow through the
heating and cooling ducts also can produce considerable background noise in an
NICU, but this can be reduced through appropriate sizing and baffling of the
ducts. These issues must be addressed in the design process, because it is
prohibitively expensive to correct a poor design after construction begins.
Traffic patterns also play a role in determining the level of noise to which babies
and staff are exposed. To the greatest extent possible, traffic flow should be
designed so that an echocardiogram, ultrasound, x-ray, or
electroencephalographic technician can get to each baby's bedside as directly as
possible, without wheeling the equipment past several other bed positions. As
noted previously, support areas should be designed so that restocking functions
can be accomplished without creating unnecessary bedside traffic.
Noise production also should be a prime consideration in design of the monitoring
and communication systems and in the selection of equipment. Whenever
possible, equipment should be selected with a noise criterion rating of less than
40. Based on available data for sound levels that do not interfere with newborn
sleep or adult conversation, overall background noise levels in the NICU should be
maintained below 55 decibels on the A-weighted slow response scale, with peak
levels not in excess of 70 dB (10).
Once all unnecessary sources of sound are minimized, the next design
consideration is the abatement of unavoidable sound such as voices, equipment
noise, and anything that might disturb a sleeping baby. Here, there is no
substitute for adequate space, and another cogent argument for individualized
environments becomes apparent. Increasing the distance between beds will
diminish noise transfer from one baby's bedside to another, as will higher ceilings,
especially those that are angled to reflect sound laterally rather than back to the
bedside. Obviously, floor, wall, and ceiling materials are crucial in this regard (see
the next section, Surface Finishes).
Finally, care practices should be evaluated as part of the design process to see
whether many sources of noise produced by the staff can be diminished or
eliminated. Radios, pagers, rounds, and reports are examples of care practices
that may create considerable noise at the bedside, and that can be modified or
eliminated.
Surface Finishes
In the past, selection of surface finishes was given little attention in NICU design,
which focused primarily on integration of the newest technology. The choice of
wall, ceiling, and floor finishes is important, however, for reasons of aesthetics,
noise abatement, and infection control.
Perhaps the most controversial design issue in this regard is the choice between
carpet and hard flooring. Hard flooring (usually a vinyl composite) is easily
cleaned, durable, and provides little resistance to wheeled equipment. Carpeting
provides noise abatement and may be more attractive and more comfortable to
those who are on their feet for several hours a day. The differences between these
two choices have started to blur in recent years, as carpet has become more
durable and cleanable, and hard flooring has become more resilient and sound
absorbent. It seems clear that vinyl or rubber flooring is the ideal for isolation,
procedure, and clean and soiled utility areas and around sinks. Carpeting may be
desirable for other areas; the direct patient care area is where this question
remains most controversial. Carpet, however, has a significantly higher
replacement factor and the hospital administration must demonstrate a
commitment to the greater maintenance required to keep it clean.
Wall finishes increasingly include quilts or soft sculpture to provide aesthetic and
sound-absorbent qualities. Extensive use of highly durable railings or moldings is
necessary throughout the NICU, as walls are easily damaged by portable
equipment.
Ceiling materials should be designated with a noise reduction coefficient (NRC) of
at least 0.90. Many states are now allowing the use of certain types of nonfriable
acoustical ceiling tile that helps with noise abatement. The method of cleaning the
ceiling and changing lights should be considered in the design process so that this
can be accomplished with minimal disruption to patient care.
Infection Control Considerations

As noted previously, an infection control practitioner will be an important member
of the design team during the discussion of the layout of the NICU and the choice
of surface finishes. In addition, they will be helpful in discussing air quality issues
in design of the heating and cooling system. They also will be very interested in
the location and design of hand washing areas.
Sinks should be readily accessible throughout the NICU, within a few steps of any
bedside. They should be large and deep, so that a full surgical scrub can be
performed with minimal splatter, and the walls and floor surrounding them should
be covered with easily cleanable surfaces. Porcelain sinks generally are more
attractive and quieter in use than stainless steel. Faucets should operate handsfree; and soap, hand-drying supplies, and large trash receptacles should be easily
accessible. These should be designed to avoid cross-contamination, to be easily
cleanable, and to minimize noise production. At least some sinks should be
provided for children and individuals with handicaps. Signage above each sink
should contain written and pictorial hand-washing instructions. A convenient space
for waterless hand-cleaning solutions should also be available at each bedside.
P.72
Headwalls
The area surrounding the baby's bedside, containing service outlets, shelving, and
bedside storage, is commonly referred to as the headwall. It is the focal point
for creating a self-contained workstation at each bedside. This area must be easily
adaptable to changes in census and acuity and to future changes in care practices.
It must support and provide easy access to necessary equipment and supplies, as
well as to the baby. The headwall design also should contain a comfortable
working area for the staff and provide space for the family to personalize the
baby's surroundings. There are several vendors that supply headwall systems, or
these can be built on site, using either moveable or fixed-rail systems that can be
customized for local equipment and practices.
A complete headwall system might include the following items and capabilities:
Dimmer (rheostat)-controlled task lighting
Three oxygen ports
Three compressed air ports
Three vacuum ports
Twenty to 30 electrical outlets
Telephone jack
Computer terminal jack
Consideration also should be given to allowing space for future needs.

All electrical, vacuum, and gas outlets need to be simultaneously and conveniently
accessible, that is, usage of one outlet cannot obstruct other outlets, even when
equipment contains oversized plugs. Some electrical outlets should provide normal
power, whereas others should be on emergency power, because either system
could be temporarily incapacitated.
A fixed or moveable shelf to contain monitoring equipment should be located as
near to eye level as possible and within easy reach. A countertop should be
provided as a work surface, with at least 24 cubic feet of cabinet space below to
hold supplies. Raceways within this cabinetry to hold wiring and gas lines can
minimize the danger and unattractiveness of a clutter of cords, but should be
easily accessible for repairs or modifications. Support beams should be placed to
avoid visual obstruction or access problems. A ledge at the base of the headwall
will prevent moveable equipment, such as incubators or warmers, from damaging
the wall.
Heating and Cooling Systems

Heating, ventilation, and air conditioning (HVAC) is perhaps the most mundane
segment of NICU design, and the one in which most members of the design team
feel they have the least expertise; yet it is important to consider for several
reasons. We have previously discussed the contribution HVAC makes to
background noise in the NICU, a problem that is amenable to some design
considerations. The number of air changes per hour needs to be specified for the
patient care area in general, and for the isolation and procedure areas in
particular; minimums usually are dictated by state code. In the case of the
isolation area(s) and soiled utility room, a negative air pressure design, with
100% of the air exhausted to the outside, is imperative. In all cases, a highefficiency filtering system is needed to remove particulate matter from the air.
Control of temperature and humidity is particularly important when designing the
HVAC system for an NICU. The system should be able to maintain ambient
temperature in the NICU between 72 and 78 degrees Fahrenheit throughout the
year, even at the extremes of outside temperatures for that particular locale. A
relative humidity of 30% to 60% should be maintained as well, again, even at
local extremes externally. Maintaining temperature and humidity within these
guidelines will minimize heat and water loss for the babies and discomfort for the
staff.
Delivery of airflow into the unit requires considerable forethought. Return ducts
should be situated near the floor so that particulate matter is not carried upward.
Supply ducts should be located where drafts will not be a problem and should be
generous in number, so that high-velocity air flow is avoided. Placement of supply
ducts near external walls and windows should be carefully planned to avoid
condensation and to minimize convective heat loss or gain to the babies nearby.
The fresh air intake into the hospital HVAC system should be planned carefully to
avoid areas that will contain exhaust fumes from vehicles, nearby buildings, or
from the hospital itself.
Communication Systems
Communication systems comprise, perhaps, the segment of NICU design that
requires the greatest anticipation of future developments. It is likely that
communication patterns among NICU staff, and between the NICU and other
support areas of the hospital (e.g., laboratory, radiology, and pharmacy), will
change dramatically in the next decade. Likewise, information transfer among
infant monitoring devices, the staff, and the medical record will accelerate rapidly
in the same time frame. Because unit design and construction typically take
several years, lack of foresight in this area can be inconvenient at best and often
quite expensive.
Those planning an NICU should anticipate digital transfer of virtually all
information. It is likely that the entire medical record will be computerized, and
work areas within the patient care area should be designed accordingly, with
adequate space to add terminals. Ergonomics and lighting should be considered,
so that staff can work at these terminals with minimal strain. Monitoring systems
will be interfaced with all the equipment supporting the baby, such as the
incubator, ventilator, and intravenous pumps, and then with the patient chart.
This will allow the obvious benefit of making data acquisition faster and more
accurate, and it will provide new alternatives for patient alarm systems. At
present, most NICUs depend on audible alarms to alert staff that a baby needs
attention; but, in the
P.73
future, these alarms can be transmitted digitally to the staff via headsets, pagers,
or other devices that will not add to the noise of the unit.
Telephone and intercom systems also will change drastically with newer
technology. Much of the noise generated in the NICU is because of these devices,
but alternatives already exist, such as wireless headsets, which allow
communication to occur very efficiently without adding to noise levels. These
systems also will facilitate communication among staff in the multiple room
arrangement used in most NICUs currently being constructed. Because using
these systems requires a considerable adjustment in the local culture of the NICU,
the design process should include discussion with end users so that the system
chosen will address their needs and concerns.
Maintenance Issues
Many NICU designs that looked good in the blueprint stage have required major
revisions soon after construction because maintenance issues were not considered
adequately. The choice of carpeting, lighting systems, HVAC design, computer
devices, and bedside equipment should not be finalized until maintenance
problems are identified fully with those who will be responsible for their upkeep.
Many of these devices or systems cost more to maintain than to buy, so the
economic perspective also should be considered carefully over the anticipated 15to 20-year lifespan of the NICU, as well as the extent to which routine or
unanticipated maintenance will interfere with patient care.
Designing for Renovations

Unless the pace of change slows dramatically, all NICUs will be faced with the
desire to upgrade their facility every few years. Many of the suggestions (listed
previously) are practical to implement without a major building program (e.g.,
reduced ambient lighting; some noise control measures; better, more welcoming
signage). Others can be accomplished through a renovation-in-place project (e.g.,
adding carpeting, wall hangings, and acoustical tile; providing more rooming-in
space for families), but some will be impossible without new construction and
significant increases in available floor space. In the course of a new construction
project, it will be impossible to anticipate all the changes that might occur in
neonatal care and technology over the next 20 years, but a couple of general
principles should be kept in mind.
First, there is no substitute for adequate space. The square footage needed for
state-of-the-art NICU care has increased relentlessly, yet is still less per patient
than that allotted to pediatric or adult ICUs in virtually all hospitals. Most NICU
new construction projects completed in the last 20 years have been forced to
accept reductions in floor space from the ideal for fiscal considerations, which
often proved to be short sighted, as renovations were required only a few years
later. It is hard to imagine new developments that would reduce the floor space
requirements for NICUs, but it is easy to suggest those that might increase them,
such as increased parental access and continued development of new technology

to improve monitoring, diagnosis, and thereby survival and outcome of extremely
low-birth-weight infants.
Second, we are in the midst of a dramatic transition from the NICU as a hightech, sterile (in every sense of the word) environment, to one that resembles a
baby's bedroom, with all the implications that carries for parental access. This
does not mean technology will disappear, or that many bedsides will not continue
to resemble an operating room setup. It does suggest that this latter situation
should be the exception, and that NICU design teams will need to consider the
norm to be an area where each baby is surrounded by its family and a nurturing
environment.
EQUIPMENT SELECTION
Equipment selection is an integral part of the planning process, whether for
renovation or new construction. It is important to recognize that equipment
features change rapidly and that, with the advance of technology, specific plans
for space allocation and cost must be easily modifiable. Categories of equipment
that will need to be specified include:
Environmental (incubators, radiant warmers)
Life support (ventilators, extracorporeal membrane oxygenation, nitric oxide

systems)
Monitors
Diagnostic (x-ray, ultrasound, electronic scales)
Treatment (infusion pumps, phototherapy, suction, chest percussion)
Communications (telephone, computer terminals, printers)
General support (breast milk, pharmacy refrigerators)
All users, as well as consultants familiar with the process of equipment inventory,
planning, procurement, and installation, and with maintenance of the equipment,
should be part of the planning team.
The first step of the equipment selection process involves preparing a list of all
fixed and portable equipment that will be needed. Next, existing equipment should
be evaluated to determine which items can be used in the newly constructed
NICU. At this point, dimensions and general space and equipment-mounting
requirements should be transmitted to the design team so that design of the
patient care and storage areas can proceed while decisions are being made
regarding purchase of new equipment.
The choice and procurement of new equipment is itself a several-step process.
After deciding exactly what equipment will be needed and the budget available,
the selection team should become familiar with options available in the market. If
a considerable amount of new equipment is anticipated, it is wise to organize an
exhibitors' day where
P.74
all the major vendors can demonstrate their products to the largest number of
staff possible. Alternatively, most large medical and nursing conferences have
displays by the major vendors.
When evaluating a new product, considerations should include ease of use,
durability, ease of maintenance, the ability to interface easily with computer and
monitor systems, hazards such as noise and electromagnetic radiation, size and
portability, ability to upgrade, and cost. After all this information is gathered, the
equipment consultant should organize it into a report that can be given to all
interested parties (users, maintenance, and procurement staff) for comment.
Procurement then can proceed with a request for bids, and a final purchasing
decision can be made when these are available. A delivery schedule should be
developed in coordination with the design and construction teams so that
equipment will arrive in sufficient time to be assembled, tested, and installed
before the NICU opens, but not so far in advance that upgrades and modifications
in the technology occur while the equipment is sitting in storage.
After selecting the equipment, the financial planning team will need to consider
decisions regarding purchase versus lease and service contracts on each item.
Although many hospitals have standing policies for such decisions, certain factors
may still be worth reviewing. If a piece of equipment is a newer model of an item
from a manufacturer with which the hospital has had considerable experience, the
biomedical maintenance department may feel quite comfortable with assuming
the responsibility for repair without the benefit of a service contract, and purchase
of the item usually will be less expensive than leasing in the long run (although
the chief financial officer should confirm this based on the specific terms offered).
However, if the piece of equipment is an entirely new device and only one or two
are being acquired, lease or purchase with a service contract has considerable
value in that repairs will be made by experienced technicians, and faulty
equipment may be replaced more readily. In either case, the hospital should have
a very clear understanding of how quickly service will be available, and whether
replacement items will be immediately available if repairs cannot be made

promptly, especially if the equipment is crucial to the management of a critically ill
child. These commitments should be obtained in writing, and their reliability
confirmed by calling other units currently using similar equipment from each
manufacturer.
Before installation occurs, the equipment consultant should review architectural,
mechanical, electrical, and plumbing drawings to ensure that all specifications are
current and compatible with the equipment that has been selected. There is
nothing more frustrating than trying to install a piece of equipment that does not
fit or has inadequate electrical or plumbing support. There are also many human
dynamics that are crucial to this process, and the equipment consultant must be
able to work with each of the specialists involved and organize a process that will
be well structured and successfully implemented. Again, the construction of a fullsized, functional mockup of a patient care area should be considered as part of
this process before final approval of the blueprints.
REVIEW AND APPROVAL PROCESS

After several months of planning, the architects will review the completed plans
with the design team for final comment and approval. The hospital administration
will be asked to sign off on a series of construction documents before submitting
them to contractors for bids. Although it is common knowledge to architects and
construction managers, many members of the design team may be unaware that
changes after this time usually are expensive and sometimes impossible. A design
that has considerable flexibility built in will become an obvious benefit, so that
interim changes in care practices or equipment can be accommodated easily.
When bids on the project are received, however, some major changes may still be
required, because financial projections by the architects are, of necessity in such
one-of-a-kind projects, only rough estimates. Sometimes compromises are fairly
straightforward, such as:
If aisles were generously sized, a reduction in all aisle widths by 1 foot

might save tens of thousands of dollars.
Bed positions might be reduced in number or size.
Offices and conference areas might be reduced in size, shared with adjacent
units, or redesigned as flexible-use spaces.
The mission statement developed at the beginning of the planning process
provides an important point of reference here. If compromises appear to be

required that would make accomplishing that mission and the accompanying goals
and objectives difficult or impossible, it may be more appropriate to revise the
budget rather than the design or (particularly in areas that already have highquality neonatal providers) to reconsider the project's feasibility altogether.
Before finalizing the blueprints, the design team should spend one more session
exploring all the possible sources of unanticipated problems. Any device or system
that could break or malfunction should be reevaluated to see if a change in the
design would minimize the impact of such an event. For example, is there
anything in, or attached to, the ceiling that could break or come loose and fall on
a baby? Is there any wiring, plumbing, or ductwork hidden in a wall for which
access might be very difficult or disruptive? Is there any passageway that might
be obstructed by a piece of equipment just when a resuscitation became
necessary? Architects cannot anticipate all of these events, which usually are not
discussed in the planning process, and staff have difficulty visualizing the impact
of such events from looking at blueprints, so this process usually is best done with
several members of each discipline committed to brainstorming together for
several hours. This is another reason visits to other NICUs can be particularly
P.75
valuable to one who is as interested in the flaws as well as the positive features of
the design.
After a bid is accepted (a process that, like the choice of an architect, should be
based not only on cost but on the extent and quality of the contractor's previous
experience), a meeting should be held with the contractor and members of all the
planning teams to be sure the construction team has caught the vision so
carefully thought out over the previous months. Far from being simply
contractors, the construction team can offer many helpful suggestions along the
way if they understand the planning teams desires, especially with regard to
specifics such as easily accessible wiring, minimizing noise generation from HVAC
and plumbing supplies, and so forth. Walkthroughs should be performed regularly
throughout the construction process, as multiple fine-tuning issues that were not
anticipated in advance undoubtedly will occur. Any resulting changes from the
architectural drawings should be documented carefully on as is drawings, so that
future renovations are not hampered by unpleasant surprises.
HUMAN DYNAMICS CRUCIAL TO A SUCCESSFUL

MOVE
In renovation projects, construction often must proceed in stages, while patient
care continues to be provided in the existing unit. With either new construction or
renovation, the implementation of new equipment and care practices can be both
exciting and stressful, and it is not unusual for turnover of staff to increase around
the time of the move. There are two contrasting strategies that have been used to
ease the transition to a new unit. One school of thought suggests that introducing
new equipment and practices to the greatest extent possible before the move is
valuable to minimize the culture shock of transition. An alternative strategy is
based on the concept that acceptance of new practices is most successful when
undertaken wholesale, especially if some issues that are anxiety provoking (e.g.,
increased access for families) are balanced by others about which the staff will be
excited (more space, better equipment). In practice, the transition period requires
belief of both philosophies, because some changes may not be possible until the
new unit is built (e.g., rooming in for parents), whereas others will be desirable to
implement as soon as possible (e.g., a new ventilator).
Certainly the most important strategy in this regard is to integrate the staff as
fully as possible into the planning, design, and construction process. The staff
must understand and buy into the conceptual changes intended by the mission
statement and around which the design process proceeded. Attendance at
committee meetings and posting blueprints are helpful in this regard, especially if
comments are encouraged and used. Additionally, working with a full-scale
mockup of a patient care area and occasional visits to the construction site are
very helpful to those who have difficulty visualizing two-dimensional renderings.
CHANGE-IN-PLACE
The final step of any NICU construction is a commitment to change-in-place. We
are just beginning to understand the biological effects of the environment on
premature infants, especially the positive and negative effects of light, sound,
touch, movement, and smell at each gestation stage. Likewise, a better
recognition of the role of parents and staff in the care and nurture of their babies
will lead to improved care practices, and technological improvements are sure to
continue. Each of these trends will influence our concept of the optimal NICU
design and should be incorporated to the greatest extent possible on an ongoing
basis within an existing structure, rather than waiting until new construction again
becomes feasible. An agreement by all members of the planning process that the
NICU will be considered a work in progress, rather than a finished edifice, will
enhance the readiness of all disciplines to implement change when the need
becomes apparent and will drive the design teams to build in as much flexibility as
possible.
REFERENCES
1. Gilstrap LC, Oh W, eds. Guidelines for perinatal care, 5th ed. Elk Grove
Village, IL: American Academy of Pediatrics, 2002.
2. Recommended standards for newborn ICU design. J Perinatol 2003;23:S3S21.
3. American Institute of Architects Academy of Architecture for Health.

Guidelines for design and construction of hospital and health care facilities.
Washington, DC: The American Institute of Architects, 2001.
4. White RD (ed). The sensory environment of the MCU: scientific and designrelated aspects. Clin Perinatol 2004;31:199393.
5. White RD. Enhanced neonatal intensive care design: a physiological

approach. J Perinatol 1996;16:381384.
6. Pediatrix University, NICU Design Center Section. Available online at http://

www.natalu.com.
7. Bullough J, Rea MS. Lighting for neonatal intensive care units: some critical
information for design. Lighting Res Technol 1996; 28:189.
8. White RD. Individual rooms in the NICUan evolving concept. J Perinatol

2003;23:S22-S24.
9. Rea MS, ed. The IESNA lighting handbook, 9th ed. New York: Illuminating
Engineering Society of North America, 2000.
10. Philbin K, Graven S, Robertson A. The influence of auditory experience on

the fetus, newborn, and preterm infant: report of the sound study group of the
national resource center. J Perinatol 2000;20:S1-S142.
Chapter 7
Organization of Care and Quality in the
NICU
Richard Powers
Carolyn Houska Lund
Providing current, research-based care to critically ill new borns requires the
collaboration of highly skilled, dedicated, and motivated caregivers from a variety
of disciplines. Professional nurses, physicians, respiratory therapists, social
workers, developmental care specialists, pharmacists, clinical dieticians, and
occupational and physical therapists have roles to play in planning, implementing,
and evaluating care for infants and their families in the neonatal intensive care
unit (NICU). This chapter includes a brief review of the basic organization and
components of the NICU. An in-depth discussion of quality improvement is
presented, including the systems to monitor and improve the quality of care in the
NICU. These systems involve all professional disciplines and are integral in
providing care to high-risk infants.
ORGANIZATION OF CARE
The organization of care in the NICU includes the medical staff working in
collaboration with nursing and other departments in the care of patients in the
NICU. Decisions about delivery of care, unit philosophy, and future directions are
best made through this collaborative process rather than by any one department
or discipline.
However, most NICUs are organized structurally and financially around the nursing
component. The organization and functioning of a NICU is dependent on nursing
leadership that can provide knowledgeable support and nursing input for the
following functions: strategic planning, budget development and implementation,
staff development, education, quality assurance and improvement,
interdepartmental collaboration, and clinical standards development.
Nursing leadership is provided by nurses with advanced education, training, and
experience in the following roles: nurse manager, clinical nurse specialist (CNS),
neonatal nurse practitioner (NNP), nurse educator, transport or extracorporeal
membrane oxygenation (ECMO) coordinator, and case manager. Depending on the
size and complexity of the NICU, some of these roles may be combined.
Caregivers who work in the NICU require emotional support because of the high
stress and sensitive nature of their work with critically ill infants and families. The
nursing leadership group actively seeks out situations that are stressful to staff
and provides support to staff through stress debriefing and staff case conferences.
Referrals to hospital ethics committees and appropriate professionals such as
psychologists, psychiatric nurse liaisons, chaplains, and employee assistance
programs should be made early to assist staff when needed.
Nurse Manager
The nurse manager has overall responsibility for the day-to-day operation of the
NICU and for coordinating and collaborating with the medical staff, other
department directors, and nursing managers from other units. The manager is
usually expected to plan and implement both capital and operations budgets.
Further, the manager works with the nursing leadership team to ensure that
education, ongoing development, and competency of the nursing staff are
attained. The nursing leadership group, in collaboration with the medical team and
ancillary disciplines, is also responsible for assuring that the quality of care
delivered in the NICU is safe and appropriate, meets regulatory standards, and
demonstrates a commitment to ongoing quality improvement.
Although the nurse manager and nursing leadership group are accountable for the
care delivered in the NICU, the staff nurses are the keystones of care delivery.
Therefore, facilitating staff participation at every level of decision making is critical
to the success of any unit operations. In some units, this collaboration may be
formalized through a system of shared governance in which staff nurses are
empowered to govern many aspects of unit operations. However, even in more
traditional organizational
P.77
structures, participation by staff is critical to the successful operation of the unit.
Advanced-practice Nurses
The CNS is an advanced practice nurse at the master's degree level. In a nonline
position in the nursing structure, the CNS generally has no direct authority over
other staff. The CNS role involves direct clinical care, consultation to nursing staff
and other professionals, and education of staff and parents. Research is a
component of the CNS role, and this is accomplished by keeping abreast of current
research applicable to neonatal care, implementing research-based practices,
supporting and facilitating research efforts in the NICU, and participating in
research studies as primary investigator or co-investigator. Maintaining quality of
care is another aspect of CNS practice; monitoring care practices, identifying
problems, and participating in the NICU multidisciplinary practice committee are
essential aspects of the CNS role and its effective implementation in the NICU (1).
Many units use the CNS in case management functions. Case management, a
model of care delivery, focuses on multidisciplinary care throughout hospitalization

to achieve desired patient outcomes in an expected time frame and with efficient
use of resources (2,3). Within this model, the CNS follows a caseload of patients
throughout their lengths of stay in the NICU, consulting with nursing staff,
physicians, and other team members about the patients' expected courses and any
variations that occur. The CNS coordinates the interdisciplinary model through
weekly interdisciplinary rounds to discuss medical and social aspects of care for
each infant in the NICU and identify needed interventions, procedures, or family
communication. The CNS may identify a need for individual patient care
conferences during which medical staff, nurses, and specialty consultants discuss a
complicated patient and establish reasonable goals with evaluation criteria.
The NNP has formal education and certification in the medical management of highrisk newborns. NNPs generally carry a caseload of neonatal patients with
consultation, collaboration, and supervision from a neonatologist. With extensive
knowledge of physiology, pathophysiology, and pharmacology, the NNP functions
both independently and interdependently with physicians in the assessment,
diagnosis, and implementation of specific medical practices and procedures. Other
responsibilities may include delivery room resuscitation, stabilization and transport
either within the hospital or to other facilities, education, consultation, and
research at varying levels (1,4,5). In addition to advanced-practice nurses, other
nursing roles include neonatal clinical educators, outreach educators, transport
nurses, discharge planners, nursing shift coordinators, and ECMO specialists.
The role of the charge nurse in the NICU is an important leadership position in the
daily operations of the unit. Charge nurses may be nurse managers, shift
coordinators, or experienced staff nurses responsible for the smooth functioning of
the unit during each shift in a 24-hour day. Among their responsibilities are
evaluating the number and level of acuity of all infants to determine the number of
nursing and support staff required, and communicating the assignments of
patients and staff. Charge nurses also arrange incoming and outgoing transports
and may, in some settings, attend high-risk deliveries. They are often the extra
pair of hands needed during emergency situations or special procedures, are
consulted by other nurses in problem situations and conflict resolution, and
become involved in crisis intervention with families. Astute assessment and
problem-solving skills, along with excellent communication abilities, are necessary
in the successful implementation of the charge nurse role and the daily running of
the NICU.
Staff nurses help families in their adaptation to and resolution of crisis by
providing consistency in the delivery of daily nursing care. Infants are regularly
assigned to a nurse or a group of nurses on admission or shortly thereafter, and
consistent care is delivered by this primary nurse or primary team throughout the
hospital course.
Primary nurses provide direct patient care, organize and write individual care
plans, and collaborate with the neonatologist, social worker, and advancedpractice nurse to facilitate the smooth transition of the infant throughout the
hospital stay and discharge from the NICU. They have extensive knowledge about
the individual responses of patients for whom they care on a daily basis and are
invaluable to the neonatologist and other team members. Primary nursing care
lends itself well to the developmental care that is specific in both assessment and
interventions aimed at the individual needs and unique characteristics of neonatal
patients.
Primary nursing is highly valued by families of infants in the NICU. Seeing the
same person caring for their infant is comforting and establishes trust during this
period of crisis and disequilibrium for families. Families often share their feelings
and reactions with someone they have come to know and trust; this is often the
primary nurse.
Other essential roles are necessary for the safe and effective functioning of the
NICU. Respiratory specialists provide expertise in the NICU in areas of pulmonary
care and assisted ventilation. Clinical dieticians consult regularly for both
parenteral and enteral nutrition issues. Pharmacists assist regarding the safe and
appropriate use of the multitude of medications administered in the NICU, as well
as monitoring for adverse drug reactions and side effects. Developmental
specialists, along with occupational and physical therapists, are responsible for the
integration of developmentally appropriate interventions for specific infants, as
well as educating other team members about developmental care and assisting in
environmental modifications which can improve patient comfort and possibly even
outcomes. Social workers assist in crisis intervention and provide psychosocial
assessment and emotional support as they advocate for the wide range of families
that encounter the NICU experience. In culturally diverse settings, translators are
indispensable in the NICU to ensure that information is accurately provided to
families.
P.78
Family-centered Care
The initial phase of hospitalization for high-risk infants results in significant
disequilibrium for families. The expected outcome of their pregnancy has been
changed from a healthy, full-term newborn to a premature newborn or a newborn
with significant medical or surgical problems. With prenatal detection of problems,
as well as perinatal care for premature labor, the families may have some idea of
the situation they are facing. Yet, many have not faced a crisis of such importance
and may need help in developing coping skills, understanding complicated medical
information, and learning how to be an advocate for their infant. Nurses use
therapeutic communication, crisis intervention, and supportive techniques to assist
families during this time (6).
Because many families may have additional social risk factors, including language
or cultural differences, poverty, chronic illness, or substance abuse, knowledge
about the impact of these factors on coping with crises and parenting is needed.
The importance of early intervention cannot be emphasized strongly enough, and
interventions by neonatal nurses along with NICU social workers and
neonatologists can have considerable positive effects for high-risk families during
this time of disequilibrium.
Family-centered care is both a philosophy and approach to care that can enhance
the potential of families to cope with the crisis and experience a positive outcome.
Principles for family-centered neonatal care include open and honest
communication in both medical and ethical considerations, providing in-depth
medical information in terms that are meaningful, and accessibility to other
parents who have had infants in similar circumstances. Information is provided to
families early if neonatal problems are diagnosed prenatally. Parents are allowed
to make decisions for their infants about aggressive treatments once they are fully
informed with adequate medical knowledge. Additional areas addressed in familycentered care are alleviation of pain, ensuring an appropriate environment,
providing safe and effective treatments, and policies and programs that promote
parenting skills and maximum involvement of families with their infants in the
NICU (7,8,9). Key elements of family-centered care are outlined in Table 7-1.
TABLE 7-1 THE KEY ELEMENTS OF FAMILY-CENTERED CARE
Incorporating into policy and practice the recognition that the family is the
constant in a child's life while the service systems and support personnel
within those systems fluctuate.
Facilitating family/professional collaboration at all levels of hospital, home,
and community care:
Care of an individual child
Program development, implementation, evaluation, and evolution
Policy formation
Exchanging complete and unbiased information between families and
professionals in a supportive manner at all times.
Incorporating into policy and practice the recognition and honoring of
cultural diversity, strengths, and individuality within and across all families,
including ethnic, racial, spiritual, social, economic, educational, and
geographic diversity.
Recognizing and respecting different methods of coping and implementing
comprehensive policiesand programs that provide developmental,
educational, emotional, environmental, and financialsupports to meet the
diverse needs of families.
Encouraging and facilitating family-to-family support and networking.
Ensuring that hospital, home, and community service and support systems
for children needing specialized health and developmental care and their
families are flexible, accessible, and comprehensive in responding to
diverse family-identified needs.
Appreciating families as families and children as children, recognizing that
they possess a wide range of strengths, concerns, emotions, and
aspirations beyond their need for specialized health and developmental
services and support.
(Reprinted from Shelton T, Stepanek JS. Family-centered care for children
needing specialized health and development services. Bethesda:
Association for the Care of Children's Health, 1994, with permission.)
Parent and family education are necessary throughout the hospitalization in the
NICU. Initially parents need information about their infant's medical condition and
what the prognosis is, as well as an introduction to the NICU personnel they
encounter (who does what). Pamphlets and booklets about premature infants or
specific disease conditions may be helpful. There are also several books written by
parents or NICU professionals that contain detailed information, illustrations, and
accounts of other parents' reactions to the experience in the NICU (10,11). The
Internet is another source of information for parents. Each unit should wisely
evaluate which resources on the Internet contain the most up-to-date, factual,
nonbiased information about specific conditions and post these resources for
parents to access if they wish. Although written information is valuable, it is not a
substitution for conferences and verbal interchange with parents. These
conferences are focused on what the professional staff expects that the family
needs to hear, what the parents are concerned about, and the parents' feelings
and reactions to what is happening to them and their infant. Nurses can help
parents become involved in the physical care of their
P.79
infant by showing them the things they can do such as comfort measures, bathing,
skin or mouth care, changing diapers, taking the temperature, holding as soon as
their infant is stable on the ventilator or oxygen (Fig. 7-1), and providing breast
milk.
Figure 7-1 Photo of a mother holding her ventilated preemie skin to skin.
Discharge teaching is an important aspect of family-centered care. This includes

well-baby care with knowledge about specific patterns of feeding, sleeping,
urination, stooling, breathing, skin care, and appropriate use of infant car seats.
Teaching about medication administration, including the purpose, method of
administration, side effects, and where to obtain the medications, is essential.
Special needs are described and skills taught, including gavage or gastrostomy
feeding, oxygen administration, ostomy care, tracheostomy care, cardiorespiratory
monitoring, and others. Different disease processes or conditions are identified for
parents, such as bronchopulmonary dysplasia, short-bowel syndrome, and
hydrocephalus, and symptoms are identified for parents so they can assess and
seek medical care appropriately after the infant is home (12,13).
QUALITY IN THE NEONATAL INTENSIVE CARE

UNIT
The many disciplines and personnel described above provide the structure around
the delivery of care in the NICU. However, to move beyond day-to-day care
delivery and improve the quality of care, a sound knowledge base in quality
improvement theory and practice is essential.
Background of Quality Improvement

Over the past decades, medicine has witnessed a rapid expansion of knowledge
and technology. This expansion has occurred in parallel with financial pressures
brought on by inexorable increases in per capita costs of health care and
limitations in financial resources available to the health care system. These forces
are especially applicable to intensive care subspecialties such as neonatology on
which a significant amount of research and technology is focused and for which
delivery of care can be extremely costly.
Research in the form of randomized controlled trials (RCTs) has become the gold
standard for evaluating the efficacy of health care interventions. In 1966, about
100 articles were published annually in all fields of medicine from RCTs; by 1995,
more than 10,000 were published (14). More than 3,000 articles about
neonatology RCTs have been published alone (15).
In the face of this avalanche of information on clinical efficacy and rapid infusion of
technology driven by the computer and pharmaceutical industries, health care
workers and institutions have major challenges before them. Individuals in health
care organizations need to efficiently evaluate new interventions and adopt the
most compelling ones in a timely manner to provide optimal patient care and avoid
preventable complications. It is through the principles of quality improvement,
along with organizational adaptability, that continuous integration of research,
technology, and improved patient care outcomes is accomplished.
Quality of care is defined by the Institute of Medicine as the degree to which
health services for individuals and populations increase the likelihood of desired
health outcomes and are consistent with current professional knowledge (16).
This definition, first proposed in 1990, has become widely accepted and is still
considered the best definition of health care quality today. The concept of health
services for individuals and populations is especially important in neonatology
where evaluation is often determined by population data such as infant and
neonatal mortality rates or the incidence of neurological deficits among a specific
subgroup such as extremely low-birth-weight survivors.
The definition also emphasizes that quality care increases the likelihood of
beneficial outcomes, a reminder that quality is not merely the achievement of
positive outcomes. Poor outcomes occur despite excellent care because diseases
vary in severity and can defeat even the best efforts. Conversely, patients may do
well despite poor quality of care. Assessing quality thus requires attention to both
processes and outcomes of care. The last part of the definition of quality,
consistent with current knowledge highlights the dynamic and evolving body of
knowledge available to health care professionals and the need to revise and
update measures of quality as new interventions become standards of care.
Problems in quality of health care can be classified in three categories: underuse,
overuse, and misuse (14). Underuse is the failure to provide a health care service
when it would have produced a favorable outcome. For example, failure to provide
surfactant in a timely manner after the delivery of an extremely low-birth-weight
infant with respiratory distress syndrome would indicate underuse. Overuse occurs
when a health care service is provided despite the fact that its potential for harm
exceeds its possible benefit.
P.80
The widespread use of postnatal steroids for chronic lung disease popular in the
1990s is an example of overuse in neonatology. Misuse occurs when a preventable
complication arises during administration of an appropriately selected treatment.
Misuse includes many of the common medical errors that occur during
hospitalization or other health care encounters. Medical errors have been
extensively discussed (17,18), driving numerous initiatives by the United States (U.
S.) government and regulatory agencies aimed at understanding the human and
systems factors that contribute to the errors. External reporting systems that
collect information on adverse events and errors are important in the reduction of
future errors by alerting practitioners to new hazards, using the experience of
individual hospitals using new methods to prevent errors, and revealing trends
that require attention (19). In neonatology, medical errors have been collated and
classified as part of an anonymous error-reporting project in conjunction with the
NIC/Q Quality Improvement Collaborative of the Vermont Oxford Network (Table 72) (20).
Regulatory agencies, in conjunction with federal and state governments, have
traditionally been charged with the task of motivating health care professionals
and organizations to maintain and improve quality. The Joint Commission on
Accreditation of Healthcare Organizations, formed in 1951, initially developed
standards for hospitals and evaluated compliance to these standards,
hypothesizing that compliance with these standards would correlate with quality
care and positive outcomes for patients in hospitals. In accreditation, quality is
evaluated by monitoring adherence to accepted standards and measuring
outcomes. Standards used by accreditation organizations are derived from a
variety of sources, including government (via regulatory agencies at both the
federal and state levels), as well as professional and community-based, standards
of practice.
Regulation is for the most part successful in establishing minimal standards of
performance and is an important means of protecting the public from egregiously
poor providers. It has, however, numerous limitations. Standards are difficult to
enforce uniformly, and regulation tends to be inflexible with difficulty in adapting
quickly as knowledge changes. Regulation also fails to stimulate organizations to
integrate new technologies or developments and does not motivate them to
continuously improve. Continuous quality improvement (CQI) can supplement the
deficiencies of regulation alone, while providing an impetus for individuals and
organizations to strive for the highest quality of care.
TABLE 7-2 CLASSIFICATION OF MEDICAL ERRORS IN NEONATOLGY(20)
Error Classification
Wrong medication/dose/schedule/infusion rate
Percent of Errors
53.4%
Error in performance of operation, procedure or test

Patient misidentification
12%
7.8%
Error in administration or method of using a treatment

Error or delay in diagnosis
7%
4.7%
Equipment failure
Failure of communication
2.3%
1.3%
Other system failure
6.7%
Continuous Quality Improvement

CQI emerges from the industrial sector as an effective system to reduce errors in
production. It motivates good performers to excel, emphasizes identification of
potentially successful change opportunities, and facilitates change implementation
throughout all levels of the organization. CQI provides the framework for
organizations to keep abreast of current knowledge and innovations, identify
appropriate changes, and implement them in a timely manner. There are three
components to CQI: measurement, benchmarking, and collaboration.
Measurement
The first and most basic element of CQI is the acquisition of data. Data acquisition
drives information, which in turn drives action. Over the past 25 years, numerous
systems of quality measurement have been developed, encompassing the areas of
outcomes, processes, and patient satisfaction.
Outcome measures represent the most objective and often the most meaningful
data for health care organizations. When applied to populations, outcome
measures provide essential feedback to leaders charged with resource allocation,
managers charged with developing successful and efficient organizations, and
individual health care providers.
Due to variability in disease severity among patients from different socioeconomic
and cultural backgrounds, as well as differences in the type of patients cared for in
highly specialized tertiary centers compared to community health facilities, data
based on outcomes alone can be inaccurate or misleading. Process measures are
also important in evaluating overall quality. Measures of process are needed to
determine that accepted standards of care are being met regardless of good or
bad outcomes.
TABLE 7-3 TYPICAL QUALITY IMPROVEMENT MEASURES
Process measures
Intrapartum antibiotics for mothers with positive group B strep cultures
Intrapartum antibiotics for mothers with group B strep risk factors
Antenatal steroid use
Admission temperature
Postnatal steroid use
Surfactant administration
Incidence of hypocarbia (PaCO2 <30)
Outcome measures
Mean 1-minute and 5-minute Apgar scores
Survival rate
Length of stay
Incidence of chronic lung disease
Incidence retinopathy of prematurity
Incidence of intraventricular hemorrhage
Nosocomial infection rate
Patient satisfaction measures
Comfort of facilities
Appearance of room
Nurses' attitudes toward requests
Facilities for family information provided
Doctor's concern for questions/worries
a
From Press, Ganey Associates, Inc., South Bend, IN. 2002

P.81
A third category of measurement in assessing quality of care is patient and family

satisfaction. This is the result of applying traditional marketing techniques to the
health care industry, and it has accompanied the adoption of CQI principles from
industry to health care. It is also a natural result of a larger movement throughout
health care that recognizes the autonomy and accountability of the patient and
family. As the health care community expands its expectations that patients and
their families take on a larger part of the responsibility to maintain their own
health and wellness, the feedback obtained from patients and families regarding
their interaction with the health care system is crucial. Table 7-3 shows examples
of neonatology quality measures in the three areas of outcomes, processes, and
patient satisfaction.
Benchmarking: Use of Comparative Databases

The second step in the process of CQI is benchmarking of outcomes. Opportunities
for benchmarking are numerous, thanks to the recognition in the past 15 years of
the contribution it brings to CQI. A number of regional, national, and international
databases have been organized in neonatology, providing benchmarking
opportunities through voluntary participation and confidential reporting of
individual center outcomes.
One of the first neonatology databases, and currently the largest, is the Vermont
Oxford Neonatal (VON) Network. Started in 1990 with 36 hospitals, this network
has grown to 380 centers (21,22). The network includes data on more than
25,000 very low-birth-weight (<1,500 g; VLBW) infants each year, more than
50% of all VLBW infants born in the United States annually.
In the VON Network, centers report outcomes, including survival and lengths of
stay for all VLBW infants admitted to the NICU. They also report incidence of
chronic lung disease and complications, including nosocomial infection,
pneumothorax, necrotizing enterocolitis, intraventricular hemorrhage, retinopathy
of prematurity, and other conditions. All participating centers receive a confidential
annual report showing their performance compared with the database as a whole.
Each center can see how they rank with all other centers, and with centers that
are grouped in similar categories by number and type of NICU admissions. In the
VON Network, all of the variables are reported in aggregate form showing the
mean incidence rate and highest and lowest quartile of each measure. The
mortality rate and length of stay for each center is also adjusted for patient acuity.
Other databases have been formed at regional and national levels. The National
Institute of Child Health and Development Neonatal Research Network provides a
venue for participating institutions to submit outcome measures; their aggregate
data has been published to serve as a reference for other centers to compare their
performances. These include general survival and complication rates in VLBW
infants (23,24) and rates of neurological abnormalities (25). The Canadian
Neonatal Network has published reports tracking overall outcomes and
complications in infants of all gestational ages (26,27,28). Other national networks
include the Scottish Neonatal Consultants and Nurses Collaborative Study Group,
New South Wales Neonatal Intensive Care Unit Study Group, and Paulista
Collaborative Group on Neonatal Care in Brazil (29,30,31).
Variability in Neonatal Intensive Care Unit

Outcomes
Variability in neonatal outcomes becomes apparent when data from multiple
newborn intensive care units is analyzed in comparative databases. The VON data
describing the first full year of the collaborative in 1990 showed a number of
differences in outcomes among VLBW infants in the 36 reporting centers. These
included the incidence of respiratory distress syndrome, pneumothorax, chronic
lung disease, patent ductus arteriosus, nosocomial infection, intraventricular
hemorrhage, and retinopathy of prematurity (22). The annual VON reports provide
the distribution of outcomes among centers by ranking the data and calculating
the 25th and 75th percentiles for mean incidence of the given outcome from each
center. Percentile ranking of mean values from individual centers represents a
simple and effective means of illustrating one center's ranking among the entire
sample of participants. Table 7-4 shows the 2001 data from the VON database,
illustrating this methodology for selected outcomes (32).
Investigators also report variation in outcomes as part of multicenter prospective
interventional trials or retrospectively as multicenter independent research. Brodie
and
P.82
associates (33) studied nosocomial bloodstream infections in VLBW infants in 6
NICUs in the Boston area from 1994 to 1996; mean incidence of infections was
19.1% for the whole group, but varied from 8.5% to 42% among the 6 units. After
adjusting for patient- and treatment-related variables, significant variation
persisted. Variations in blood transfusions among 6 perinatal centers in
Massachusetts and Rhode Island were studied, showing a mean total transfused
volume ranging from 95.5 mL/kg (highest) to 35.0 mL/kg (lowest) (34). Avery and
associates (35) described the variability in incidence of chronic lung disease among
8 units surveyed. Later, an in-depth review of the practices related to respiratory
support for infants with respiratory distress syndrome was undertaken at the 8
centers, triggering the study and dissemination of a number of innovations in
respiratory care practices from the unit reporting the best outcome.
TABLE 7-4 OUTCOMES AND INTERVENTIONS (VERMONT OXFORD
NETWORK 2001, 32)
Outcome/Intervention
Respiratory distress syndrome

Pneumothorax
Oxygen use
Nasal CPAP use
Ventilation
Chronic lung disease
Patent ductus arteriosus
Necrotizing enterocolitis
Nosocomial infection
Intraventricular hemorrhage
Grade 3
Grade 4
Retinopathy of prematurity
Percent
Interquartile Range (25th,

75th Percentile)
72
6
90
60
71
29
33
6
21
63,82
3,7
86,96
60,73
63,80
17,36
22,40
2,8
12,26
4
5
2,6
2,7
Stage 3
10
Stage 4
1
CPAP, continuous positive air pressure.
5,13
0,0
Wide variation in outcomes among centers is often found when centers participate
in comparative outcome studies. Even when the data has been adjusted for
confounding risk factors, marked variability still exists in many cases. Explanations
for this persistent variation include differences in case mix, data quality, and case
finding. However, the final and most important factor is often variation in
effectiveness of clinical practice.
It can be extremely useful and important for units to recognize how clinical
effectiveness contributes to variability, especially in areas where the outcomes for
the unit are in the lowest quartile. One of the major benefits of participating in the
comparative analyses of outcome measures lies in the understanding that
changing clinical care practices truly can influence their outcomes. In most cases,
individual units find outcomes in the lowest quartile for only a few variables of the
data set, with the majority falling within the interquartile range (25th-75th
percentile) or even exceeding the 75th percentile. The lowest quartile outcomes
provide target areas to which focused improvement efforts can be directed.
Furthermore, centers in the database that report better outcomes can be used as
resources to identify practices that may benefit centers in the lowest quartile.
Identification of benchmark data through concurrent measurement among centers
represents only one model for benchmarking. Published data can also be used as a
benchmark when concurrent data is not available. It is important to review the
methodology and data definitions in the published benchmark paper to allow for
consistency in data acquisition before any extrinsic comparison can be made. An
example is the nosocomial infection rates published by the National Nosocomial
Infection Surveillance project of the Centers for Disease Control which is used by
many NICUs to analyze their infection prevalence (36).
Collaboration
The concept of institutions collaborating with each other for the purpose of
improving overall quality of care is novel, yet essential, in successfully and
efficiently changing health care. Comparative databases with institutions
prospectively reporting outcomes to identify opportunities for improvement are an
example of how collaboration can benefit organizations. Collaboration is equally
important for implementing practices that enable institutions to adopt new
technology and bring about improved methods of delivering care.
Collaboration among different and sometimes competing companies is found in
industries outside of health care. For example, in the semiconductor manufacturing
industry
P.83
corporations must maintain the pace of technologic research and development to

compete. Yet 10 of the largest U.S. semiconductor manufacturers, including Intel,
Motorola, Advanced Micro Devices, and Texas Instruments, joined forces in the
mid-1990s to form a company called Semanteck to foster mutual information and
advancement (Austin, TX) (37).
This motivation to achieve common improvement goals among different
organizations has led to the formation of numerous quality improvement
collaboratives in health care. A number of factors serve as incentives to
collaboration. First is the substantial gap that exists between current knowledge
and the actual provision of care. Second, collaboration among institutions can
accelerate the infusion of existing knowledge into practice by providing insight into
how the centers with the best outcomes were able to implement systems that
yielded these results. Despite abundant literature describing practices that should
or should not be done, little is written about specific activities that facilitate
implementing desired practices (37). Dialogue between units, as well as visits
involving a multidisciplinary team from one unit to another, provides an exchange
of ideas and solutions to clinical problems.
The Institute of Healthcare Improvement Breakthrough Series is an example of
large collaboratives working to accomplish common aims, including reducing
cesarean section rates, improving asthma care, reducing adverse drug events and
others (38,39,40). The NIC/Q Collaboratives on Quality Improvement in
Neonatology, organized by the VON, have identified the goal of improving neonatal
outcomes in chronic lung disease, nosocomial infection, nutrition and other areas
(22). The California Perinatal Quality Care Collaborative is a regional collaborative
with the goal of improving perinatal care in California. It involves more than 100
NICUs throughout the state and brings, in addition to health care providers,
representatives from state certification agencies and various business groups into
the process. Through this diverse collaboration, goals with a broader definition of
quality are identified.
During the collaborative process, the decision to adopt new interventions or
evaluate existing practices is often encountered. A critical review of the available
evidence supporting new interventions and clinical practices is essential. With the
pace of development of new drugs and technology, the consistent and systematic
evaluation of evidence supporting new interventions becomes a priority.
Evidence-based Medicine
Evidence-based medicine is defined as the conscientious, explicit, and judicious
use of current best evidence in making decisions about the care of individual
patients (41). In the context of continuous quality improvement, the definition is
expanded beyond the individual patient to decisions regarding institutional
guidelines and policies in the care of multiple patients with similar diagnoses. In
both applications, the principles are the same: an answerable clinical question is
formulated, the best evidence is located, and the evidence is critically appraised.
These steps are essential whether answering a question regarding treatment for a
VLBW infant with a patent ductus arteriosus or creating a policy for management
of all VLBW admissions who develop the diagnosis of patent ductus arteriosus.
Randomized Controlled Trials

Randomized controlled trials (RCTs) represent the highest level of evidence and
are the foundation for establishing effectiveness of interventions. The methodology
of RCTs is geared to minimize selection bias by the random allocation of study
subjects at the time of enrollment. This unique characteristic of the RCT
establishes it as the basis for traditional statistical comparisons in both clinical and
basic science research. Sound scientific methodology must also address other
potential biases that can impact validity of the results. These include performance
bias, or nonuniform exposure to the intervention; exclusion bias, or incomplete
follow-up in obtaining postintervention data; and assessment bias, or inaccurate
measurement of outcomes.
Metanalyses
Although RCTs are considered the best single methodology to evaluate an
intervention, there are sometimes more than one RCT for a given intervention.
RCTs can be combined using the technique of quantitative systematic review, or
metanalysis. Pooling results of similar RCTs can increase the statistical power
lacking in multiple small RCTs or provide more support for decision making when
conflicting results are reported in separate studies on the same treatment. The
techniques used in performing metanalysis are rigorous. The methodology of
metanalysis has been formally outlined and includes 5 stages: (a) specify the
objectives of the review, (b) identify and select studies, (c) assess validity, (d)
combine results of independent studies, and (e) make inferences (15).
Objectives of the review must be clearly and succinctly stated at the outset, with a
principle objective and often secondary objectives given. The strategy for
identifying and selecting studies must also be clearly stated in the review. This
addresses one of the most challenging issues in any search and analysis of
available literature, publication bias. Publication bias refers to the tendency for
investigators to preferentially submit studies with positive results and the tendency
for editors to preferentially select studies with positive results for publication.
Klassen and associates (42) report that only 59% of abstracts presented at
meeting of the Society for Pediatric Research between 1992 and 1995 were
subsequently published. Abstracts were more likely to be published as a full study
if they reported good news about newer therapies. To document their rigor in
minimizing publication bias, authors of metanalyses must include a prospectively
designed search protocol, a comprehensive and explicit search strategy, and strict
criteria for inclusion and exclusion of studies.

P.84
Validity of each study in the metanalysis must be determined independently,
based on each study's merits and exclusion of bias in all areas. Results of multiple
studies are usually combined based on their principle outcomes or measures of
treatment effect. These are typically expressed using relative estimators, the
relative risk or odds ratio, and absolute estimators, event rate difference or risk
difference. Derived from the estimators are the relative risk reduction, computed
as (1 - relative risk), and the number needed to treat, computed as (1/risk
difference). For continuous outcome data measurements, the effect of treatment is
measured as the mean difference. These outcomes are determined for individual
trials in the metanalysis, and then computed for the entire systematic review using
statistical methodology that weights each trial based on its population size. The
statistical significance of a treatment effect is determined when the confidence
interval of the combined estimate for relative risk or odds ratio does not include 1,
or when the confidence interval for the risk difference or weighted mean difference
does not include 0.
Estimates of treatment effect or lack of effect are much stronger in metanalyses in
which there is less heterogeneity. Heterogeneity refers to the observation of large
differences in point estimates among the included studies. The ability to make
inferences from a metanalysis depends on the methodologic quality of the primary
trials on which the review was based, degree of consistency of results among the
trials contributing to the review, and degree of confidence that the search for all
trials relevant to the review was comprehensive (15).
The quality of a metanalysis is dependent upon the rigor of the methodology
applied by its authors. The process of metanalysis has its critics; attempts at
pooling results from various studies not only incorporates the biases of the primary
studies, but may add additional bias attributable to study selection and
heterogeneity of the selected studies (43). In addition, pooling data from small
studies not adequately powered cannot answer questions about the potential side
effects of new therapies, as compared to larger RCTs.
TABLE 7-5 EVIDENCE-GRADING SYSTEMS (45)
Grade
Description
Agency for Healthcare Research and Quality (45)

I.Metanalysis of multiple well-designed controlled studies
II.At least one well-designed experimental study
III.Well-designed, quasi-experimental studies, such as
nonrandomized controlled, single-group, pro-post cohort, time
series, or matched case-controlled studies
IV.Well-designed, non-experimental studies, such as comparative
and correlational descriptive and case studies.
V.Case reports and clinical examples
Canadian Task Force on the Periodic Health Examination (46,47)
I.Metanalysis of multiple well-designed controlled studies
II-1.Evidence obtained from well designed controlled trials without
randomization
II-2.Evidence obtained from well-designed cohort or case-control
analytic studies, preferably from more than 1 center or research
group
II-3.Evidence obtained from multiple time series with or without
intervention
III.Opinions of respected authorities, based on clinical experience,
descriptive studies or reports of expert committees
(Adapted from The periodic health examination. Canadian Task Force on
the Periodic Health Examination.
CMAS 1979;121:11931254; Strom KL. Quality improvement
interventions: what works? J Healthc Qual 2001;23:414, with permission.)
Evidence-Grading Systems
Although RCTs and metanalyses of RCTs represent the best sources of evidence,
such high-quality evidence is not always available. Multiple systems for grading
the strength of evidence have evolved that account for sources other than RCTs.
The majority of grading systems place the most value on inferences from a
systematic review of RCTs, with evidence from an individual RCT second, followed
by evidence from well-designed trials without randomization, evidence from
nonexperimental studies and, finally, opinions of respected authorities or reports
of expert committees (44). Sources that may not meet the gold standard of an
RCT are nevertheless important when this gold standard is either not yet achieved
or is not achievable. Numerous systems have been developed to rate the strength
of evidence, targeting evidence from metanalyses down to the opinion of experts
(45). Table 7-5 shows a summary of 2 typical evidence-grading systems (46,47).
Despite the obvious value of evidence provided by RCT and metanalyses,
examples of other forms of validation can be found in collaborative processes. The

Vermont Oxford Quality Improvement Collaborative for Neonatology reported that
one of the most fundamental components in reducing nosocomial infection in the
NICU observed
P.85
consistently in best-performing units was staff accountability for nosocomial
infections (48). Implementation of this valuable insight was critical for
collaborative participants desiring to reduce infections in their individual units.
Although not meeting the usual criteria for high-grade evidence since it has not
been formally studied scientifically, this form of observational data nevertheless
represents an important category of evidence, i.e. evidence gained from
collaborative benchmarking.
In addition to determining the efficacy of a certain intervention, disease
prevalence, complication rates, financial impact of treatment versus the disease
itself, and cost of implementation of the policy are all critical issues before
implementation of evidence-based treatments. Many of the questions can be
answered by a critical systematic review of available studies. Decisions to
implement an intervention for a population of patients should be thoroughly
researched and analyzed to account for case mix, variation over time, availability
of subspecialty resources, and baseline complication rate. Although only somewhat
relevant in an individual treatment decision, financial cost-effectiveness and
feasibility are vital issues to be addressed in adopting policies on a unit-,
institution-, or system-wide basis. Especially when financial and staff resources are
limited, the implementation of one costly intervention could preclude the
implementation of other interventions with more impact for a greater number of
patients.
As critical analysis of research and benchmarking allows an organization to identify
specific processes to improve, the actual incorporation of change into the
organization requires an approach unique to each separate organization.
Implementation of change requires individual commitment and open mindedness,
a multidisciplinary group culture that embraces new ideas, and an institutional
environment established by the management structure that encourages new ideas.
Organizational Change and Unit Culture

Each organization has a culture that is unique, representing the individuals and
relationships that make up that organization. As Baker points out, teams and
organizations that attempt to implement quality improvement often identify
organizational culture as an important barrier to or facilitator of success (49).
Change, and an organization's readiness for change, has been the topic of
numerous treatises in the field of industrial psychology. Although the traditional
view of change management is that change begins at the top of an organization
pyramid, recent views emphasize that change must occur at all levels within the
organization, especially on an individual level (50). O'Connor and Fiol (51)

describe 3 elements: energizers or motivators that compel people to take action,
impediments to change, and action steps needed for change to occur. Examples of
energizers include objectives, benefits, and negative consequences. Barriers to
change include apathy, errors, disturbed relationships, lost power, status, and
money. Action steps are the steps taken to overcome barriers and resistance to
change. It is important in CQI to understand common obstacles and the unique
interrelationships that characterize different organizations in their quests for
providing the same quality of care (51).
Tools for Quality Improvement

Fundamental to CQI is the use of improvement teams working together to solve
problems or improve processes in which they are key players. In this approach, a
process needing revision is analyzed using a series of tools, with multiple group
discussions among the team, progressing through specific phases of analysis. This
diagnostic journey progresses through description, cause-effect analysis,
elaboration of potential improvements, and prioritization of these improvements
based on research and study. The end result is the identification and
implementation of 1 or 2 improvements that will have the most constructive
impact on the related process.
One of the tools used in this process is flowcharting, which allows teams to depict
the sequence of activities that make up the process. As teams chart the sequence
of activities in a process, they often identify problem areas that become potential
opportunities for improvement. Subsequent analyses using another tool, the causeeffect diagram, allow the team to attribute causes to the problem areas in the
process being studied. Causes are classified in generic categories, which include
personnel, equipment, supplies, information, methods, measurement, and
environment. The determination of causes often requires analysis and chart
review, with group discussion and consensus determining the likelihood of a causeeffect relationship. After group consensus and further study using a Pareto
analysis (based on the 80-20 rule, ascribing 20% of the causes to account for
80% of the problems), the next step is to elaborate on the relationship between
potential problems identified and their relative effect on the global process under
review. The goal in this stage is to identify the relationship between factors
identified with the adverse outcome of the process under review, as well as to
systematically describe those factors which have the most potential to favorably
influence the process under review.
Traditional CQI incorporates several tools for data collection and analysis. Data are
essential in establishing the baseline activity or outcome of a process, ascertaining
the relative role of various potential interventions, and measuring the process to
document the desired improvement. Examples of data collection tools include
simple check sheets, data sheets, interviews, and surveys. Analysis of the data is
often performed using bar charts, histograms, line graphs, and scatter diagrams.
Another important and more sophisticated tool useful in this stage is the process
control chart (52,53) in which data is plotted over time; analyzing variation in the
data allows the distinction of special-cause variation from common-cause
variation. This distinguishes variation caused by the influence of extrinsic factors
from normal variation intrinsic to the process itself.
P.86
The sequence described represents traditional CQI, which has evolved over time
and led to important improvements in industry and health care. However, due to
the demands placed on modern health care organizations to implement numerous
changes over much shorter time periods, the traditional model may not be best, as
it may be tedious, be unnecessarily time consuming, and tie up precious resources
when used every time a change is needed.
Rapid-Cycle Continuous Quality Improvement

Most organizations are looking for results at a pace that meets objectives in a
more timely manner than traditional CQI. The impetus for getting results more
efficiently has led to the development of a new approach called rapid-cycle CQI.
Rather than merely speeding up the discrete stages of traditional CQI, rapid-cycle
CQI represents a completely different approach. Rapid-cycle CQI is a system of
parallel processes as compared to traditional CQI, a system composed of multiple
sequential processes performed in series. Originally designed by Nolan, Langley,
and colleagues (54,55), rapid-cycle CQI involves the identification of changes
chosen as goals by an organization; introduction of change concepts in small
cycles, or tests of change; and measurement of the effects of the change.
Rapid-cycle CQI has several basic prerequisites. Organizations with smaller, less
hierarchical leadership structures are best suited for this methodology. The use of
smaller sample size in initial assessment is another important element as it allows
for more rapid feedback regarding the success or failure of a test of change. Rapidcycle CQI also works best when potential solutions or successful change concepts
are readily available through prior applications in other settings. For more
complex, interdepartmental problems, other problem-solving methods, including
traditional CQI teams, may be more effective.
The process of rapid-cycle CQI begins with the question. What are we trying to
accomplish? The answer to this question constitutes the aim statement, the
starting point in applying rapid-cycle CQI. Next is establishment of the
measurement necessary to answer the question, How will we know that a change
is an improvement? Measurement is key to any methodology of CQI, and it is
especially key to rapid cycle because each small test of change is evaluated by
measurement. The progression toward the final outcome requires successive or
simultaneous cycles depending on measurement to provide the appropriate
feedback regarding the value of the change concept being evaluated. Change
concepts can be based on evidence in the literature, models of practice identified
in benchmark centers, ideas generated during brainstorm sessions by teams within
units, or modifications based on the result of previous tests of change.
Although not unique to the rapid-cycle CQI model, Plan-Do-Study-Act (PDSA)
cycles are important elements of the process. The small tests of change are set in
place with PDSA cycles. This involves the systematic planning of a specific
adaptation of the change concept followed by the unit making the adaptation. The
unit staff then studies its effects through measurement, and finally acts on the
outcome of the cycle, either deciding to continue with a further adaptation in
another cycle, or determining that the aim has been achieved and concluding the
process. The concept of building knowledge through PDSA cycles has a long
history with roots that can be traced to the British philosopher John Dewey (56),
and it has been adapted by Shewhart (57) and Deming (58) to the science of
quality management. Figure 7-2 shows the flow diagram used in rapid-cycle CQI.
Plsek (56) points out numerous examples of successful applications of rapid-cycle
CQI. These include 2 teams in the Breakthrough Series Collaborative sponsored by
the Institute for Healthcare Improvement. One team, from the Mayo Family
Medicine Clinic, identified the following change concepts: build capacity for routine
assessment of patient outcomes, reduce unintended variation in care, streamline
the process of care, and build information system capacity. By implementing these
concepts through a series of small-scale cycles of change, the Mayo team reduced
hospitalization in asthma patients by 47% and reduced emergency visits by 22%
(39). In another example from the Breakthrough Series, a second team
implemented a series of 3 change concepts and decreased readmission to the
intensive care unit from 15.6% to 9.8% (40).
The Vermont Oxford NIC/Q 2000 collaborative for quality improvement in
neonatology brought together
P.87
teams from 34 participating centers, and provided education, content experts, and
resources to facilitate the identification of change concepts in several clinical areas
and the application of rapid-cycle CQI principles. One of the focus groups of the
collaborative identified and prioritized 3 change concepts for reducing nosocomial
infection in the NICU: improved hand hygiene, attention to the care of deep line
connections, and standardized diagnosis of coagulase-negative staphylococcal
sepsis. They were able to demonstrate a reduction in coagulase-negative
staphylococcal sepsis from 25% to 16% (59).
Figure 7-2 Flow diagram of rapid-cycle CQI. (From Plsek PE. Quality
improvement methods in clinical medicine. Pediatrics 1999;103:203-214, with
permission.)
In rapid-cycle quality improvement, pace is important. Successful application has

been shown to depend on running small tests of change, short PDSA cycles, rather
than running longer ones over an extended period of time. Each cycle provides the
feedback forming the bases of learning and giving information on which to base
further cycles of change. Efficiency of measurement is also critical for successful
application of rapid cycle CQI. Teams have found that measurement within each
cycle must be objective and sufficient enough to draw conclusions about the effect
of the change concept. This often does not require the use of statistical techniques
that would be very difficult given the sample sizes chosen for rapid cycle tests of
change. Although not a requirement of many organizations successful at
implementing widespread changes, statistical testing may be reserved for
comparison of performance measures after large projects are completed, such as
the examples of reduced hospitalization rates and nosocomial infection rates cited
above.
CONCLUSION
The delivery of care in the NICU is a complex process involving numerous
disciplines and personnel. The day-to-day management is important to the overall
organization and keeps operations in motion. However, to continually improve
practice and reduce medical errors, a system of continuous quality improvement is
needed. The integration of comparative databases and benchmarking, evidencebased medicine principles, along with rapid-cycle processes that are compatible
with concepts such as unit culture and change cycles, will help an individual unit to
maintain quality in the face of technologic advances and organizational change.
REFERENCES
1. Stafford M, Appleyard JA. Clinical nurse specialists and nurse practitioners:
who are they, what do they do, and what challenges do they face? In:
McCloskey J, Grace HK, eds. Current issues in nursing. St. Louis: Mosby-Year
Book, 1994.
2. Elizondo A. Nursing case management in the neonatal intensive care unit.

Part 1: pioneering new territory. Neonatal Netw 1994; 13:912.
3. Strong AG. Case management and the CNS. Clin Nurse Specialist 1992;6:64.
4. Farah AL, Bieda A, Shiao SY. The history of the neonatal nurse practitioner in
the United States. Neonatal Netw 1996;15:1121.
5. Buus-Frank ME, Conner-Bronson J, Mullaney D, et al. Evaluation of the

neonatal nurse practitioner role: the next frontier. Neonatal Netw 1996;15:31
40.
6. Kenner C. Caring for the NICU parent. J Perinat Neonatal Nurs 1990;4:7887.
7. Harrison H. The principles for family-centered neonatal care. Pediatrics

1993;92:643650.
8. Johnson BH, Jeppson ES, Redburn L. Caring for children and families:
guidelines for hospitals, 1st ed. Bethesda: Association for the Care of Children's
Health, 1992.
9. Stepanek JS. Moving beyond the medical/technical: analysis and discussion of

psychosocial practices in pediatric hospitals, 1st ed. Bethesda: Association for
the Care of Children's Health, 1995.
10. Zaichkin J. Newborn intensive care: what every parent needs to know. 2nd
ed. Petaluma, CA: NICU INK Book, 2002.
11. Linden DW, Paroli ET, Doron MW. Preemies: the essential guide for parents
of premature babies. New York: Pocket Books, 2000.
12. Kenner C, Bagwell GA, Torok LS. Assessment and management in the
transition to home. In: Kenner C, Lott JW, eds. Comprehensive neonatal
nursing: a physiologic perspective, 3rd ed. St. Louis: WB Saunders Co., 2003.
13. Durfor SL, Murphy-Ratcliff M. Home- and community-based care. In: Kenner
C, Lott JW, eds. Comprehensive neonatal nursing: a physiologic perspective, 3rd
ed. St. Louis: WB Saunders Co., 2003.
14. Chassin MR, Galvin RW. The urgent need to improve health care quality.
Institute of Medicine National Roundtable on Health Care Quality. JAMA
1998;280:10001005.
15. Sinclair JC, Bracken MB, Horbar JD, et al. Introduction to neonatal
systematic reviews. Pediatrics 1997;100:892895.
16. Lohr KN, ed. Medicare: a strategy for quality assurance. Washington, DC:
17. Institute of Medicine. Crossing the quality chasm: a new health system for
the 21st century. Washington, DC: National Academy Press, 2001.
18. Institute of Medicine. To err Is human: building a safer health system.

Washington, DC: National Academy Press, 1999.
19. Leape LL. Reporting of adverse events. N Engl J Med 2002;347: 16331638.
20. Suresh G, Horbar J, Plsek P, et al. Voluntary anonymous reporting of

medical errors for neonatal intensive care. Pediatric Res 2003;53:426A.
21. The Vermont-Oxford Trials Network: very low birth weight outcomes for
1990. Investigators of the Vermont-Oxford Trials Network Database Project.
Pediatrics 1993;91:540545.
22. Horbar JD, Plsek PE, Leahy K. NIC/Q 2000: establishing habits for
improvement in neonatal intensive care units. Pediatrics 2003;111:e397-e410.
23. Donovan EF, Ehrenkranz RA, Shankaran S, et al. Outcomes of very low birth
weight twins cared for in the National Institute of Child Health and Human
Development Neonatal Research Network's intensive care units. Am J Obstet
Gynecol 1998;179:742749.
24. Lemons JA, Bauer CR, Oh W, et al. Very low birth weight outcomes of the
National Institute of Child health and human development neonatal research
network, January 1995 through December 1996. NICHD Neonatal Research
Network. Pediatrics 2001;107:E1.
25. Vohr BR, Wright LL, Dusick AM, et al. Neurodevelopmental and functional
outcomes of extremely low birth weight infants in the National Institute of Child
Health and Human Development Neonatal Research Network, 19931994.
Pediatrics 2000;105: 12161226.
26. Fernandez CV, Rees EP. Pain management in Canadian level 3 neonatal
intensive care units. CMAJ 1994;150:499504.
27. Lee SK, McMillan DD, Ohlsson A, et al. Variations in practice and outcomes
in the Canadian NICU network: 19961997. Pediatrics 2000;106:1070-1079.
28. Sankaran K, Chien LY, Walker R, et al. Variations in mortality rates among
Canadian neonatal intensive care units. CMAJ 2002;166: 173178.
29. Risk adjusted and population based studies of the outcome for high risk
infants in Scotland and Australia. International Neonatal Network, Scottish
Neonatal Consultants, Nurses Collaborative Study Group. Arch Dis Child Fetal
Neonatal Ed 2000;82:F118-F123.
30. Sutton L, Bajuk B. Population-based study of infants born at less than 28
weeks' gestation in New South Wales, Australia, in 19923. New South Wales
Neonatal Intensive Care Unit Study Group. Paediatr Perinat Epidemiol
1999;13:288301.
P.88
31. Zullini MT, Bonati M, Sanvito E. Survival at nine neonatal intensive care
units in Sao Paulo, Brazil. Paulista Collaborative Group on Neonatal Care. Rev
Panam Salud Publica 1997;2:303309.
32. Horbar JD, Carpenter JH, Burlington, VT: Vermont Oxford Network, 2002.
33. Brodie SB, Sands KE, Gray JE, et al. Occurrence of nosocomial bloodstream
infections in six neonatal intensive care units. Pediatr Infect Dis J 2000;19:56
65.
34. Bednarek FJ, Weisberger S, Richardson DK, et al. Variations in blood

transfusions among newborn intensive care units. SNAP II Study Group. J
Pediatr 1998;133:601607.
35. Avery ME, Tooley WH, Keller JB, et al. Is chronic lung disease in low birth
weight infants preventable? A survey of eight centers. Pediatrics 1987;79:2630.
36. National Nosocomial Infections Surveillance (NNIS) System report, data

summary from January 1992-June 2001, issued August 2001. Am J Infect
Control 2001;29:404421.
37. Kilo CM. Improving care through collaboration. Pediatrics 1999;103:384

393.
38. Nolan TW, Schall MW, Roessner J. Reducing delays and waiting times
throughout the healthcare system. Boston: Institute for Healthcare
Improvement, 1996.
39. Weiss KB, Mendoza G, Schall MW, et al. Improving asthma care in children
and adults. Boston: Institute for Healthcare Improvement, 1997.
40. Rainey TG, Kabcenell A, Berwick DM, et al. Reducing costs and improving
outcomes in adult intensive care. Boston: Institute for Healthcare Improvement,
1996.
41. Sackett DL, Rosenberg WM, Gray JA, et al. Evidence based medicine: what it
is and what it isn't. BMJ 1996;312:7172.
42. Klassen TP, Wiebe N, Russell K, et al. Abstracts of randomized controlled

trials presented at the society for pediatric research meeting: an example of
publication bias. Arch Pediatr Adolesc Med 2002;156:474479.
43. Soll RF, Andruscavage L. The principles and practice of evidence-based

neonatology. Pediatrics 1999;103(1 Suppl E):215224.
44. Muir Gray JA. Evidence-based health care: how to make health policy and
management decisions. New York and London: Churchill Livingstone, 1997.
45. West S, King V, Carey T, et al. Systems to rate the strength of scientific
evidence. Rockville, MD: Agency for Healthcare Research and Quality, 2002.
46. The periodic health examination. Canadian Task Force on the Periodic Health
Examination. Can Med Assoc J 1979;121:1193- 1254.
47. Strom KL. Quality improvement interventions: what works? J Healthc Qual
2001;23:414.
48. Kilbride HW, Powers R, Wirtschafter DD, et al. Evaluation and development
of potentially better practices to prevent neonatal nosocomial bacteremia.
Pediatrics 2003;111:e504-e518.
49. Baker GR, King H, MacDonald JL, et al. Using organizational assessment
surveys for improvement in neonatal intensive care. Pediatrics 2003;111:e419e425.
50. Baker GR, King H, MacDonald JL, et al. Using organizational assessment
surveys for improvement in neonatal intensive care. Pediatrics 2003;111:e419.
51. O'Connor E, Fiol CM. In: Lowery JE, ed. Culture shift: a leader's guide to
managing change in health care. Chicago: American Hospital Publishing, Inc.,
1997:39-.
52. Plsek PE. Tutorial: introduction to control charts. Qual Manag Health Care
1992;1:6574.
53. Carey RG, Lloyd RC. Measuring quality improvement in healthcare: a guide
to statistical process control applications. New York: Quality Resources, 1995.
54. Langley GJ, Nolan KM, Norman CL, et al. The improvement guide: a practical
approach to enhancing organizational performance. San Francisco: Jossey-Bass,
1996.
55. Nolan TW, Schall MW, Roessner J. Reducing delays and waiting times
throughout the healthcare system. Boston: Institute for Healthcare
Improvement, 1996.
56. Plsek PE. Quality improvement methods in clinical medicine. Pediatrics

1999;103:203214.
57. Shewhart WA. Economic control of quality of manufactured product. New

York: Van Nostrand, 1931.
58. Deming WE. Out of the crisis. Cambridge, MA: MIT Press, 1986.
59. Kilbride HW, Wirtschafter DD, Powers RJ, et al. Implementation of evidencebased potentially better practices to decrease nosocomial infections. Pediatrics
2003;111:e519-e533.
Chapter 8
Law, Quality Assurance and Risk Management
in the Practice of Neonatology
Harold M. Ginzburg
Mhairi G. MacDonald
BASIC LEGAL AND REGULATORY CONCEPTS

A brief historical overview of the transition from the past to present will help explain some of the
current legal and social difficulties within the practice of medicine.
Health care professionals have always been subject to punishment should their medical skills or
judgment fall below a community standard. Reports from the Middle Ages in England indicate that
compensation was paid when patients were injured. In other nations during the Middle Ages the
health care provider, if found responsible for poor medical care, was deformed in the same
manner as the injury that he caused. Physicians were not motivated to treat patients with
complicated illnesses unless the patient and his or her extended family clearly understood that
treatment would be palliative. Although litigation has replaced amputation in modem western
societies, health care providers and institutions still practice under the specter of retribution and
public criticism.
Prior to the middle of the 20th century members of the healing profession functioned as an
integral part of their local community. Since the 1940s there has been a progressive separation of
health care providers from the communities they serve; this process has been accelerated by the
advent of highly subspecialized and expensive intensive care. Neonatologists function in a crisis
environment with little or no prior knowledge of their patients' family unit.
Isolation of patients from health care providers, except in times of crisis, can lead to poor or
limited communication and unrealistic expectations. The family expects success.
Litigation may be initiated when negligence has occurred or is perceived to have occurred. A lack
of good communication skills and empathy can be as self destructive as a faulty knowledge base,
performing unacceptably, or functioning in an impaired manner. Thorough knowledge and
understanding of one's medical subspecialty is insufficient to prevent involvement in malpractice
litigation. Impersonal institutional and individual behaviors often precipitate a lawsuit. The failure
to communicate and document patient and health care system interactions may place a health
care system or provider in an untenable position when his or her actions are reviewed in an
arbitration, mediation, or courtroom environment.
Health care providers are never able to succeed all the time. Sometimes the nature and extent of
the disease is too severe, sometimes the course of treatment is too risky or aggressive to be
tolerated by the patient, and sometimes errors of skill and judgment are made. Relatively recently
the medical community and their legal advisors have come to the realization that explaining to a
patient and family what went wrong in language that they can understand is not only the right
thing to do but also appears to decrease the incidence of litigation (1,2). As part of its
accreditation process, the Joint Commission on Accreditation of Healthcare Organizations (JCAHO)
now requires health care organizations to disclose unanticipated injuries to the affected patients,
investigate their root causes, and take action to prevent their recurrence (3).
Responsibility and Liability

The mundane aspects of medical care and treatment, such as scheduling of appointments,
documentation of procedures, and understanding of the federal, state, and local
P.90
guidelines, procedures, policies, regulations, and laws, are frequently the basis for confrontations
between members of the medical and legal professions.
The vast majority of medical education pertains to understanding basic sciences and providing
clinical services. Little formal attention is given to the myriad of governmental policies,
procedures, and regulations that control all aspects of health care delivery.
Medical care is a contract between the health care professional and the patient. In almost all
instances, if the patient is unable, because of age or illness, to render informed educated consent,
then others must provide such consent. Thus, as a patient is examined or interviewed, law and
medicine become intertwined.
The basic legal considerations relating the care and treatment of any patient, and particularly, a
neonate, flow from the following four concepts:
The duty to act. When does the health care professional-patient or health care facilitypatient relationship commence?
Knowledge and application of hospital policies and local, state, and federal mandates. What
resources are available to facilitate information transfer to health care facilities and service
providers?
Responsibility and accountability of the health care provider (individual or organization) to

provide adequate care and treatment. Who is responsible for the decisions made in the
provision of health care? Who monitors the quality of the services provided? Who ensures
that the services provided are consistent with hospital policies and local, state, and federal
mandates?
Information transfer to patients and their families or guardians. Who obtains educated
informed consent, in what manner, and with what documentation? Who is responsible for
providing ongoing medical information to the families or guardians of neonates and
ensuring that the information, and the implications of the information, is understood? State
and federal legislation, such as the Health Insurance Portability and Account-ability Act of
1996 (HIPAA) (4), do not set standards for the manner in which a clinician may or should
communicate with a patient, family member or significant other.
Defensive Medicine
Defensive medicine has become a medical term of art. However, it can connote a thoughtful
systematic approach to health care rather than the excessive ordering of investigatory studies
because of anticipatory fear of litigation for malpractice. Medical malpractice lawsuits are based
on the principle of negligence. Negligence implies some wrongful act of commission or omission
(5). The essence of negligence is unreasonableness.
Due Care
Due care is simply reasonable conduct (6). In order for negligence to be demonstrated in a
courtroom, the injured person/plaintiff must demonstrate that (a) there was a legal duty owed to
him; (b) there was a breach of that duty (a deviation from the accepted standard of care); (c) as
a result of the duty and the breach thereof, damages or an injury occurred; and (d) the damages
or injury can be determined to have been caused by, or shown to have flowed from, the care or
lack of care provided by the health care provider and/or organization responsible for the
environment in which the health care was provided.
Quality assurance and risk management aspects of medical care are relatively recent innovations
designed to improve patient care and outcome; they are discussed in greater detail in the second
section of this chapter. Quality assurance activities accept the legal and medical position that a
health care provider owes a duty to the patient to provide reasonable medical care, consistent
with available resources. There are inherent, irreducible risks in the delivery of medical care and
treatment, and quality assurance and risk management assessments are designed to identify and
limit the risks. There are always risks in a medical intervention, and there are always risks in not
rendering a medical intervention. The balance of relative risks needs to be understood by both the
health care provider and the patient and/or parent/guardian.
The Duty to Act

The duty to act is determined when the health care professional- or organization-patient
relationship commences (7). A duty is a legal and ethical responsibility. There is no legal duty
under most circumstances for a health care provider or institution to accept a patient for care
unless they hold themselves out as providing emergency care or they are required to do so by
law, regulation, or contract. If a medical center, hospital, or physician represents itself to the
public as a source of emergency medical care and the community has come to expect such care,
then a patient cannot be arbitrarily denied such services (8). Once a health care service is
initiated, a health care provider/institution-patient relationship exists, a duty is created, and there
is then a legal and moral obligation not to abandon the patient. Further, the care that is being
provided must be adequate under the circumstances. A moral and legal obligation attaches, which
precludes abandonment or dumping of the patient (9). A referring hospital transferring an infant
to another institution for further care is not perceived as abandoning the patient, as long as the
reason for transfer is medical and not financial. The senior medical person responsible for the
transport, whether stationed at a hospital or directly providing patient care during transport, is
deemed to be supervising the health care until the transport team transfers care to the clinical
staff at the receiving referral medical facility. The transport team's legal and ethical duty to the
patient exceeds that of other parties, such as the referring or referral hospital.
The complexity of health care responsibility and liability increased rapidly during the 20th century.
Public health
P.91
care facilities have existed since the Middle Ages. Commencing in the 13th century, the Hotel Dieu
in Paris provided indigent care for many centuries. In the United States, city, municipal, state,
and federal public hospitals provided care for those unable to obtain it elsewhere, and the
physicians and hospitals were not held liable for the outcome of care that was provided free of
charge. The doctrine of charitable immunity protected hospitals from legal liability if medical
negligence occurred within their boundaries. However, an individual's inability to pay for medical
care no longer affects his or her ability to demand and receive services that are commensurate
with those provided to patients who pay for their care directly or through third-party payment
systems. Thus, providing care to patients who are unable to pay no longer protects a health care
provider or medical institution from liability for negligence or malpractice. Physicians, other health
providers, suppliers, and manufacturers of equipment, medical devices, and medicines can now all
be sued for negligence and be held individually or jointly liable for their own actions, those that
they supervise, and those that are performed by members of their health care team.
Licensure; Interstate-International Practice

In the United States, health care providers (physicians, nurses, emergency medical technicians,
etc.) may be licensed in more than one state. They must be licensed in the state in which they
maintain their primary office or place of employment. The authorities in most, but not all, states
are not concerned about whether a health care provider who enters the state solely to transport a
patient to another health care facility is licensed in that state; however, they are concerned that
the individuals involved in the transport are competent to perform their job.
An individual who enters a state, regardless of the reason, will be subject to that state's laws. If a
driver is involved in an accident, he or she is subject to the laws of the state in which the accident
occurred, not to those of the state that issued the driving license; this legal principal also applies
to medical transport vehicle operators (10). Failure to obtain informed consent may result in
litigation in the state in which it was inadequately obtained or in the state to which the patient
was transferred.
Patients and their guardians may initiate medical malpractice litigation in the state in which they
reside, the state in which the alleged negligence occurred, the state in which the hospital is
located, or the state in which the physician resides. If the patient/plaintiff can show that his or her
residence is in a different state from that of the defendant/hospital and defendant/health care
provider, then the plaintiff can commence the litigation in a federal court because the matter
involves diversity of jurisdiction, i.e., opposing parties are located in two or more states. The
defendant can request that the matter be removed to the federal court system for a similar
reason (11). Most plaintiffs prefer state courts, especially if the defendant is from a different
state. Some state courts are known for their large awards to plaintiffs, whereas others are known
to be more sympathetic to defendants/health care providers.
Telemedicine
Recent advances in telephone-linked care (TLC) or telemedicine (see also Chapter 5) have
initiated new questions regarding the practice of medicine across state lines (12). TLC has been
applied as a supplement to direct patient care, to monitor patient progress, and as a service
expander for specialized medical expertise and technology (e.g., the interpretation of neonatal
radiologic or cardiologic films and tracings). Landwehr and associates (13) demonstrated the
feasibility of telesonography for the interpretation of fetal anatomic scans from a remote location.
Lewis and Moir (14) in Scotland and Landquist (15) in Finland demonstrated that telemedicine is
an international technology.
TLC has been practiced for more than 30 years (in its simplest form, it includes giving advice over
the telephone). In 1996 the U.S. Congress passed the Telecommunications Reform Act, which
required a study of patient safety, efficacy, and the quality of services (16). The neonatologist has
the opportunity to engage in telehealth, which includes consultation, transportation, and
interpretation of radiographic, cardiologic, and other data, as well as professional education,
community health education, public health, and administration of health services. The American
Medical Association and American Telemedicine Association have urged medical specialty societies
to develop appropriate practice standards. Managed care organizations have begun to embrace
telemedicine. Louisiana, in 1995, became the first state to enact legislation dealing with
telemedicine reimbursement (17) that specifies a certain reimbursement rate for physicians at the
originating site and includes language prohibiting insurance carriers from discriminating against
telemedicine as a medium for delivering health care services.
California in 1996 (18), Oklahoma in 1997 (19), Texas in 1997 (20), and Kentucky in 2000 (21)
have also passed telemedicine legislation. However, at the present time, issues relating to crossstate licensure are perceived to be potential barriers to the expansion of telemedicine, especially
now that reimbursement is possible. States license physicians and other health care providers
within their boundaries, but the federal government has the authority to prepare national
licensure standards as they relate to national programs such as Medicaid and Medicare. In the
future, there may be alternative approaches to licensure (22). Regardless of the end result of the
issues surrounding telemedicine, neonatologists increasingly cross state and international
boundaries and need to appreciate that the laws of political jurisdictions other than their home
state may significantly impact the manner in which they practice.
The U.S. Food and Drug Administration (FDA) has become involved in telehealth activities on the
Internet.
P.92
Over the past few years, some Web sites have offered illegal drugs or prescription drugs based on
questionnaires rather than a face-to-face examination by a licensed health care practitioner.
Some offshore sites offer prescription drugs without any prescription or medical consultation. The
FDA works with the U.S. National Association of Boards of Pharmacy, which created a program in
1999 called Verified Internet Pharmacy Practice Sites to provide the consumer with the ability to
verify the safety of medications being sold over the Internet (23). A number of U.S. federal and
state regulatory agencies are working together to address health-related consumer problems on
the Internet. They include state health authorities, FDA, Justice Depart-ment, and Federal Trade
Commission. The Federal Trade Commission plays a key oversight and enforcement role in
Internet commerce.
Medical Torts and Contracts

The legal system is divided into two broad areas. Civil litigation is based on the need to correct or
remedy a wrong between one individual (corporation or partnership) and another. Criminal
prosecution is instituted to correct a wrong against the community. In a civil litigation matter (or
case), the plaintiff is the party bringing the lawsuit and alleging the wrong; the lawsuit is filed
against the party (defendant) who is accused of causing the damage. In a criminal prosecution,
the plaintiff is the government (local, state, federal) alleging that the community has been
harmed by the action or inaction of a party (also known as the defendant).
Civil matters resulting in litigation generally are either contract disputes or torts. A contract
dispute occurs when two or more parties have entered into an agreement and one or more parties
believe that the terms and conditions of the agreement, either an oral or written contract, have
not been met. It is important to appreciate that, in a court; most oral contracts have the same
weight as written contracts.
A personal tort is an injury to a person or his or her reputation or feelings that directly results
from a violation of a duty owed to the plaintiff (in medical malpractice cases, this is usually the
patient) and produces damage. The remedy in any civil matter, after the nature and extent of the
damages have been proved to the court (a judge with or without a jury), is determined by a
preponderance of the evidence (more than 50.01%). Thus, to a medical degree of certainty
means that it is more likely than not, and it is that standard upon which there are usually
monetary damages awarded if the matter is found in favor of the plaintiff. In most instances in
the United States, each side pays for its own legal services, regardless of the outcome of the case.
Battery
Battery is a tort; it is an intentional and volitional act without consent which results in touching
that causes harm (e.g., the touching of a patient's body without consent). A technical battery can
occur when there is no actual harm but touching occurred without consent. Patient care, even
with a beneficial outcome but without informed consent, may be considered battery.
Plaintiffs may sue for an injury that occurred as a result of negligence or a tort (physical or mental
harm), or both. Because the criminal court usually will not award monetary damages to the victim
of a crime and because the standard of proof for conviction is beyond a reasonable
doubt (quantitatively, this can be conceptualized as at least 95% certain), plaintiffs usually prefer
to sue for injuries from a tort in civil court. In civil litigation, monetary damages may be awarded,
and if the injury was determined to be egregious, punitive damages also can be assessed against
the defendant. The standard of proof in civil litigation is the preponderance of the evidence or
the more likely than not standard; it is a superiority of weight test that requires that for the
plaintiff to be successful, 50.01% of the evidence must weigh in his or her favor (24). Thus, the
preponderance of the evidence rule is a threshold test (25). In general, either the plaintiff proves
that the damages were more likely to have been caused by the defendant agent than by any
other source and is; therefore, entitled to full compensation, or he or she fails to meet the burden
of proof and is entitled to nothing (26).
Professional Negligence
Negligence is conduct, and not a state of mind (27), involves an unreasonably great risk of
causing damage (27) and is conduct which falls below the standard established by the law for
the protection of others against unreasonable risk of harm (28,29).
Professional negligence, or medical malpractice, is a special instance of negligence. The medical
profession is held to a specific minimum level of performance based on the possession, or claim of
possession, of special knowledge or skills that have been accrued through specialized education
and training.
Ely and associates (30) found that when family physicians recalled memorable errors, the
majority fell into the following categories: physician distracters (hurried or overburdened),
process of care factors (premature closure of the diagnostic process), patient-related factors
(misleading normal results), and physician factors (lack of knowledge, inadequately aggressive
patient management). Under-standing the common causes of errors alerts the practitioner to
situations when errors are most likely to occur.
The Elements of a Malpractice Case

To establish a prima facie medical malpractice case (one that still appears obvious after reviewing
the medical evidence), the patient/plaintiff must demonstrate (Fig. 8-1): that (a) there is a duty
on the part of the defendant/health care provider and/or defendant/health care facility to the
patient/plaintiff, (b) the defendant failed to conform his or her conduct to the requisite standard
of care required by the relationship, and (c) an injury to that patient/plaintiff resulted from that
failure (31).
Figure 8-1 The elements of negligence. There must be an unbroken chain for successful
litigation. If any link is not proved, the plaintiff will lose the case. (From Ginzburg HM. Legal
issues in patient transport. In: MacDonald MG, ed. Emergency transport of the perinatal patient.
Philadelphia: Little, Brown and Company, 1989:163, with permission.)
P.93
Generally, in order for the plaintiff to establish a claim of medical malpractice, the plaintiff must
establish by medical expert testimony (a) what the applicable standard of care is, (b) how the
defendant breached or violated that standard of care, and (c) that the breach or violation (also
referred to as the negligence) was the proximate cause of the injury.
A medical malpractice action can only proceed if the court determines that there is a genuine
issue of material fact and if damages are quantifiable (e.g., the future costs of treatment,
economic lost value of productive activities, etc.).
The most difficult element to prove is whether or not the standard of care was adequate. The
plaintiff usually must provide expert witnesses to establish what a prudent health care provider in
similar circumstances might have done. A conspiracy of silence may have existed in prior years
in the United States, but today there are many experts willing to testify anywhere about
anything. However, in Japan and other nations, it is difficult to find a local or even national expert
to testify in medical malpractice cases (32).
More than 100 years ago in Massachusetts, it was held that a physician in a small town was
bound to have only the skill that physicians of ordinary ability and skill in similar localities
possessed. The court believed that a small-town physician should not be expected to have the
skill of surgeons practicing a specialty in a large city (33). It was also held that a physician was
required to use only ordinary skill and diligence, the average of that possessed by the profession
as a body, and not by the thoroughly educated (34). However, a physician is now not excused for
failing to keep himself or herself informed of medical progress. State requirements for continuing
education and the effects of telemedicine consultations and educational programs essentially have
removed clinicians' ability to say that they are too busy or so geographically inaccessible as to be
precluded from keeping current with new treatments and new understanding of the illnesses that
affect their patients.
Courts admit medical evidence based upon rules of evidence. In 1993, the U.S. Supreme Court in
Daubert v Merrill Dow Pharmaceuticals, Inc. ruled that the Federal Rules of Evidence standards for
acceptance of evidence would be used (35). This was an attempt to remove junk science from
distracting the jury. The court held that scientific (medical) evidence had to be grounded in
relevant scientific principles. The four criteria the court established are: (a) whether the theory or
technique has been tested; (b) whether the theory or technique has been subjected to peer
review and publication; (c) the known or potential rate of error of the method used and the
existence and maintenance of standards controlling the technique's operation; and (d) whether
the theory or method has been generally accepted by the scientific community. Thus, publication
in a peer-reviewed or peer-refereed journal was not the only qualification for acceptance of
evidence in a courtroom. The district (trial) court judges have the latitude to permit or exclude
experts, based on the perceived scientific merit of the information they intend to provide to the
court and, thus, to the jury. The fundamental issue for a clinician is not an understanding of the
rules of evidence and the workings of the civil justice system but practicing medicine and acting in
a professional manner, as documented in a patient's medical record.
Many states have medical peer-review panels in place. In these states, before a medical
malpractice case may be heard in a court, the facts of the case are presented to the review panel
on behalf of both the plaintiff and defendant. The medical facts often are buttressed by the
opinions of retained medical experts for both sides. In some states the medical review panel is
comprised of attorneys and physicians; in other states the medical review panel is chaired by an
attorney and comprised of physicians in the same or similar medical specialty as the physician
being accused of having committed malpractice. Even when there is a finding for the defendant by
the medical review panel, the plaintiff may continue litigation in the local court. However, the
findings of the medical review panel are admissible on behalf of either the plaintiffs or defendants.
Res ipsa loquitur

There are circumstances in which no expert witness is required to corroborate the findings of
negligence. The doctrine of res ipsa loquitur essentially means that the thing speaks for itself.
Under such circumstances, the negligence is inferred from the act itself, i.e., proof from
circumstantial evidence. In the classic case Ybarra v Spangard, a patient was well prior to being
anesthetized for an appendectomy, and when he woke up he had an injury to his arm (36).
Clearly, he could not determine how his arm was injured; the operating room and recovery room
staff either could not or would not explain the etiology of the injury. The court found for the
injured plaintiff without the introduction of any expert witnesses because (a) the plaintiff had not
done anything that in any way could have contributed to the injury, (b) the injury could not have
occurred unless someone
P.94
was negligent, and (c) the instrumentalities (hospital staff and physicians) that allegedly caused
the injury were at all times under the control of the defendant hospital.
Informed Consent
Informed consent requires that sound, reasonable, comprehensible, and relevant information be
provided by a health care professional to a competent individual (patient or guardian) for the
purpose of eliciting a voluntary and educated decision by that patient (or guardian) about the
advisability of permitting one course of clinical action as opposed to another (28). Physicians and
other health care providers are held to have a fiduciary duty to their patients. Such a duty exists
when one individual relies on another because of the unequal possession of information. The
failure to obtain proper informed consent may result in the defendant/physician or defendant/
hospital being sued for battery in some states or for negligence in others.
According to the battery theory, the defendant is to be held liable if any deliberate (not careless
or accidental) action resulted in physical contact. The contact must have occurred under
circumstances in which the plaintiff/ patient did not provide either express or implied permission
and the defendant/health care provider knew or should have known that the action was
unauthorized. If the scope of consent obtained from the patient is exceeded, a claim of battery is
proper. The plaintiff in Mohr v Williams consented to have surgery performed on her right ear
(37). During the procedure, the surgeon determined that the right ear was not sufficiently
diseased to require surgery, but the left ear required surgery. Because the patient was already
anesthetized, the surgeon performed the operation. The operation was a success, but the patient
successfully sued for battery. The court held that there was no informed consent for an operation
to the left ear. Thus, it is not necessary for injury to occur for damages to be awarded;
demonstration that there was unauthorized touching is sufficient. In this instance, the court found
that there was no medical emergency that would have threatened the plaintiff/patient if the
surgery had not immediately commenced. If there were evidence of a medical emergency, the
court's decision might have been significantly different.
Failure to specifically identify the risks that accompany a surgical procedure also can result in a
successful claim of battery. In Canterbury v Spence, the plaintiff/patient successfully proved that
he was not informed of the risks attendant to the surgical procedure and that had he known them
he would not have given permission (38). The court held that the physician has a duty to disclose
all reasonable risks of a surgical procedure, and because he failed to perform that duty, the court
held him liable for damages to the patient. The court noted that the concept of informed consent
might be more appropriately replaced with the concept of educated consent. The court also
articulated an objective standard that could be used in legal cases involving informed consent.
This objective standard is based on what a reasonable person in circumstances similar to that of
the patient would have decided if he or she had been provided with an adequate amount of
information. Therefore, the central issue in a medical battery is whether an educated, effective, or
valid consent was given for the procedure that actually was performed.
A physician is not required to disclose every possible risk to a patient for fear of being guilty of
battery (39). The court in Cooper v Roberts held that [t]he physician is bound to disclose only
those risks which a reasonable man would consider material to his decision whether or not to
undergo treatment (40). Thus, the court stated that such a standard creates no unreasonable
burden for the physician. However, the physician must disclose risks that are material and
feasible alternatives that are available. The information should be provided in a language and
manner that reflects the emotional and educational status of the patient or, when the patient is a
neonate, the parents. In Davis v Wyeth, the court held that any medical complication or risk that
has a probability of greater than 1:1,000 should be included in the informed consent (41).
When a therapeutic procedure is for the benefit of a minor, the decision to proceed usually
belongs to the parent or legal guardian. The failure of the parent to consent to blood transfusions
(even if the refusal is based on sincere religious convictions) or other now routine procedures for
a small child that are clearly medically indicated and required for the maintenance of life can be
overridden by the physician and/or hospital petitioning the court for the appointment of a
temporary legal guardian (42).
The unavailability of a parent in a life-threatening circumstance should not preclude therapeutic
action. Just as informed consent is imputed to an unconscious accident victim who has a lifethreatening condition that requires surgery, such rational behavior can be imputed to the absent
parent in the case of a sick neonate. However, in such circumstances if time permits, detailed
documentation and consultation with the hospital administration is recommended.
Informed consent in neonatal/perinatal medicine is not an empty gesture to reduce liability, but
rather an interaction with the physician that helps parents to become full partners in decision
making. Informed consent documents are intended to support decision makers in their choices,
rather than to merely have them ratify decisions already made (43). Informed consent documents
need to be routinely reviewed to determine that the reading level required to understand them is
consistent with the educational and cultural experiences of those being asked to read,
understand, and sign the forms (44).

The informed consent process can be extremely complex, with several legal gray areas. For
example, are maternal rights any more definitive in making critical decisions for a fetus or
neonate than paternal rights? Conflicts can arise even when the putative or alleged father is not
the legal spouse of the mother. Emergency hearings in front of local judges may be required to
resolve conflicting opinions, especially when the decision of one parent may lead,
P.95
to a medical degree of certainty (more likely than not), to the death of the infant.
It is one of the ironies of the law, in most states, that an unwed teenage mother has the ultimate
legal responsibility for the care of her child, unless the court is petitioned to appoint an alternative
guardian. In many states the live birth of a child, regardless of the age of the mother, results in
the mother being declared an emancipated minor. In contrast, a non-pregnant teenager, living at
home and attending school, does not have the legal right to make decisions about many aspects
of her own medical care.
The disclosure of risks in the informed consent process tends to underscore a parent's sense of
helplessness and to portray the physician as somewhat helpless as well. The powerlessness of the
parents and their wish that the physician be omnipotent creates unrealistic expectations for the
outcomes of procedures and treatment. Gutheil and associates (45) suggest that the physician
acknowledge the parent's wish for certainty and substitute the mystical with a physician-parent
alliance in which uncertainty is accepted.
Advanced Care Planning

Decision making in life and death situations is not easy. Advanced care planning efforts initially
evolved in the care of the elderly (46). Advanced care planning, or contingency medical care
planning, is no longer generally reserved for adults. The American Academy of Pediatrics
Committee of Bioethics developed a policy statement about parental permission and informed
consent (47). A distinction is made in this public policy statement and in case law in the United
States between emergency medical treatment, life support efforts, and elective surgical
procedures, such as circumcision or removing a kidney from one child to aid a sibling (48,49).
Neonates need others to make decisions about their treatment and viability. Their treatment in a
neonatal intensive care unit (NICU) is also never a series of leisurely scheduled elective
procedures. There must be a designated person to whom the health care providers relate and who
is held to render decisions about treatment. When potential conflicts in decision making are
recognized, a conference among the concerned individuals, with representation from the hospital
administration, may be a worthwhile endeavor to clarify who actually has the final decisionmaking authority. There may be times in the treatment of a patient when one person has to
render an immediate decision, even if that decision is not a consensus. Decision-making issues
are often confounded when the mother of the neonate is an adolescent herself and not wed to the
father.
While all 50 states in the United States and the District of Columbia have passed legislation on
advance directives, reinforcing the fact that adherence to such directives is mandatory rather than
optional continues to be problematic (46). The majority of states do place restrictions on proxy
decision making. If identification of the responsible party for decision making is not clear or keeps
shifting, legal consultation, initiated by the health care providers, is recommended. Most
jurisdictions have the ability to hold emergency judicial proceedings when an impasse has been
reached and a medical decision must be made before irreversible damage or death to the patient
occurs (42).
Medical Records
As evidenced by HIPAA (4) and state regulations, medical records are legal documents. Medical
and hospital records are designed to be a contemporaneous record of the available clinical
information and medical and other decisions that flow from the clinical information and
interactions with the patient's significant others. Records provide an opportunity for adequate
documentation. Documentation is the key to management of patients, and it is the key to
protecting physicians against malpractice litigation, especially in the instances in which the
patients have difficult and complex clinical presentations. The course of treatment and meeting or
failing to meet therapeutic goals should be noted in the hospital chart. Treatment options,
including the option of no treatment when pertinent, should be explained to the patient's family
and, if necessary, others potentially involved in the decision-making process; these interactions
should be documented in the patient's chart. The patient's family's understanding, or lack thereof,
of the various treatment options also should be noted, especially if there are divergent views
among family members. Ultimately one family member or guardian has to be acknowledged by
the family and health care providers as the decision maker. Identifying such an individual in the
medical record will facilitate treatment decisions and posthospital treatment care and
management. Such documentation may preclude the need for judicial intervention.
Adequate medical records will document that the risks of a given procedure have been shared
with the patient's decision makers. A record that indicates that specific known adverse side
effects, or rare but serious untoward events, were discussed with a patient's family members or
guardians helps protect the clinician should one of these untoward events actually occur. No
medical procedure is risk free, and although families may be informed of the relative risks of the
various procedures and pharmacologic interventions, the stress of the moment may shorten their
attention span, concentration, and recall. For example, prescriptions at the time of discharge are
often for a limited period of time, with follow-up care being provided in either a hospital
outpatient, clinic, or private clinical environment. Prescriptions must be written clearly, identifying
the patient, date, dose, dose schedule, and route of administration. The parent or guardian to
whom the prescription is provided needs to understand why the medication is being prescribed,
adverse side effects, therapeutic effects of the medication, and consequences to the infant if the
medication is not provided. Medical record documentation, including discharge instructions and
prescriptions, at least provide a contemporaneous record of what information actually was
provided.
P.96
Medical records provide the basis for reimbursement of the costs for patient care and treatment.
The severity of a condition, justification for laboratory and other investigations, need for
consultations, and manner in which the consultative advice is incorporated into patient care
should be present within a patient's medical record. In the spring of 1998 the U.S. Justice
Department announced the hiring of 250 Federal Bureau of Investigation (FBI) agents for the
purpose of investigating Medicare and Medicaid fraud. The U.S. federal government continues to
pay increasing attention to the issue of fraudulent billing for health care services whether the care
is provided directly or via telemedicine. The FBI estimates that 10% of the money paid out for
health care services under the Medicaid program is the result of fraudulent billing (50). Not all the
money paid for fraudulent services is recovered. However, in 2002, $1.6 billion was collected in
connection with health care fraud cases and matters (51). The absence of consistent,
comprehensive reimbursement policies is frequently cited as one of the most serious obstacles to
total integration of telemedicine into health care practice. This lack of an overall telemedicine
reimbursement policy reflects the multiplicity of payment sources and policies within the current U.
S. health care system.
Adequate documentation facilitates medical audit, permits those who prepare invoices for
reimbursement or payment to justify the categories or International Classifica-tion of Diseases
codes placed on the universal billing forms (sometimes identified as HCFA-1500 forms), and
prevents errors that may result in the appearance of fraud (52).
Medical Confidentiality of Oral and Handwritten

Communications
The Hippocratic oath states, in part, Whatever, in connection with my professional practice or not
in connection with it, I see or hear, in the life of men, which ought not to be spoken of abroad, I
will not divulge, as reckoning that all should be kept secret (53). This oath, taken by many
physicians upon graduating from medical school, has been codified by state and federal law. The
confidentiality of the information obtained by the physician is considered to be privileged (54).
The privilege, however, belongs to the patient or his or her guardian; it does not belong to the
physician or other health care provider. A court may order a health care provider to breach
medical confidentiality; legislation may require it.
As the delivery of medical care becomes more complex, individuals not generally thought of as
direct health care providers are requiring access to confidential medical information. The
legitimate needs, for example, of physical therapists, occupational therapists, hospital
administrative personnel, and medical insurance personnel may appear to conflict with the
principle that medical information is confidential and should be restricted only to physicians. The
ethical basis for confidentiality derives from the concept that assurance of confidentiality
encourages patients to seek the medical care that they require and to be candid with their health
care providers. The legal basis for confidentiality is derived from statutes, which have been
enacted in essentially all states within the United States.
There is significant variation among the states as to what classes of health care providers
(doctors, nurses, social workers, etc.) can assert that they cannot share medical information
provided to them without the patient's express permission or a court order directing them to do
so.
Legislatures create state disease-reporting requirements that require medical confidentiality to be
breached as historically the courts have found that the state has a compelling need to protect its
citizens from certain infectious diseases (55). Certain behaviors, such as dangerousness, also
have resulted in reporting requirements that compel a health care professional to notify
authorities when a patient has made an explicit threat against a known individual (56) or even an
unknown but identifiable group of individuals (57). Dangerous behaviors include spousal and child
abuse.
Medical records also are protected by state and federal legislation. The U.S. Public Health Service
Act provides for the explicit protection of medical records dealing with treatment for drug (58)
and alcohol abuse (59). However, even the contents of substance abuse treatment records may
be released to medical personnel to the extent necessary to meet a bonafide medical
emergency (60,61). If release of the medical records is authorized by a court order, then
appropriate safeguards need to be put into place to preclude unauthorized disclosure.
The courts recognize that they must balance the public interest and the need for disclosure
against injury to the patient, physician-patient relationship, and treatment services (4,62,63).
A hospital medical record is the property of the hospital, but it is kept for the benefit of the
patient, the physician, and the hospital (53). A patient's medical record is preserved to document
events in a contemporaneous manner for later use. Later use includes further medical care and
treatment, documentation for financial payment by third-party payers, and defense of medical
malpractice claims.
The medical record librarian becomes the custodian of the medical records. The American
Association of Medical Record Librarians has a code that is similar to the Hippocratic Oath taken
by physicians. Hospital policies and procedures, consistent with state and federal statutes and
regulations, prevent a medical record from being released without either a patient release or court
order. Under certain defined circumstances, medical records can be admitted as evidence in a
court of law. They can be authenticated as business records. They also can be used to refresh a
doctor's memory and document his or her actions.
The American Hospital Association, JCAHO, and other health care professional organizations have
asserted that a patient's medical records are to be protected from unauthorized and unnecessary
access. Their positions are generally consistent with current HIPAA regulations and state
P.97
regulations. Hematology, blood chemistry, urinalyses, and radiographic, sonographic, and
electrodiagnostic findings are all considered a part of a patient's authorization or court order.
Third-party payers, as a condition of their insuring the patient, almost invariably are given access
to patient medical records, as are state and federal auditors.
The privilege of medical confidentiality does not extend to third parties present who are not part
of the health care being delivered. That is, police officers present during an evaluation or
treatment cannot be prevented from sharing with other law enforcement personnel or the courts
any information that they obtained under such circumstances. When an individual engages in
litigation in which his or her physical or mental status is an issue, that individual cannot assert the
privilege to prevent unfavorable information from reaching the ear of the court. Thus, in a lawsuit
for malpractice, the patient/ plaintiff has waived his or her right to medical confidentiality of any
oral, written, or electronic communication concerning his or her medical history, diagnosis,
treatment, or prognosis. Medical confidentiality statutes were and are designed to protect a
patient's privacy and to encourage treatment for conditions that may bear social or other stigmas.
The laws were not made to give the moving party, the patient/plaintiff in a medical malpractice
case, an unfair advantage by allowing the individual to select only those records considered
testimony favorable to his or her case. A medical record cannot be used as both a sword and a
shield.
In 1991 the Institute of Medicine (IOM) advocated the adoption of the computer-based patient
record as standard medical practice in the United States (64). A computer-based record is
perceived as a continuous chronological history of a patient's medical care. The medical care
record can be linked to various aids, including reminders and alerts to clinicians and clinical
decision-making instruments. However, a computer-based record increases access to a patient's
record and increases the array of data maintained in a single record (65). The more information
compressed into an easily accessible single location, the greater the precautions needed to
prevent misuse. As Annas (66) notes, in a setting of private practice, medical information that
identifies a patient is supposed to be transferred from physician to physician only with the
patient's written informed consent. In contrast, he explains, within a medical institution,
information is generally passed around on a perceived, usually self-designated, need to know
basis without first obtaining a patient's informed consent. Individuals who receive their medical
care through managed care facilities and integrated health care delivery systems that have
multiple treatment sites and use computer-based patient records can anticipate that the
traditional standards of medical confidentiality will be diminished. Even a minor error, such as
dialing an incorrect fax number and sending an electronic report to an unintended recipient, can
result in damage to the patient and ultimately cost to the individual who authorized the report to
be transmitted in error.
A release of medical information request can be general or rather specific, depending on the
clinical and social circumstances, needs of the treating health care providers, and instructions of
the patient or guardian. In general, medical information is released upon written instruction;
however, oral instructions frequently are sufficient or necessary. This may be the case in a
medical emergency. Documentation of oral permission for release of medical information is
recommended.
Health Insurance Portability and Accountability Act of

1996
The HIPAA, Public Law 104191 (4), is concerned in general about the written word and not oral
communications, especially those provided in an ongoing manner in the course of clinical
treatment.
HIPAA set national standards for the protection of health information, as applied to the three
types of covered entities: (a) health plansan individual or group plan that provides or pays the
cost of medical care that includes the diagnosis, cure, mitigation, treatment, or prevention of
disease; health plans include private entities and government organizations such as Medicaid,
Medicare, and the Veterans Health Administration; (b) health care clearinghousesa public or
private entity, including a billing service or community health information system that processes
medical, financial, or other data; and (c) health care providers who conduct certain health care
business transactions electronically. Health care providers such as physicians, hospitals, and
clinics are covered entities if they transmit health information in electronic form in connection with
a transaction for which a HIPAA standard has been adopted by the Department of Health and
Human Services (DHHS).
The compliance date has passed; for larger entities it was April 14, 2003, for smaller entities it
was April 14, 2004. If HIPAA is more stringent than a state's law, it governs; if the state law is
more stringent than HIPAA, the state law governs. Generally, covered entities are required to
comply with both HIPAA and state law whenever possible. HIPAA preempts any contrary provision
of state law, including state law provisions that require written records rather than electronic
ones. State law, however, is not preempted in the following circumstances: regulation of
insurance or health plans; prevention of fraud and abuse; reporting on health care system
operations and costs; regulations concerning controlled substances; when a state law requires the
reporting of disease or injury; child abuse, birth, or death; public health surveillance,
investigation, or intervention; or when a provision of state law is more stringent than the
requirements of HIPAA.
As with every issue of conflict between state and federal laws, there are exceptions. Exceptions
mean health care lawyers need to be consulted when a clinician is unsure as to his or her course
of action with regard to HIPAA and state confidentiality laws. There are other federal laws that
may be relevant to parental conduct, such as the federal
P.98
Drug and Alcohol Confidentiality Regulations (58) which govern the release of medical records
that contain information about drug and alcohol use, and the terrorism provisions of HIPAA which
allow access to medical records if matters of national security may be involved.
HIPAA was designed to protect patients' medical records and their confidentiality. Other laws have
existed to do this. The U.S. Congress believed, in consultation with many patient advocates,
health care providers, and other commentators, that more than 1,000 additional pages of laws
and regulations were required. If HIPAA serves to demonstrate two basic principles that health
care professionals have long recognized, it is that first, medicine is no longer a one-on-one
relationship and second, medicine has evolved with all its disciplines, providers, and equipment to
be a major economic force and component of every industrialized nation.
HIPAA has forced health care professionals to better communicate with patients and their
relatives. More ironically, 911 and the potential for future terrorist activities, has demanded that
the health care professions better communicate with potential patients, the general public,
politicians, and the media. Risk communication has been transformed from an arcane technique
used by military and emergency preparedness personnel to a staple of daily communication. The
basic principles of risk communication are simple and totally applicable to the practice of
medicine, especially neonatology: Tell what you knowclearly, succinctly, and in a manner that
every sixth grader can understand. Tell what you do not know clearly, succinctly, and in a manner
that every sixth grader can understand. Do not guessinformation has to be data driven. It is
better to say we do not know now, and when we do, we will share it with you rather than to
provide incorrect information initially because incorrect information causes credibility to suffer,
believability to suffer, cooperation to suffer, and ultimately there is more confusion, anxiety, and
poorer outcomes. Communication issues are discussed further in the second section of this
chapter.
HIPAA delegates to local health departments the task of collecting vital statistics that include
birth, death, and marriage information. These documents generally are protected to some degree.
Autopsy reports usually are not released beyond the treating physician and/or coroner/ medical
examiner, unless there is authorization to do so. Civil or criminal courts can order the contents of
an autopsy revealed if it would assist a party in asserting his or her civil claim, criminal
prosecution, or defense from prosecution.
Legal Issues Specific to Managed Care

Although the incidence of malpractice litigation against health professionals has increased
exponentially in the U.S. since the 1970s (67), it is not a recent phenomenon. Documented
medical professional liability cases date to 1374, when a surgeon in England was sued for
negligent treatment of a wound (68). However, a hybrid of lawsuits against health care providers,
focusing on managed care issues, is new (69,70). Profit is the bottom line in today's health care
system (71). Managed care developed in response to a perception on the part of both government
and society in general that there were unnecessary medical procedures being performed with an
associated overutilization of medical resources and services, and an unrestrained lack of
uniformity of cost in health care. Managed care filled a vacuum; it was developed and marketed
because of a perceived need to control by those who paid costs (the employers) and those who
managed health care dollars (health care insurance companies). Health care providers did little to
inform their constituencies of the potential problems associated with managed care.
Managed care may be defined as any entity capable of negotiating to deliver health care to a
group of recipients at a predetermined rate on a per capita basis (71). The American Association
of Health Plans claims that 150 million Americans were enrolled in health maintenance
organizations (HMOs) or other managed care entities in 1995 (72). Almost 75% of those who
receive health insurance through their employers are covered by some type of managed care plan
(73).
Managed care has become big business, with the administrative costs and profit deriving in
major part from cost savings and efficiencies (74). The focus of the decision maker to authorize or
perform a service or procedure or order a medication now has an aspect of a real or perceived
profit motive: There has been considerable media coverage of managed care programs that firmly
control costs to the detriment of those paying premiums and expecting the best services
available. In many cases, unqualified personnel are placed in a position of second-guessing and
overruling physician decision-making responsibilities and obligations. Delays in receiving

permission for expensive procedures or outright denials for critical but costly procedures occur. In
some instances, so-called gag orders had been inserted in physicians' contracts, which
effectively prohibited physicians from disclosing expensive, and potentially more effective,
alternative treatments to their patients (71). However, in February 1997 President Clinton
requested that the DHHS send a letter to all state Medicaid directors informing them that gag
rules are prohibited for Medicaid HMOs. He also supported national legislation to ban gag rules
for all managed care organizations within the United States. Approximately half the states in the
United States have enacted or introduced their own laws regulating gag rule provisions in
medical care policies. In Moore v Regents of the University of California, the California Supreme
Court held that the concept of informed consent is broad enough to include the physicians' duty to
inform their patients that they have an economic interest which might affect that physician's
professional judgment (75).
Although managed care attempts to reduce costs by creating economies of scale and coordinating
care among health care providers and resources, they also attempt to reduce costs by eliminating
unnecessary care, as they define it. In some instances, a managed care program's criteria
P.99
for emergency or extensive and expensive care may conflict with those held by the physician. The
physician's contract may have a provision in which a percentage of fees are paid only if utilization
goals are met. The physician's contract also may have a capitation provision, which means that
the physician is provided a fixed amount regardless of the level of service provided to each
patient. In the latter circumstances, the physicians essentially become coinsurers and may have a
perceived conflict of interest in the manner and degree to which they provide services to their
patients.
The most unique feature of an HMO is that an enrolled patient pays a prepaid, fixed fee for
medical services. This is in contrast to fee for services in which the patient pays a separate fee
for each service rendered by the independent physician. Preferred provider organizations (PPOs)
differ from HMOs in that a PPO is an organized group of health care providers offering their
services at a discount. Services are provided on a predetermined fee-for-service basis. Members
may choose physicians regardless of whether they are in the members' plan but receive
discounts only from physicians within the plan. PPO physicians have relatively minimal financial
incentive to limit services, and physicians practicing under such an arrangement would appear to
have greater control of their practice and potential liability.
The HMO may contract directly with physicians and pay them a salary (staff model), or they may
contract with a group of physicians. The physician group may or may not devote a majority of its
time to serving the needs of a specific HMO. It may offer a mixture of financial arrangements to
its patients, including fee for service. The HMO may contract with an individual practice
association (IPA), usually a partnership or corporation of physicians, to provide health care
services to the HMO membership. The IPA then contracts with its physicians to provide services to
the HMO. The fundamental difference between the IPA and the staff and group physician
arrangements is that IPA physicians usually work in their own facilities, use their own equipment,
and keep their own records. The HMO pays the IPA a capitation fee (a specific amount per
enrollee or subscriber); the IPA pays the treating physicians on a fee-for-services basis (76).
Physicians and their managed care carriers are being sued jointly. Allegations of malpractice now
are joined with allegations of bad faith, and contract theories of liability are being added to the
more traditional negligence/medical malpractice theories of liability (69,70,77). Under the
doctrine of respondeat superior, HMOs have been found responsible for the alleged negligence of
their employeephysicians (78). The doctrine of respondeat superior states that an employer is
vicariously liable for negligence of an employee acting within the scope of his or her employment
(79). The doctrine does not apply if the negligent party is a true independent contractor (80). The
distinction between an employee and an independent contractor is control or independence; that
is, an employee is subject to the immediate direction and control of the employer, and
independent contractors use their own judgment and are not subject to direct control. In some
instances, an HMO can be held vicariously liable for the negligence of a consultant requested by
the HMO's physician (81). This issue may be decided on the wording in the promotional material
provided to subscribers. A managed care contract can increase a physician's obligations to a
patient; the contract cannot decrease or even limit the physician's obligations and legal liability to
the patient. When a physician employed by an HMO makes a treatment decision, that treatment
decision has financial consequences to the HMO (69). In a malpractice suit, there are now
assertions that the HMO-paid physician's treatment decisions were economically motivated rather
than being made in the best interests of the patient (69). Before a physician accepts a contractual
obligation, whether it is the product of an employment contract with a hospital, group practice,
HMO, or other form of managed or prepaid care program, it is strongly suggested that the
physician seek legal advice regarding the terms of the contract and the consequences of those
terms.
Regardless of the employment status of the health care provider, the hospital has an obligation to
oversee the quality of patient care and services (82). Ultimately, however, the courts have held
that it is the physician's responsibility to uphold good medical practice in the face of improper or
incorrect cost-containment procedures put forth by HMOs or other managed care organizations.
The California Court of Appeals has held that, while we recognize, realistically, that cost
consciousness has become a permanent feature of the health care system, it is essential that cost
limitation programs are not permitted to corrupt medical judgment (83). That court found for the
plaintiff because the physician did not protest the health insurer's determination that a prolonged
hospitalization for his patient was not necessary, with severe negative consequences for the
patient (83).
The Employee Retirement Income Security Act of 1974 (ERISA) (84) supersedes any and all
state laws insofar as they may now or hereafter relate to any employee benefit (85). In Shea v
Esensten, a federal court found that ERISA requires HMOs to disclose to their enrollees the
compensation agreement between the HMO and its physicians (86). That is, there is an
affirmative duty by the HMOs to inform their subscribers of any financial incentives that the health
care providers may receive as they manage their patients' care. These financial incentives must
be disclosed and the failure to do so is a breach of ERISA's fiduciary duties (86). Physicians who
have challenged managed care decisions about patient care have been removed from the HMO
panels. Suits for reinstatement on the grounds that they were removed from the panel without
good cause and in violation of public policy and the implied covenant of good faith and fair dealing
traditionally read into contracts have had mixed results (87,88). Just as medicine has evolved
significantly during the past several decades, so too has the law. Both will continue to change,
and their progress, conflicts, and
P.100
resolutions will be documented in the media, professional journals, legislature, and courts.
Physician Liability Arising from Nonphysician Providers

Liability exposure arising from the use of nonphysician providers (NPPs) (e.g., neonatal nurse
practitioner, clinical nurse specialist, physician assistant) is relatively new. It can be assumed,
however, that as NPPs increase their roles, there will be a concomitant increase in malpractice
allegations against NPPs and the physicians working with them, based upon the fact that the NPP
is acting as an agent for the physician.
To the extent that NPPs exercise control over their professional activities, they will be held
responsible for their negligent acts. Under the old captain of the ship (respondeat superior)
doctrine, the physician was presumed to be responsible for the activities of all the NPPs working
with the physician. Contemporary courts have moved away from this theory and have placed
responsibility on those professionals exercising control. However, this does not eliminate the
liability risk for the supervising physician.
Potential areas of risk for neonatologists working with NPPs and methods for reducing risk are
listed in Table 8-1.
TABLE 8-1 LIABILITY RISK AND RISK REDUCTION FOR NEONATOLOGISTS WORKING
WITH NONPHYSICIAN PROVIDERS
Areas of Risk
Inadequate supervision by
physician
NPPs working beyond their scope
Parent unhappy about access to
their newborn's physician
Physician viewed as deep
pocket by plaintiff's bar
Apparent physician delay in
seeing critical patient
Risk Reduction
Check NPP credentials before hiring and maintain

documentation
Establish, and review at least annually, written
policies, protocols, and procedures for:
patient examination
treatment
delegation
supervision
patient's/parent's right of access to

physician
Instruct NPP to adhere as closely as possible to
patient care protocols, using physician input when
significant deviation may be required; the
rationale for deviations should be documented
Document completion of skills inventory check
Document current competency and establish a
quality monitoring system
Educate other staff about the role and limits of the
NPP
Understand and keep current regarding statutory
requirements for NPPs
Use name tags to identify the professional status
of the NPP
Introduce NPP to parents and explain role
Review NPP's charts on a regular basis and cosign
orders in a timely fashion
Establish protocol for physician to see patient and
parents at set intervals
Obtain adequate liability insurance for NPP;
physician's liability insurer should be informed that
they are supervising NPPs
Ensure that NPP complies with hospital
credentialing requirements and has sufficient time
allotted for continuing professional education;
require that copies of licensure renewal and
certification of continuing education are kept on

file
Inform physicians covering call of NPP's role
The National Practitioner Data Bank

Some malpractice insurance carriers retain the right to settle lawsuits independent of the wishes
or sentiments of the health care provider. The federally mandated and operated National
Practitioner Data Bank contains information, available to hospitals and individuals, and lists
settled medical malpractice claims (89). General medical malpractice carriers may not always be
knowledgeable regarding the specific nuances of specialty practice.
Health professionals should do more than comparison shop the price of a malpractice policy. They
need to determine the financial soundness of the malpractice carrier and their own rights to
control settlement and to determine the sophistication and knowledge of the legal representation
provided by their insurance carrier.
P.101
QUALITY ASSURANCE, PERFORMANCE IMPROVEMENT,

ANDRISK MANAGEMENT
The process of quality monitoring and improvement (QMI) is also discussed in Chapter 7. This
part of the chapter emphasizes how QMI and risk management are closely interwoven and how
both monitoring systems are used to improve patient care.
The Growth of Risk Management

In the United States, risk management programs grew out of changes in the medical malpractice
insurance market (insurance became more difficult and more expensive to obtain) as malpractice
claims accelerated in the 1970s (67). As the cost of liability insurance increased, the
establishment of risk management programs made health care facilities more attractive to
insurance underwriters. When institutions formed their own self-insurance trust funds, risk
management programs provided some assurances to hospital board members that the financial
liability of the institution would be limited to the greatest extent possible by their own risk
managers. Depending upon the size and complexity of the health care institution involved, there
may be one set of professionals who provide risk management guidance relative to standards of
care and a second set of professionals under the leadership of a compliance officer who assure
compliance with the requirements of regulatory bodies. Ensuring compliance requires a thorough
knowledge of applicable standards and a commitment of time specifically directed towards
understanding compliance and related issues. In smaller institutions the compliance officer and
risk manager may be the same individual.
Basic Elements of a Compliance Program

A compliance program is a set of procedures within a medical practice or other health care facility
which are designed to ensure adherence to laws and regulations associated with either federal or
privately insured health care programs and thereby to prevent or reduce improper conduct.
Although the size (e.g., number of employees) and type of practice or organization determines in
large part the extent and formality of its compliance program, even small, private, neonatology
practices are expected to implement one. Compliance programs establish and communicate
standards of conduct; provide channels of communication to inform employees of rights, benefits,
and obligations, and to answer questions; enforce standards through the consistent application of
disciplinary measures; and require ongoing auditing and monitoring to ensure that identified
errors are corrected and future similar occurrences are prevented.
The rapid rise in health care costs in the United States during the past three decades has
prompted a significant increase in governmental scrutiny. The results of government audits have
created a dramatic change in public opinion regarding the level of waste, fraud, and abuse in the
health care system. This has led to the passage of legislation that creates a trust fund for
investigations and expands governmental prosecutorial powers and also has markedly increased
penalties for noncompliance. For instance, a claim for payment for medical services to Medicaid or
Medicare which is deemed to be fraudulent under the False Claims Act (90) requires payment of
three times the amount of the overpayment plus a mandatory $5,000 to $10,000 fine per claim.
And since each individual service billed for is a claim under the definitions used in the statute,
penalties can become enormous. On the other hand, erroneous claims (i.e., those resulting from
innocent errors) require only the return of the amount of the overpayment. Fraudulent claims
result from three circumstances: (a) actual knowledge that the claim is false, (b) reckless
disregard of the truth or falsity of the claim, or (c) deliberate ignorance of the truth or falsity of
the information. To mitigate the potential and corresponding penalties for noncompliance,
voluntary compliance programs have become commonplace and necessary in the health care
world. The voluntary compliance programs recommended for health care providers are designed
to detect and/or prevent illegal activity through self policing.
Providers meeting the requirements of an effective compliance program, listed below,
demonstrate their commitment to creating an environment in which payment claims are accurate;
fraudulent behavior does not occur; improper practices are prevented, detected, or rectified;
wrongdoing is reduced; administrative liability is mitigated; and the mental state of reckless
disregard is negated. Although all compliance programs do not need to be alike and the degree to
which each element needs to be addressed varies among practice types, compliance programs
should address the following seven basic elements (91):
establish written standards of conduct, policies, and procedures
designate a compliance officer or contact
provide mandatory training and education
create and publish accessible lines of communication
audit and monitor compliance with guidelines
enforce through clear disciplinary guidelines
respond to violations and take corrective action
Although neonatal practitioners are not expected to be experts in law or regulation, they need to
have a good working knowledge of relevant legal and regulatory requirements that relate directly
to their duties and responsibilities. Regulations often are complex, ambiguous, and sometimes
silent on key issues. Whenever in doubt about a legal or administrative issue, it is important for
the practice or practitioner to consult with an expert before proceeding. The list of areas of law/
regulation which the practitioner should understand includes, but is not limited to, the following
(91):
billing and coding
patients' rights
health care antifraud and antiabuse laws
P.102
Medicare/Medicaid antikickback statutes

conflicts of interest
harassment in the workplace
environmental protection laws (e.g., disposal of hazardous material and
medical waste)
health and safety programs and Occupational, Safety, and Health
Administration regulations
HIPAA regulations
receipt and giving of gifts, gratuities, and other business courtesies
credentialing requirements, including sanction screening
insider information and securities laws
patent, copyright, and intellectual property laws
record retention requirements
professional standards of care
Stark legislation (e.g., physician antireferral statues)
the Privacy Act
Establishing a Risk Management Program

Health care errors rarely are due solely to a mistake made by an individual; they are usually an
indication of failure of a health care system. System issues are referred by risk managers to staff
trained in quality improvement and to the hospital administration, and vice versa. An effective
QMI program works in close cooperation with the risk management team, with the ultimate aim
being the improved care and safety of patients and lowered liability exposure. The importance of
monitoring all aspects of a health care system in the prevention of medical errors was emphasized
by the IOM in their 1999 report (92) on the nature and extent of medical errors in the practice of
medicine.
Risk Reduction as It Relates to Quality Monitoring and

Improvement and Risk Management
Risk is the possibility of harm, the uncertainty of danger, and the probability of loss. The essence
of a QMI program and of risk management lies in maximizing the areas in which there is some
control over the outcome while minimizing the areas in which there is absolutely no control over
the outcome. The existence of risk does not imply that there are always alternatives. The terms
QMI and risk management are used to describe a process of examining any potential alternatives
and thereby influencing, in a positive manner, the level of risk to which patients, health care
providers, and institutions are exposed. An outcome of managing risk is the identification of a
level of harm or danger that is the irreducible minimum. However, QMI and risk management
differ in the order of priority of their primary goals. In the health care setting, harm or danger
expressed in human terms pertains to the safety of patients, visitors, and employees, and to
health care delivery practices; however, harm or loss can also be expressed in financial terms,
such as losses in settlements, jury verdicts, legal expenses, and payment of increased insurance
premiums. QMI is the term used for the process in which the primary expected reduction is
expressed in human terms; risk management is the term used when the primary expected
reduction in risk is expressed in financial terms. Risk reduction is achieved using a process that
involves investigating, evaluating, planning, organizing, and implementing procedures. The
resulting benefit of risk reduction is improved quality and understanding of patient care and also
preservation of the financial resources necessary to provide optimal patient care when there are
fewer claims made and fewer lawsuits brought against the health care providers (93,94). Financial
and psychological benefits also are realized when malpractice cases are dismissed or settled for
reduced amounts, based on the strength of the defense.
The level of risk exposure in the health care environment is inversely proportional to the control
maintained by the health care service providers. Total control is not possible in the face of
complex disease processes and complex care and treatment managed in complex patient care
delivery systems. In the critical care environment, a risk reduction program must be established
that anticipates, identifies, and responds to risk. When it is possible in the process of delivering
care to patients, fail-safe systems are established to prevent errors from affecting the patient.
The elements of a risk reduction program also reflect the health care organization's attempt to
define, describe, and impute liability within the organizational structure. Functional definitions of
health care executives and their roles and responsibilities for supervising and administering health
care activities enable the health care providers to identify chains of management decision making
and accountability. Health care providers need to understand who supervises each administrative
component of a health care facility so that complaints, comments, and recommendations reach
those who are in a position to respond affirmatively. The chain of responsibility for developing,
implementing, supporting, monitoring, and evaluating a risk reduction program reflects
multidisciplinary health and management skills and activities. It must be recognized that a risk
reduction program is both a proactive (preventing potential incidents from occurring) and reactive
(responding to an incident or seminal event and preventing future similar incidents) system of
intervention to protect the patient, institution, and clinician. Although assignment of responsibility
does occur in either a proactive or reactive risk assessment, responsibility is not to be perceived
as the same as negligence. For example, there is no negligence in a situation in which a physician
orders a specific dose of medication, the nurse responsible administers the proper dosage, and
there is a resultant catastrophic effect on the patient. In many instances, either the risk was
known and shared with the parents or the adverse consequences could not reasonably have been
foreseeable.
Risk reduction programs strive to reduce risk to the greatest extent possible and to manage the
remaining risk. Quality improvement committees, safety committees, and specialized
subcommittees provide the administrative and
P.103
management structures to review the various aspects of clinical care and support and to provide
constructive guidance and assistance. The outcome of this effort is improved quality of care within
the context of the goals of the health care team and the patient's family.
From the perspective of risk management, risk financing involves the financial arrangement
(insurance) made for the payment of losses that inevitably do occur. The liability insurance
program may be coordinated by the risk management department. The risk manager notifies the
insurance carrier of reportable events, claims, and lawsuits. Members of the health care team
have a shared responsibility to report adverse events, contribute to the data that will track these
events, and alert the malpractice insurance carrier to potential financial loss. The risk manager
may select defense counsel and assist counsel in the defense of the institution and/or health care
providers. In addition to contributing to the quality of patient care, the risk manager, under the
direction of counsel, contributes to the quality of the defense of the health care provider in a claim
or suit (94,95). Conflicts of interest between the institution and health care provider need to be
identified and addressed as soon as possible. Separate counsel for the medical institution or group
practice and medical practitioners may be required when there are divergent institutional and
personal interests. For instance, one party may wish to settle their portion of the case while the
other named party or parties may wish to litigate the allegations against them.
From the QMI perspective, adequate risk financing, malpractice insurance, requires the
acknowledgment that reducing the risk to patients and staff in the health care system never
means totally eliminating it. There will always remain some minimal risk for an untoward event.
Hence, malpractice insurance is needed to cover the cost of defending the claims and the cost of
paying any judgment, should the defense efforts prove unsuccessful.
To achieve the goals of risk management, the following activities must take place:
Systematic and continuous investigation of risk exposures. Risk managers must be made aware of
potential and actual exposures (adverse events) as they occur and of system changes that may
affect patient and staff potential exposure to risk. To achieve this level of awareness, it is
necessary for risk managers to establish a communication network that provides real-time
feedback. The risk manager utilizes the knowledge, impressions, and experience of the health
care team to investigate events that have occurred and to identify events that might occur in the
future. Methods of communication to risk managers include incident reports, risk management
presence on committees, and ready accessibility by electronic means.
Evaluation of risk loss exposure. An examination of the nature, frequency, severity of risk, and
potential impact on the patient/organization detects patterns in recurrent events, which can be
used to predict their recurrence. For example, if medication errors are observed to occur just
before a change of shift, examination of the human dynamics that contribute to these errors and
breaking the pattern can lead to correction of an error-producing systems flaw. Administration
and clinical staff are called upon to identify potential risk areas for evaluation.
Planning and organization of appropriate risk avoidance and prevention techniques to efficiently
minimize loss to the organization. Currently in the United States, order, stability, and consistency
are not the hallmarks of a successful health care organization. Risk management demands the
establishment of an information-gathering network, which identifies emerging organizational
changes. Areas of risk can be identified as projects are developed. The monitoring of system
changes after implementation is used to detect any negative impact on patient care.
Recommendations can then be made to adjust system changes. Anticipation is one of the most
important considerations. Identifying pertinent questions regarding the level of risk for potential
adverse events and those who should be involved in evaluating a system helps the direct clinical
service providers avoid some significant errors (96).
Implementation of risk reduction. Good clinical staff are flexible, innovative, and creative; thus,
there is constantly the potential for an adverse event to emerge from new technology or
treatment modalities. Simplifying the organization or health care practices to gain control is not a
realistic goal in today's complex health systems. By using reporting and communication
connections within and outside the organization and analyzing data from the investigation of
adverse events, areas of greatest risk can be identified. The information about potential risk is
communicated to committees or individuals who then put in place mechanisms to reduce the risk
to the patient and institution. That the risk cannot be controlled completely should also be
communicated, particularly to the patient's family.
The Risk Management Plan

Risk management activities are outlined in a program description. Such a document provides clear
and concise information about the risk management program, which is available to all those
within the institution.
Staff Education
Personnel involved in QMI or risk management must recognize that they are educators. Every
committee meeting, exchange with a staff member, or participation in orientation is an
opportunity to dispel misinformation and increase understanding of the risk management process.
The fundamentals of risk management are not highlighted in medical and nursing schools but are
necessary to prepare future health care providers for dealing with risk issues that arise in the
clinical setting. The development of a risk management curriculum gives structure to educational
P.104
efforts and provides a framework to cover essential topics, ranging from the orientation of new
employees, staff physicians, and residents to seminars for advanced practitioners, department
chairpersons, and administrators.
As an outcome of a core curriculum in risk management, all individuals in health care should be
able to:
define risk
know the basics of how to set up a risk reduction study, including definitions of area of risk
to be studied and sentinel events
describe what a risk manager does and why
understand what events need to be communicated to risk managers
recognize that if a facility, treatment, or system is replaced, the replacement must be

monitored until proven to decrease risk
describe how to access risk managers for reporting and receiving information
describe what relationship risk managers have with professional liability insurers
Grupp-Phelan and associates (97) reported that pediatric residents were named in 26% of
malpractice suits. All health care providers, including nurses, nurse practitioners, residents, house
staff, and ancillary staff such as respiratory therapists and clinical nutritionists, should know what
kind of process a risk management investigation will follow (Fig. 8-2).
Figure 8-2 The risk management investigation process begins with the receipt of notice that an
adverse event has occurred. An attorney directs some investigations in jurisdictions where the
risk management activities are not protected from discovery in litigation. The personal notes of
health care providers are not protected from discovery. After many rounds of questions,
opportunities for improvement are reported to quality improvement councils. In a separate
activity, the results of investigation directed by the defense attorney are sent to him or her.
Health care professionals should avoid making editorial comments in patient records and should
not keep personal notes about clinical events. Objective clinical impressions belong in a patient's
medical record. Physicians, nurses, and other health care professionals should understand that
investigations take time and that the findings may be very different from the initial views of the
staff. Speculation about the causes of adverse events or discussions of the events outside the
peer-review/quality improvement areas should be discouraged.
With an understanding of the fundamentals of risk reduction/management, the health care
provider can move on to identify systems and processes in his or her care that can be examined
for safety and effectiveness. The risk manager analyzes the frequency or severity of adverse
events or unexpected outcomes and the findings during the course of discovery in lawsuits, and
then identifies topics that can be presented in an advanced risk reduction curriculum. These
include: retroactive peer review (e.g., clinicopathologic review, regular review of perinatal or
neonatal patient care statistics), documentation, communication, supervision, monitoring and
assessment, coordination of care, medication administration, system failure(s)
P.105
and human error, and changes in local, state, or national regulations or laws.
Adverse Events
Adverse sentinel events were defined by the JCAHO in 1996 as unexpected occurrences involving
death or serious physical or psychologic injury or risk thereof. The term sentinel is used because
the event should sound a warning that requires immediate attention. A root cause analysis is
recommended by the JCAHO as the format for an intensive assessment to reduce the variation
and prevent the event in the future. This type of analysis addresses these questions: what
happened, why did it happen, and what processes were involved when it happened? The analysis
can be used to piece together the reasons why a mistake or complication occurred. If the system
worked previously, what elements have changed to produce the potential for error (98)
In some instances, it is not possible to sort out the exact sequence of events. Interruptions in the
system of care delivery that are identified should be repaired or redesigned to prevent further
breakdowns. Because serious occurrences do not occur very often, it is necessary to optimize the
lessons learned from each event, including an application of those potential events that did not
occur. Adverse, unexpected, or poor patient outcomes do not necessarily indicate that a mistake
has been made. It is important not to make this assumption within the health care team, nor to
accuse or place blame. When there is an unexpected outcome, the risk manager is called upon to
perform the process of identifying the possible contributing factors.
Receiving an incident report and investigating problems that have occurred or potentially might
occur is viewed by the risk manager as an opportunity for improvement in the system of care
delivery. This positive outlook may not be shared by the staff. The stigma of punishment or
retaliation in relation to incident reports exists and is a deterrent to obtaining the information on
which to base a recommendation for change and improvement in health care delivery.
Accountability versus Blame

System and practice patterns are the focus of a risk management investigation. Investigations
may reveal systems and patterns that do not provide sufficient or effective safety checks to
prevent human error or miscommunication. In addition to identifying the flaw in systems or
processes that potentiate mistakes, adverse event investigations studied in the aggregate identify
the types of patients most likely to be affected. Andrews and associates (99) found that more
patients with a serious illness had an initial adverse event, and that the likelihood of an adverse
event increased 6% for each additional hospital day. The analysis of event databases provides the
health care team with information that they can use to modify practice or alert other practitioners,
thereby improving care and reducing the likelihood of lawsuits.
Reported predictors of the outcome of pediatric litigation include severity of resulting disability
(100), defensibility of the case (101), and parental anger (102) (often exacerbated by a
perception that communication with health care providers was less than satisfactory). The
importance of patient or parental anger was demonstrated by Andrews and associates (99), who
found that of 1,047 patents with serious injury, 17.7% experienced serious adverse events that
led to longer hospital stays, but only in 1% to 2% of these adverse events was a legal claim made
for compensation.
Pichert and associates (102) studied 43 closed claims, covering 8 years, involving children. A
significant portion of the claims were associated with high-risk medical conditions (37%) and
problems associated with intravenous therapies, central lines, and catheters (21%). In 9% of the
claims physicians relied on faulty information, and in 19% the parent(s) became particularly
enraged by an injury.
Conventional wisdom holds the view that human beings are intrinsically unreliable. From this it
follows that when something goes wrong, someone must have erred. It is assumed that mistakes
such as these are the result of inattention, carelessness, and negligence. The remedies invoked to
prevent further lapses include finding the culprit, assigning blame, and acting to correct future
misdeeds. The truth is that the errors people make often are traceable to extrinsic factors that
predispose an individual to fail. Whenever human error is suspected, it is a sound policy to trace
the error to its root cause(s). To assign blame to an individual who makes an error is no
assurance that the same error will not be made again by that individual or another (103).
The health care provider should not assume that he or she or another member of the team is
responsible or liable for an unexpected outcome. Frequently the first impression regarding the
cause of poor patient outcomes is significantly different from the cause(s) recorded at the
conclusion of an investigation. There is also a possibility that a cause may never be found. There
should be no speculation by those directly or peripherally involved as to the cause or etiology of
the event. A defensive position in response to the family's accusations should be avoided, as
should the assignation of blame to other disciplines, departments, and/or systems.
The analysis of a single event and analysis of event trends provide the health care team with
information that they can use to modify practice or alert other practitioners, thereby improving
care and reducing the likelihood of repeated errors.
Communication
Risk communication includes those techniques for disseminating information that may be complex
and include elements of uncertainty as to diagnosis, treatment, and prognosis. Chess and
associates (104) note that some common myths often interfere with the development and
implementation of effective risk communication. The manner in
P.106
which the information is provided (the style and process of information delivery) is just as
important as the content of the message. Treatment options need to be provided, including the
option of no treatment. Confrontation should be avoided. Questioning should be encouraged and
not perceived to be a challenge to a clinician's knowledge, treatment interventions, or plans.
Neonatal intensive care units are highly emotionally charged environments, both for the parents
and concerned family members and health care providers.
Communication with Parents and Guardians

Many factors that lead to medical malpractice lawsuits are perceptual and qualitative in nature. In
many cases, lawsuits are not the result of malpractice, but are instituted because the patient or
patient's family feels that information was withheld from them or that they were not told the
truth. Effective communication skills used in data collection, developing relationships, and dealing
with parents' emotions help to prevent anger and possible malpractice litigation (105).
The perception by the family of the competency of the physicians, nurses, and other health care
team members is drawn from their interactions with them. Ideally, contact with the family
inspires trust, confidence, and openness. Positive contact with the family of a neonate is based
upon the fundamentals of communication: body language that conveys attention to the family and
infant, taking time to answer questions, asking for input from the family regarding their
impressions of the baby's progress, and speaking in terms that can be understood by the family.
These basic clinical and social skills represent the foundation of defensive medicine techniques. It
is important to convey the message that team members taking care of the infant are
communicating with each other, are fully aware of the treatment plan, and are coordinated in
their approach to the care of the infant. Judgments and preconceived opinions about the family
should be avoided, regardless of their lifestyle or behavior toward the health care team. A study
by Cuttini and associates (106) concluded that parents generally are satisfied with the information
received, although some complain about the style of communication and especially the need to
ask repeatedly in order to be informed. The latter complaint may be more common in teaching
hospitals because of the greater complexity of the health care hierarchy produced by the presence
of medical students, interns, residents, and attending physicians, who may change every month.
No matter how large the team caring for a neonate, the attending physician should be known to
the parents and guardians and readily accessible to all salient parties.
Those who provide health care need to understand the manner in which they communicate to
parents and other concerned parties. Health care providers must communicate in a manner which
is linguistically, contextually, and culturally appropriate. The reluctance of many health care
providers to convey the limitations as well as the achievements of the health care can significantly
increase the family's distress when negative changes in their infant's condition occur. When the
neonate or infant is not doing well, that is the critical time for more communication and more
interaction with parents and significant others, not less. A decrease in communication at a critical
juncture can and will damage the health care team's credibility. Training in risk communication, in
conveying negative information, is essential.
Covello and Allen (107) identify seven cardinal rules of risk communication. Though these rules
are designed for individuals who have to deal with the general public, they generally can be
adapted to a hospital setting (91):
Accept and involve the family or guardian as a partner. Information is to be shared;

concerns may not always be eliminated.
Present the plan or course of action that may even include alternative plans.
Listen to the concerns being stated. A concerned individual cares more about trust,
credibility, competence, and empathy than cold statistics and complex details.
Be honest, frank, and open. Recognize that once trust and credibility are lost, they are
almost impossible to regain.
Different health care providers may have different opinions; different family members may
express different concerns. This is to be expected. Open and frequent communication
minimize doubt and distrust. A single spokesperson, or a limited number of spokespersons,
should be designated to minimize the diversity of opinions being provided.
Concerned parties are more interested in risks, simplicity of explanations, and the danger to
the patient. The focus of the discussion needs to be on the patient, not on the health care
provider.
Clear communication in a language and vocabulary understood by all parties needs to

occur. Communication should be accompanied by compassion.
For the family of a critically ill neonate, the distinction is unclear between a bad outcome due to
disease progression and unfortunate circumstances/errors/negligence. Anxiety over the
uncertainty inherent in the care of the seriously ill may evoke feelings of helplessness. Outcomes
that are expected by the physician but not clearly conveyed to the family may be a shock for the
family. The nonoccurrence of expected outcomes leaves the parents feeling that a mistake has
been made. Families who experience disappointments in clinical outcomes may take out their grief
and despair on the health care provider. The provider may be unprepared for the parents' anger
and distrust. Appearing uncertain, handling questions improperly, apologizing for oneself or the
team, not knowing the available information, not involving the family in the decision-making
process, not establishing rapport, appearing disorganized, and providing the wrong information
(such as discussing the wrong patient) can all lead to a breakdown in communication.
P.107
Acknowledgment of hostility is acceptable; this indicates that the health care provider recognizes
the reality of the situation. Anxiety undercuts confidence, concentration, and momentum. Listen
to what others have to say, even if you disagree with them; recognize expressions of frustration.
State conclusions and then provide supporting data. Above all, do not lie.
If an adverse event occurs, the honest, sincere, and compassionate response of the health care
team to the family will reduce the family's view that there is a need to bring a lawsuit. Without
burdening the family with personal feelings of inadequacy, regret, or vague misgivings, it is
helpful to show concern and to express empathy regarding the patient outcome. Silverman (108)
offers insight into how parents of infants blinded by retrolental fibroplasia (RLF) (retinopathy of
prematurity) view the health professional who fails to convey appropriate concern:
As compared with conversations with RLF-blind young adults, discussions
with parents, singly and in groups, were much more difficult for me (and
for them). Most of the parents were still bitterly angry at the medical
profession, but not for the reason I imagined. They understood and
accepted the fact of limited knowledge at the time their children were
born. Most were convinced that physicians had rendered excellent care
and had used supplemental oxygen liberally in well-meant efforts to
improve the chances of the small babies for intact survival. But, almost
without exception, parents recalled (with rancor) that once the diagnosis
of RLF was made, a chill in relationships developed. At the very time
when they needed support and advice, their physicians became distant
and defensive, the parents recalled. Most blamed their doctors for failing
to maintain interest and concern, not for the failure of clairvoyance! The
parents said it was anger at personal, not professional, behavior of
physicians which prompted many of the RLF legal suits charging
malpractice which burgeoned in this country.
A member of the health care team should be designated as the primary communicator to share
with the family the treatment plans that have been formulated in response to the adverse event.
At the request of the family, extended family and other support may be included in family
conferences and discussions of the causes or possible causes of the event. Maintenance of a
coordinated and ongoing communication with the family may be difficult in the face of the family's
anger and despair. However, the goal of the health care team must remain the support and
treatment of the patient and family.
Communication with and within the Health Care Team

From the significant number of dangerous errors that occur in the intensive care unit, Donchin
and associates (109) found many were attributed to problems of communication between
physicians and nurses. They suggest that studying the weak points of communication in a unit
may reduce errors, and that errors should not be considered an incurable disease but rather a
preventable phenomenon. Clear oral communication is vital in a NICU; the medical record alone
cannot be relied upon to adequately convey urgent information between members of a health care
team.
The medical record should be designed to achieve clear communication with and among the
members of the health care team. Clinical information is focused on the patient and family, and it
derives from repeated assessments, analyses, judgments, and actions aimed at achieving specific
goals. Documentation in the patient record allows ongoing evaluation of patient progress and
review of the clinical management plan. The individual contributions of the members of the health
care team are recorded and used by other team members as a basis for planning each new step
in patient care. This documentation of the exchange of information among the team and the
recording of the outcome of each team member's analysis of the information also helps to support
the team's choices and judgments. Documenting patient care, contemporaneous with the events,
provides the clinical picture that will be used in the defense of the care provided should an
adverse event occur. If the clinical record has the appearance of being a battleground for warring
or defensive factions of the team, the record will reflect a lack of team cohesion and direction that
is difficult to defend. The record should contain the facts of the event, assessments of the patient,
the decision-making processes, and the interventions undertaken.
Any documentation of an abnormality in the medical record should be accompanied by the
reassuring factors that support the overall interpretation of the clinical findings. All factors that
explain decisions made in the face of abnormal finding(s) should be recorded. In fact,
documentation of reassuring factors should accompany the recording of abnormal factors (110).
It is very important not to draw conclusions or relate events to outcomes that are speculative in
nature. Avoid placing an inappropriate medical diagnosis as a label to clinical findings. The
members of the team will erroneously draw upon speculative causal relationships and refer to
them further in the record as an absolute. Applying an outcome, diagnosis, or finding from
incomplete data does not lessen the impact of such a premature diagnosis during the course of
either treatment or litigation. Concluding, for example, based on insufficient data that the
encephalopathy found on a computerized tomographic scan is hypoxic encephalopathy and that
the hypoxic encephalopathy is related to a specific hypotensive episode serves no meaningful
clinical purpose and may lay the foundation for an accusation of medical negligence on those
purported to be responsible for causing or failing to detect that specific hypotensive episode when
the abnormal finding may not be secondary to hypoxic encephalopathy in the first place.
Imposition of unsubstantiated diagnoses is a frequent occurrence during follow-up care of highrisk infants and children. Recognizing that the medical records of infants discharged from an NICU
may be voluminous, if a diagnosis cannot be confirmed by review of records, it should not be
made in a follow-up summary.
P.108
Clinical findings, if possible, are best graded using numeric values. Modifiers, such as the words
extreme, severe, or massive do not provide objective information and will feed into the
drama of a courtroom presentation. Contributors to the evaluation and care of the patient must
avoid inflammatory language, markings, or punctuation that attempt to draw attention to the
writer and imply that the team may be inattentive to the remarks otherwise.
TABLE 8-2 STEPS IN THE CONSULTATION PROCESS
Identify when it is appropriate to obtain consultation

Choose the type of consultant (sequencing of a variety of consultants)
Share the purpose of the consultation with the family
Decide what information is necessary to provide to the consultant, including the purpose of
the consultation and provide as a written consultation request
Arrange a formal consultation or designate this responsibility to a specific individual
Give a time frame within which the consultation is to be done
Agree with the consultant on how the consultation will be reported and who will discuss the
findings with the family/parent(s)
Share the outcome of the consultation with the family
Tell the family what will be done with the resulting information
Document in the medical record the reasons for requesting a consultation, discussions with
the family, and plan of treatment, including the reasons why the recommendations of the
consultant will or will not be followed
It is possible that an infant's condition is changing so rapidly that thorough, contemporaneous

notes in the medical record cannot be made. Although not documented, not done (111) is a wellknown legal motto, there is much that occurs that cannot be documented in the patient record at
the time of a critical event. It is often hindsight that focuses the care provider on events not
recorded in contemporaneous records. A decision must be made as to whether an addendum to
the medical record would contribute to patient care, or whether the additional information relates
to system issues and should be conveyed to the risk manager by completing an incident report.
An incident report is a confidential document that provides a means to communicate system
issues. References to the completion of the incident report or discussions with the risk manager
are not appropriate for, and should not appear in, the medical record (110). An incident report
relates to an adverse event; the patient's record reflects care and treatment delivered, the
patient's response, and the further plan of care.
When additional information, identified after an adverse event, is necessary to the care of the
patient, communication of this information is achieved in an addendum to the medical record.
Such entries must include the date and time that they were written. Under no circumstances
should an attempt be made to make this entry appear contemporaneous with the adverse event.
The addendum must be placed sequentially in the medical record in the next available space, not
on a separate page, and never inserted into the text of previous entries. The purpose of an
addendum is not to provide a defense or to make excuses.
Communication with Consultants

The curbside consultation between colleagues is a thing of the past. This type of consultation
tends to be documented poorly, if at all, and can lead to damaging finger pointing in a court of
law. Guidelines for the consultation process are shown in Table 8-2.
When a consultant is contacted for a specific patient with a formal request for evaluation and/or
management, both the primary physician and consultant should interact in such a way as to limit
unnecessary risk for both (111).
CONCLUSIONS
Almost coincident with publication of the fifth edition of this textbook (at the end of 1999), the
IOM published their landmark report on the quality of health care in America (92). The reported
finding that between 44,000 and 98,000 patient deaths per year are due to medical errors placed
this as the eighth leading cause of death in the United States. In their report, the IOM estimated
that at least 50% of these errors were preventable and challenged the health care system to
reduce medical errors by 50% in the next 5 years. A further report entitled Crossing the Quality
Chasm: A New Health System for the Twenty-First Century was published 2 years later in 2001 by
the IOM (112) in which they listed six redesigned imperatives for the health care system in the
United States:
redesign of the care process
use of information technologies
knowledge and skills management
development of effective teams
coordination of care
use of performance and outcome measurements
As the sixth edition of this textbook goes to press, there is abundant evidence that the IOM
challenge has been noted and that the mouth of the quality chasm has been explored. There is
little evidence however, at the 5-year anniversary of the IOM challenge that construction of a
sound bridge across the quality chasm has progressed beyond the sinking of pylons.
REFERENCES
1. Witman AB, Park DM, Hardin SB. How do patients want physicians to handle mistakes? A
survey of internal medicine patients in an academic setting. Arch Intern Med 1996;156:2565
2569.
2. Gallagher TH, Waterman AD, Ebers AG, et al. Patients' and physicians' attitudes regarding
the disclosure of medical errors. JAMA 2003;289:10011007.
3. Setting the Standard. Oakbrook Terrace, Ill: Joint Commission on Accreditation of Healthcare
Organizations (JCAHO) 2003.
4. Eckert v Long Island R.R., 43 NY 502 (1871).

P.109
5. Kambat v St. Francis Hospital, 89 NY2d 489, 678 NE2d 456, NYS2d 844 (1997).
7. Keeton WP, Dobbs DB, Keeton RE. Prosser and Keeton on the law of torts. 5th ed. Eagan,
Minn: West Group, 1984:356.
8. Mancini MR, Gale AT. Emergency care and the law. Rockville: Aspen Systems Corp., 1981:50.
9. Frew SA, Roush WR, LaGreca K. COBRA: implications for emergency medicine. Ann Emerg
Med 1988;17:835837.
10. Hess v Paulowski, 279 US 352 (1927).
11. Diversity of Citizenship; amount in controversy; costs, 28 USC 1332.
12. Friedman RH, Stollerman JE, Mahoney DM, et al. The virtual visit: using telecommunications
technology to take care of patients. J Am Med Inform Assoc 1997;4:413425.
13. Landwehr JB Jr, Zador IE, Wolfe HM, et al. Telemedicine and fetal ultrasonography:
assessment of technical performance and clinical feasibility. Am J Obstet Gynecol
1997;177:846848.
14. Lewis M, Moir AT. Medical telematics and telemedicine; an agenda for research evaluation in
Scotland. Health Bull (Edinb) 1995;53:129137.
15. Landquist A. Finland is a leading country when it comes to telemedicine. Lakartidningen

l996;93:78.
16. Telecommunications Reform Act, S. 652, 104th Cong. 709 (1996).
17. S. 773, 1995 Reg. Sess. (La. 1995).
18. Cal. Assembly 1665 (1996).
19. S. 48 (Okla. 1997).
20. H.R. 2033 (Tex. 1997).
21. H.R. 177 (Ky. 2000).
22. Telemedicine Report to Congress, January 31, 1997. Legal Issues, Licensure and
Telemedicine. Available at: http://www.ntia.doc.gov/reports/telemed/.
23. Available at: http://www.napb.net/vipps/intro.asp.
24. Jackson v Johns-Manville Sales Corp., 727 F2d 506,516 (5th Cir 1984).
25. McCormick C. McCormick on evidence, 2nd ed. St. Paul: West Publishing Co., 1972: 339.
26. Morgan E. Basic problems of evidence, 4th ed. Philadelphia: Joint Committee on Continuing
Legal Education, 1963:24.
27. Terry HT. Negligence. Harv Law Rev 1915;29:40.
28. Zebarth v Swedish Hospital Medical Center, 81 Wash 2d 12,499 (P2d 1 1972).
29. Second Restatement of Torts, Section 282.
30. Ely JW, Levinson W, Elder EC, et al. Perceived causes of family physicians' errors. J Fam
Pract 1995;40:337344.
31. Oelling v Rao, 593 NE2d 189 (1992).
32. Available at: http://www.japantimes.co.jp/cgi_bin/getarticle.p15?ed20020705a1.htm.

Accessed October 28, 2004.
33. Small v Howard, 128 Mass 131 (1880).
34. Peck v Hurchinson, 88 Iowa 320,55 NW 511.
35. Daubert v Merrill Dow Pharmaceuticals, Inc. 509 US 579 (1993).
36. Ybarra v Spangard, 25 Cal App 2d 486, 154 P2d 687 (1944).
37. Mohr v Williams, 104 NW 12, 5 Ct (1905).
38. Canterbury v Spence, 464 F2d 772 (DC Cit 1972), cert denied, 409 US 1064 (1972).
39. Getchell v Mansfield, 260 Or 174,489 P2d 953 (1971).
40. Cooper v Roberts, 286 A2d 647,650 (1971).
41. Davis v Wyeth Laboratories, Inc., 399 F2d 121 (9th Cit 1968).
42. Application of President & Directors of Georgetown College, 331 F2d 1000 (DC Cir), cert
denied, 377 US 978 (1964).
43. King NM. Transparency in neonatal intensive care. Hastings Cent Rep 1992;22:1825.
44. Paasche-Orlow MK, Taylor HA, Brancati FL. Readability standards for informed-consent
forms as compared with actual readability. N Engl J Med 2003;348:721726.
45. Gutheil TG, Bursztajn H, Brodsky A. Malpractice prevention through the sharing of
uncertainty. Informed consent and the therapeutic alliance. N Engl J Med 1984;311:4951.
46. Gillick MR, Advance care planning. N Engl J Med 2004;350:78.
47. Committee on Bioethics, American Academy of Pediatrics. Informed consent, parental

permission, and assent in pediatric practice. Pediatrics 1995;95:314317.
48. Little v Little, 576 SW2d 493 (1979).
49. In Re Richardson, 284 So2d 185 (1973).
50. Medicaid frauda multibillion-dollar crime. Available at: http://www.newsmax.com/

archives/articles/2002/12/30/170723.shtml.
51. U.S. Dept of Health and Human Services and U.S. Dept of Justice Health Care Fraud and
Abuse Control Program. Annual Report For FY 2002. Washington, DC: U.S. Government Printing
Office, 2003.
52. Federal crackdown puts risk managers in hot seat [Editorial]. Health Care Risk Management
1997;19:49.
53. Hayt E, Hayt LR, Groeschel AH. Law of hospital, physician, and patient, 2nd ed. New York:
Hospital Textbook Co., 1952:637.
54. Beck JC. Confidentiality versus the duty to protect: foreseeable harm in the practice of
psychiatry. Washington, DC: American Psychiatric Press, 1990.
55. Reisner v Regents of the University of California 31 Cal App 4th 1195,37 CalRptr2d518
(1995).
56. Tarasoff v Regents of University of California, 108 Cal Rptr 878 (Cal App 1973), superseded
by Tarasoff v Regents of University of California, 13 Cal 3d 177, 118 Cal Rptr 129,529 P2d 553
(1974), subsequent op on reh Tarasoff v Regents of University of California, 17 Cal 3d 425, 131
Cal Rptr 14,551 P2d 334 (1976).
57. Lipari v Sears Roebuck & Co., 497 F Supp 185 (D Neb 1980).
58. 42 USC 290ee-3.
59. 42 USC 290dd-3.
60. 42 USC 290ee-3(b)(2)(A).
61. 42 USC 290dd-31(b)(2)(A).
62. 42 USC 290ee-3(tb)(2)(C).
63. 42 USC 290dd-3(b)(2)(C).
64. Dick RS, Steen EB, eds. The computer-based patient record: an essential technology for
health care. Washington, DC: National Academy Press, 1991.
65. Woodward B. The computer-based patient record and confidentiality. N Engl J Med
1995;333:14191422.
66. Annas GJ. The rights of patients: the basic ACLU guide to patient rights, 2nd ed.
Carbondale: Southern Illinois University Press, 1989:178.
67. Basalmo RR, Brown MD. Risk management. In: Sanbar SS, Gibofsky A, Firestone MH, eds.
Legal medicine, 3rd ed. St. Louis: Mosby-Year Book, Inc., 1995:237.
68. Kramer C. Medical Malpractice 5 (1976), citing History of Reported Medical Professional
Liability Cases, 30 Temple LQ 367 (1957).
69. Pegram v Herdrich, 530 US 211 (2000), 154 F3d 362, reversed.
70. Aetna Health Inc. v Davila, 542 US (2004), 307 F3d 298, reversed and remanded.
71. Malone TW, Thaler DH. Managed health care: a plaintiff's perspective. Tort Insur Law J
1996;32:123153.
72. Clifford RA. Physician's liability in a managed care environment. Health Lawyer 1997;10:5.
73. Bodenheimer T. The HMO backlashrighteous or reactionary? N Engl J Med

1996;335:16011604.
74. Blum JL, ed. Monograph 5, achieving quality care: the role of the law. Health Law Section of
the American Bar Association, Loyola University, Chicago, June 1997.
75. Moore v Regents of the University of California, 793 P2d 479, 51 Cal 2d 120 (1990).
76. Kanute M. Evolving theories of malpractice liability in HMOs. Loy Univ Chic Law Rev
1989;20:841873.
77. Fox v Health Net, Civ No 21962 (Riverside County Super Ct, Cal 1993).
78. Sloan v Metropolitan Health Council, 516 NE2d 1104 (Ind Ct App 1987).
79. Restatement (second) of Agency paragraph sign 216, 1958.
80. Restatement (second) of Agency paragraph sign 250, 1958.
81. Schleier v Kaiser Foundation Health Plan of the Mid-Atlantic States, Inc., 876 F2d 174 (DC
Cir 1989).
82. Darling v Charleston Community Memorial Hospital, 211 NE2d 253 (1965), cert denied, 383
US 946 (1966).
83. Wickline v State, 192 Cal App 3d 1630, 239 Cal Rptr 810 (1986).
84. 29 USC Sections 10011461.
85. 29 USC Section 1144(a).
86. Shea v Esensten, 107 F3d 625 (8th Cir 1997).
87. Harper v Healthsource New Hampshire, 674 A2d 962 (NH 1996).
88. Texas Medical Association v Aetna Life Insurance Co., 80 F3d 153 (5th Cir 1996).
89. 42 USCA 1113111137 (West Supp 1995).
90. American Academy of Pediatrics Section on Perinatal Pediatrics, Committee on Practice

Management. Understanding a practice venue: guidelines and suggestions regarding the nonclinical aspects of neonatology. J Perinatol 2002;22(Suppl 1):S1-S76.
P.110
91. Kohn LT, Corrigan JM, Donaldson MS, eds. To err is human: building a safer health system.
Washington, DC: National Academy Press, 2000.
92. Fortescue EB, Kaushal R, Landrigan CP, et al. Prioritizing strategies for preventing
medication errors and adverse drug events in pediatric inpatients. Pediatrics 2003;111:722
729.
93. Bernstein PL. Against the gods: the remarkable story of risk. New York: Wiley and Sons,
1996:197.
94. Morlock LL, Malitz FE. Do hospital risk management programs make a difference?:
relationships between risk management program activities and hospital malpractice claims
experience. Law Contemp Probl 1991;54:122.
95. Acerbo-Avalone N, Kremer K. Medical malpractice claims investigation: a step-by-step

approach. Gaithersburg, MD: Aspen Publishers, 1997:xv.
96. Stelovich S. Framework for handling adverse events. Forum 2002;22:89.
97. Grupp-Phelan J, Reynolds S, Lingl LL. Professional liability of residents in a children's

hospital. Arch Pediatr Adolesc Med 1996;150:8790.
98. Conducting a root cause analysis in response to a sentinel event. Oakbrook Terrace, Ill:
Joint Commission on Accreditation of Healthcare Organizations (JCAHO), 1996.
99. Andrews LB, Stocking C, Krizek T, et al. An alternative strategy for studying adverse events
in medical care. Lancet 1997;349: 309313.
100. Brennan TA, Sox CM, Burstin HR. Relation between negligent adverse events and the
outcomes of medical-malpractice litigation. N Engl J Med 1996;335:19631967.
101. Taragin MI, Willett LR, Wilczek AP, et al. The influence of standard of care and severity of
injury on the resolution of malpractice claims. Ann Intern Med 1992;117:780784.
102. Pichert JW, Hickson GB, Bledsoe S, et al. Understanding the etiology of serious medical
events involving children: implications for pediatricians and their risk managers. Pediatr Ann
1997; 26:160172.
103. Van Cott H. Human errors: their causes and reduction. In: Bogner MS, ed. Human error in
medicine. Hillsdale, NJ: Lawrence Erlbaum Associates, 1994:153.
104. Chess C, Hance BJ, Sandman PM. Improving dialogue with communities: a short guide to
government risk communication. New Jersey Department of Environmental Protection, 1988.
105. Levison W. Doctor-patient communication and medical malpractice: implications for

pediatricians. Pediatr Ann l997;26: 186193.
106. Cuttini M, Romito P, Del Santo M, et al. Communication in the neonatal intensive therapy
unit: the opinions of parents and of medical personnel compared. Pediatr Med Chir 1994;16:
325329.
107. Covello VT, Allen FW. Seven cardinal rules of risk communication. Washington, DC: U.S.
Environmental Protection Agency, Office of Policy Analysis, 1988.
108. Silverman WA. Retrolental fibroplasia: a modern parable. New York: Grune and Stratton,
Inc., 1980:83.
109. Donchin Y, Gopher D, Olin M, et al. A look into the nature and causes of human errors in
the intensive care unit. Crit Care Med 1995;23:294300.
110. Chilton JH, Shimmel TR. Inappropriate word choice in the labor and delivery and newborn
medical record. In: Donn SM, Fisher CW, eds. Risk management techniques in perinatal and
neonatal practice. Armonk, NY: Futura Publishing, 1996:603.
111. Hartline JV Smith CG. Risk management in medical consultation. In: Donn SM, Fisher CW,
eds. Risk management techniques in perinatal and neonatal practice. Armonk, NY: Futura
Publishing, 1996:617.
112. Committee on Quality of Health Care in America. Crossing the quality chasm: a new health
system for the 21st century. Washington, DC: National Academy Press, 2001.
Chapter 9
The Vulnerable Neonate and the Neonatal Intensive
Care Environment
Penny Glass
Environmental factors in the neonatal intensive care unit (NICU) have major implications for the care of the sick
newborn infant. Advances in medical technology during the last 3 decades are credited with dramatic reductions
in mortality, with a 50% survival rate for newborns weighing 1,500 g in 1970 to a 50% survival rate for those
weighing less than 700 g by 2000. Morbidity among survivors, however, is a problem of increasing proportions.
Whereas the rate of major morbidity has remained fairly stable, around 10%, this focus on major morbidity has
overlooked the much larger number of children born prematurely who have learning disabilities at school age.
Broad evidence implicates the environment in the NICU as a factor in neonatal morbidity. Abnormal sensory
input can be a source of potentially overwhelming stress and, at a sensitive period during development, can
modify the developing brain. The NICU environment, therefore, assumes a crucial role in the care of the sick
newborn infant.
Preterm birth is the most common single risk factor for developmental problems in childhood, and learning
disability is the most pervasive developmental problem. This is a catch-all term, but includes children of low,
average, or otherwise normal intelligence who have deficits in language, visual perception, or visuomotor
integration; deficiencies in attention span, hyperactivity; or social immaturity. Such children require either
special services to function in a regular classroom or placement in a special class. Reports of school-age children
who were of very-low-birth-weight indicate that as many as half have learning disabilities (1,2,3,4,5,6,7,8).
Such deficits may originate from overt damage to the brain or from a more general disturbance in brain
organization.
Throughout infancy, both behavioral and neurologic differences exist between full-term and preterm infants,
even when matched for conceptional age. The latter often exhibits manifestations of altered brain organization,
including disrupted sleep, difficult temperament, both hyperresponsivity and hyporesponsivity to sensory input,
prolonged attention to redundant information, inattention to novel stimuli, and poor quality of motor function
(9,10,11,12,13,14,15). These precursors of learning problems in school are not fully explained by either the
severity of illness among the preterm infants or by later conditions in the home environment (10).
The sensory environment in the NICU is different in virtually every respect, both from the environment of a
fetus in utero and from that of a full-term newborn at home. The NICU experience also contains frequent
aversive procedures, excess handling, disturbance of rest, noxious oral medications, noise, and bright light.
These conditions are sources of stress and anomalous sensory stimulation, both of which may affect morbidity.
The immediate effects of stress are autonomic instability, apnea/bradycardia, vasoconstriction, and decreased
gastric motility. Cortisol, adrenaline, and catecholamines are secreted during stress as part of an intricate
hypothalamic-pituitary-adrenocortical system (16,17). High levels of these hormones interfere with tissue
healing. Noxious stimuli disrupt sleep and can have biological consequences for the neonate. Even medical
complications commonly associated with prematurity per se, such as bronchopulmonary dysplasia and
necrotizing enterocolitis, may be, in part, stress-related diseases (18).
Sensory input is essential during maturation. Most of the cortex is part of one of the sensory systems. Abnormal
experiences, both depriving and overstimulating, can modify the developing brain. The most vulnerable period
occurs during rapid brain growth and neuronal differentiation
P.112
(19,20). The timing of these events for the human fetus corresponds to 28 to 40 weeks of gestation (21). It is
assumed that, for the fetus, the optimal sensory environment is experienced within the womb. One of the most
striking aspects of this environment is the bidirectional contingency between mother and fetus.
The potential impact of the anomalous NICU environment on the vulnerable newborn infant has raised unabated
concerns for more than 2 decades (22,23,24,25,26,27,28). A more optimal NICU environment might reduce
iatrogenic morbidity and improve the outcome of sick neonates; however, the parameters are not yet well
defined. This chapter summarizes the maturation of each sensory system during late fetal development, with
particular reference to evidence for the prenatal onset of function, compares the intrauterine and NICU sensory
experience, and critiques techniques of developmental intervention.
NEONATAL SENSORY SYSTEMS: DEVELOPMENT, DISORDERS,

ENVIRONMENT, AND INTERVENTION
Maturation of all the sensory systems begins during the latter part of embryogenesis; however, the process is
neither unitary nor fixed. Within each system some reciprocity between structure and function probably exists.
To some extent, sensory input drives maturation (29). In addition, the rate of maturation of each sensory
system varies, with the onset of function generally in the following order: tactile, vestibular, gustatory-olfactory,
auditory, and visual (30). These sensory systems also are interrelated in a hierarchical mannerstimulation of
early maturing senses (e.g., tactile, vestibular) has a positive influence on development of later-maturing ones
(e.g., visual) (31). Recent research also indicates that untimely stimulation within this sequence (e.g., visual)
may disrupt the normal maturational process of another sensory system (e.g., auditory) (Philbin MK. personal
communication, 1998.).
This hierarchical organization and integration of sensory function is well supported and bolsters two guiding
principles for developmental intervention in the NICU:
Stimulation of the senses should begin with the most mature.
The optimal form of stimulation for initial postnatal development resembles the sources naturally available
to the fetus and infantthose that come from the mother.
Tactile System
The cutaneous system includes sensation of pressure, pain, and temperature. Only pressure is discussed here;
pain is discussed in Chapter 57. Receptors in the skin respond to pressure and then transmit impulses to the
spinal cord through the dorsal root, ascending in the posterior tract and terminating in the gray matter of the
cord. At this point, connecting fibers decussate and continue in the ventral spinothalamic tract to the medulla
and the thalamus, terminating in the postcentral gyrus of the cortex. Representation here is somatotopic and
contralateral to the stimulated side. Increased stimulation to an area of the body or loss of a limb can alter the
pattern of representation in the somatosensory cortex.
Development
Like the vestibular system, the tactile sense develops early in fetal life and is thought to play a particularly
pervasive role in the early development of the organism. Receptor cells are present in the perioral region in the
fetus by 8 weeks of gestation and spread to all skin and mucosal surfaces by 20 weeks. The cortical pathway is
intact by 20 to 24 weeks of gestation, and some myelin is already present. Response to tactile stimulation has
been observed by ultrasound as early as 8 weeks of conceptional age (32). Response to stroking in the lip
region occurs first, followed by a response to stimulation of the palms. Most of the body is sensitive to touch by
15 weeks (33).
Tactile threshold is very low in the preterm infant. It is more related to postconceptional age (PCA) than to natal
age but increases by term. Infants younger than 30 weeks PCA respond by an unequivocal leg withdrawal to
pressure of a 0.50-g von Frey hair applied to the plantar surface of the foot compared to 1.7-g pressure by 38
weeks PCA (34). A qualitative shift occurs around 32 weeks PCA. Infants less than 32 weeks PCA respond to
repeated stimulation with sensitization and a diffuse behavioral response. In contrast, infants after this age
show habituation to the same stimuli.
Classic studies by Harlow and Harlow (35) demonstrated the profound importance of contact comfort for normal
development. In a parallel fashion, even preterm infants will seek and maintain contact with a physical object
within their incubator and even more so if the tactile source contains rhythmic stimulation (36). These findings
provide strong support for intervention in the tactile modality.
Disturbances
Tactile hypersensitivity, or tactile defensive behavior, is contained in clinical reports of children with
developmental delay, many of whom were born preterm. It also is seen in infants and children who otherwise
appear normal. The behavior frequently is said to be a manifestation of sensory integration deficit and thought
to have its origins in the prenatal or perinatal period. It appears as an infant's overreaction to touch, generally
the hands or oral-facial regions. With oral hypersensitivity, the infant may withdraw, gag, or retch when
touched, even around the outside of the mouth. Some infants are intolerant of food with texture and resist
transition from liquids or very smooth puree. Infants also may be hypersensitive to touch on their extremities,
with prolonged palmar-mental reflex, exaggerated
P.113
hand and toe grasp, or leg withdrawal. An extreme case was a 2-month-old (corrected age) infant who, when
supine, arched his buttocks off the table surface in response to his legs being grasped. Additional manifestations
of tactile sensitivity may appear as an intolerance for grasping toys or handling play materials of certain
textures. In another extreme example, a 1-year-old infant would gag when his hand was placed in dry
macaroni. Some children are intolerant of normal clothing and even may avoid body contact. Such aversion
adversely affects parent-infant bonding. The link between early tactile disturbances and learning disabilities at
school age is unlikely to be a causal one, but may be related to a similar mechanism of brain dysfunction.
Intrauterine Experience
The fetus is housed in a thermoneutral, fluid-filled space that is a source of cutaneous input throughout the body
surface. Fetal movement provides tactile self-stimulation. Perhaps even more important, fetal movement often
evokes a contingent maternal response. As term approaches and the intrauterine space becomes more
constraining, the normal posture of flexion evokes hand-to-mouth, skin-to-skin, and body-on-body tactile
feedback. The effect is progressive throughout gestation.
After a normal term birth, a ventral-to-ventral position is preferred by both mother and infant, with touch
followed by slow stroking (37). Traditionally, the infant is then swaddled and held. As before birth, human
proximity produces contingent touch.
Touch And Handling in the Neonatal Intensive Care Unit

After premature birth, tactile input is radically altered. The extrauterine fetus is frequently nursed naked, with
the exception of maybe a diaper and a hat. The surface of the mattress generally is unyielding. She or he is
exposed to air currents, cold stress, tape, instruments, handling by caregivers, and painful stimuli. Pressure is
not uniform.
The type and frequency of tactile stimulation imposed on a preterm newborn in the NICU would be
overwhelming even for a healthy adult. During a 2-week period, a sick neonate may be handled by more than
10 different nurses, in addition to physicians, occupational or physical therapists, laboratory and x-ray
technicians, and, finally, the parents (38).
Handling occurs more often among the sickest infants, typically is related to procedures, generally is disturbing,
and often is painful. In spite of wide attention in the literature to the consequences of excess handling in the
NICU, the amount had not decreased from 1976 to 1990, even among the more severely ill infants (39). On the
average, sick preterm newborns are handled more than 150 times per day, with less than 10 minutes of
consecutive uninterrupted rest (26). Disturbance of sleep has biologic and immunologic consequences (40,41).
Secretion of cortisol and adrenaline normally is inhibited during sleep. Growth hormone, which is released during
quiet sleep, increases protein synthesis and mobilization of free fatty acids for energy use. Thus, sleep facilitates
healing.
Excess handling has other significant physiologic consequences for the sick neonate. Blood pressure changes,
alterations in cerebral blood flow, and episodes of oxygen desaturation are associated with noxious procedures,
handling, or crying (42,43,44,45,46). Fluctuations in blood pressure may contribute to intracranial hemorrhage
in the unstable preterm infant (47). Importantly, when caretakers monitor the infant's level of oxygenation
during procedures, the severity of hypoxemic episodes can be reduced significantly (42).
In addition to the hemodynamic impact of obviously noxious procedures described earlier, more benign
manipulations, such as those that occur during neurodevelopmental assessment, also may adversely affect the
preterm infant (25,46). Decreased plasma growth hormone has been reported after administration of the
Brazelton Neonatal Behavioral Assessment Scale to preterm infants at 36 weeks PCA (48). Even at the time of
discharge, the evaluation was associated with elevated cortisol levels (17,49). It is not clear whether these
effects were from the neurodevelopmental assessment or from the stress associated with crying, which normally
occurs during administration of the Neonatal Behavioral Assessment Scale. Thus, handling could be stressful
even for stable preterm infants.
Tactile Intervention in the Neonatal Intensive Care Unit

The two general approaches to tactile intervention in the NICU provide either reduction of general handling or
provision of planned touch experiences. Touch may be pressure alone or may include stroking. The neonates
may be acutely ill or medically stable. These distinctions are important.
As part of an individualized approach to developmental care for acutely ill preterm infants, Als and colleagues
(50) provided minimal handling and clustering of routine procedures, as well as positive tactile input and
containment from bunting, rolls, positioning, and the like. Short- and long-term outcomes were improved for her
intervention group compared to a nonintervention group. This technique has had a major impact on
developmental intervention as a whole. Although it is not possible to isolate which aspect of this multimodal
approach was effective, a significant change was made in the role of the nurse/caregiver.
More specific to the tactile modality, Jay (51) evaluated the effects of planned, gentle, touch for 12-minute
periods four times a day for acutely ill preterm infants. This intervention, which consisted of hands-on contact
but not stroking or manipulating, was associated with a lower fraction of inspired oxygen (FIO2) after 5 days
compared to a similar nonintervention group.
P.114
In apparent contrast, Field and colleagues (52) and Scafidi and colleagues (53) initiated a touch intervention
that differed from that of Jay (51) and Als and colleagues (50) in several important respects. The infants were
recruited when stable and growing rather than during an acute period.
The treatment provided three 15-minute periods a day of massage (i.e., stroking and passive limb movement)
for a 10-day treatment period. The massaged infants showed a greater weight gain, even though the groups did
not differ in formula intake. These early effects on growth are thought to be mediated by induction of
catecholamine release (48). The massaged infants also spent more time awake, showed better performance on
the Brazelton neonatal assessment, were discharged to home 6 days sooner, and had better performance on
developmental assessment at 8 months past term.
Although overall effects of the tactile intervention have been positive, the response of individual infants is
variable and deserves attention. For example, an increase in periods of oxygen desaturation during parent
touch, compared to baseline, has been reported (54). A number of infants have responded with apnea and
bradycardia to the type of intervention described by Field and colleagues (52) and Scafidi and colleagues (53).
This physiologic response may occur after the intervention. It is not clear whether the intervention itself was
excessive or whether the abrupt shift after the cessation of tactile input led to an unexpected physiologic
response. Individual differences exist among caregivers that also affect the infants' responses to the stimulation.
Finally, the distinction between touch, stroking, and massage is probably important.
All of these issues advise against standardized protocols of stroking or massage for all preterm infants. When
any form of tactile and kinesthetic intervention is applied, the caregiver should continue to monitor the infant's
physiologic and behavioral responses before onset, during, and after the cessation of the intervention.
Consistency of contact across caregivers would help.
Soft swaddling or clothing may provide tactile input in a more sustained fashion than periodic hands-on contact.
Arguments that this obscures the view of the infant and interferes with temperature regulation by servocontrol
simply argue against servocontrol rather than against covers. Clothing or swaddling actually could dampen the
fluctuations in temperature that occur during incubator care and inhibit exaggerated movement or agitation,
which in itself may be stabilizing. Creative swaddling with crisscrossed strips and Velcro could allow for more
visualization of the infant. The use of swaddling in the NICU is necessary if the parent is to retain that option at
home as a means of calming and supporting sleep. Swaddling is difficult to initiate after a period of not being
swaddled.
Thus, the amount and type of stimulation are important and change with acuity, maturation, and the response
of the infant. Parents need specific guidance and modeling from the beginning. The general order of tactile
intervention might be:
If acutely illminimal handling, containment (e.g., swaddling, rolls), and gentle touch (e.g., warm hand)
without stroking
When medically stableholding, rocking gently, stroking, contining to swaddle
Minimal handling protocols have a definite place in the NICU, but not as the end point. It also is important
during the hospital stay to help the infant develop increased tolerance for social contact and gentle handling,
especially as discharge approaches. Systematic desensitization may even be necessary in cases of chronically ill
infants.
Nonnutritive Sucking
Nonnutritive sucking is an important oral-tactile intervention that supports both feeding and early behavioral
regulation. It represents an early endogenous rhythm and a manifestation of sensorimotor integration (55). As
such, it is reported in the fetus (56) and observed in the preterm newborn before 28 weeks of gestation. The
number of sucks per burst increases with maturation, whereas the duration of burst is fairly stable across ages.
Nonnutritive sucking experience may facilitate important physiologic and behavioral mechanisms and potentially
reduce cost of care. Infants provided with nonnutritive sucking during gavage feeding showed significantly
improved gastrointestinal transit time, greater suck pressure, more sucks per burst, and fewer sporadic sucks.
They initiated bottle-feeding earlier, showed better weight gain, and thereby had shorter hospital stays (57,58).
Having a pacifier continuously available, however, may not be beneficial and may, in fact, encourage
inappropriate sucking patterns, particularly in the chronically ill neonate.
Nonnutritive sucking also acts as a behavioral organizer or facilitator. It has been shown to decrease motor
activity and increase quiet states in stable preterm infants (59). It dampens an infant's behavioral response
after a painful procedure such as circumcision or heelstick (60,61), although it does not appear to dampen the
cortisol response (17). It is noteworthy that sucking on a pacifier before the onset of repeated painful
procedures, such as heelsticks, may be inappropriate, because aversive conditioning to the pacifier could occur.
Nipples used for nonnutritive sucking abound, varying in size, configuration, consistency, and utility. A feeding
nipple is not designed for nonnutritive sucking and is inappropriate. It readily collapses; the infant experiences
little resistance to his or her suck and may loll the device. Gauze inserted in the nipple may absorb oral
secretions and breed bacteria. In larger infants, the nipple is unsafe because of possible aspiration. A variety of
commercially available pacifiers should be available in any NICU to suit the individual needs of each infant.
Some neonates who are hypersensitive to touch in the perioral region often respond positively to contact (and
perhaps smell) from their own hands.
P.115
Vestibular System
The vestibular system, situated in the nonauditory labyrinth of the inner ear, responds to movement as well as
directional changes in gravity. The three fluid-filled semicircular canals, one for each major plane of the body, lie
at right angles to each other. The ampulla, located at the end of each canal, contains hair fibers in a sac, or
cupula. Motion of the body or head causes pressure changes that move the cupula, which stimulates the hair
cells and transmits an impulse along the vestibular portion of the eighth cranial nerve to the vestibular nuclei of
the medulla. The vestibular organs consist of the utricle and saccule, which respond to changes of head position
involving linear motion. The macula, a thickening in the wall of the utricle and saccule, contains hair cells
sensitive to the position of the head. Impulses from the macula transmit along the vestibular nerve to the
medulla and cerebellum. From there, information is transmitted to motor fibers going to the neck, eye, trunk,
and limb muscles. There are no connections to the cortex (63). Vestibular stimulation affects level of alertness.
Slow, rhythmic, continuous movement induces sleep. Periodic or higher amplitude swing increases arousal.
Development
Initial vestibular development is concurrent with auditory development, emanating from the same otocyst early
in gestation. The three semicircular canals begin to form before 8 weeks of gestation, reaching morphologic
maturity by 14 weeks, and full size by week 20 (29). The vestibular sacs probably develop at the same time.
Response to vestibular stimulation has been observed by 25 weeks of gestation (33). The traditional vertex
presentation of the fetus at term gestation is thought to occur from fetal activity in response to vestibular input.
Disturbances
Considerable research with animals has demonstrated the importance of both tactile and vestibular input
(35,63). Lack of normal vestibular stimulation in the developing organism is thought to affect general
neurobehavioral organization (31). Children who were born preterm are reported to have deficits in balance at
preschool age (8), but this is not necessarily a vestibular problem.
The fetus experiences both contingent and noncontingent vestibular stimulation that varies during gestation.
From the beginning of embryonic life, the fluid environment of the womb provides periodic oscillations and
movements that emanate from normal movements of the mother as well as activity of the fetus itself. Reports
by mothers of fetal movement occur around 16 weeks. After 28 weeks of gestation, there is a decrease in the
relative amount of amniotic fluid, and, thus, the movement of the fetus becomes partially constrained by the
more limited physical space. Vestibular experience is then less contingent on self-activation and more related to
normal maternal activity and position change, which often occurs in response to fetal activity. In general,
maternal activity level slows as parturition approaches.
After birth, the infant is held normally. Movement is slow from maternal breathing and shifting. Change of
position is gradual, even by experienced parents. Vestibular stimulation is used to affect statemoving to
upright or laying down increases arousal; monotonous side-to-side rocking and walking in the form of parental
pacing reduce the level of arousal.
Vestibular Experience in the Neonatal Intensive Care Unit

Vestibular stimulation after preterm birth is limited to efficient manipulation or turning of the neonate by the
caregiver. It clearly lacks any of the temporal qualities or contingencies that the maternal environment may
have provided. Spontaneous limb movement generally is diffuse, often unrestricted, and typically disorganizing
in its effect.
Intervention in the Neonatal Intensive Care Unit

Like the tactile sense, the early development of the vestibular system provides a theoretical basis for primary
intervention with preterm neonates. More than 3 decades ago, Neal (64) demonstrated that daily rocking
facilitated the development of preterm infants. Subsequent research simulated the intrauterine environment and
provided compensatory vestibular stimulation. An oscillating waterbed was devised, which moved with the
rhythm of maternal respirations but with an amplitude of less than 2.5 mm at the surface of the unoccupied
waterbed. The safety of this paradigm, as well as efficacy in reduction in apnea of prematurity, has been well
demonstrated (65). In addition, the infants on waterbeds demonstrated more organized sleep state and motor
behavior, decreased irritability, enhanced visual alertness, and improved somatic growth (65,66,67,68).
Difficulty in nursing sick infants on an oscillating surface may have precluded more widespread adoption of the
waterbed. An infant's own movement also can induce more than optimal movement of the surface, as, for
example, an infant with gastroesophageal reflux; however, some babies still are likely to benefit. At the least, it
would seem prudent to provide a trial on a waterbed for a nonventilated infant with apnea of prematurity,
before introduction of pharmacologic intervention.
Other sources of vestibular stimulation, such as rocking chairs, swings, and hammocks, have not been
investigated formally. Rocking chairs probably belong in any nursery. Swings are questionable, given the
excessive upright position of the baby and the standard rate of oscillation (i.e., too fast). A crib has been
devised that provides controlled
P.116
motion similar to a woman walking; the duration of motion may be individually controlled and proportionally
reduced over time (69), but the rate of oscillation appears too rapid for a preterm infant. The device appears to
be effective in modulating fussiness in full-term infants and is used with preterm infants. The infants are well
swaddled and further contained on each side by rolls.
Positioning
The physical position of an infant is part of the NICU tactile-vestibular experience. Nursing sick preterm infants
routinely has been with the infant in the supine position and exposed, which may simplify management but may
not be advantageous for the infant. Prone positioning in the NICU has been strongly supported physiologically.
The current NICU dilemma is that the prone sleep position is contrary to the recommendation by the American
Academy of Pediatrics (AAP), which now supports supine positioning because epidemiologic data associate
supine positioning with a lower rate of sudden infant death syndrome. The optimal position of the infant needs
to address anatomic and physiologic consequences. Positioning for optimal care in the NICU needs to take
account of the AAP recommendation before the infant is ready for discharge to home.
Yu (70) demonstrated that gastric emptying was facilitated in either the prone or right lateral position compared
to the supine or left lateral position. This was particularly significant for the sick preterm who already showed a
delay in gastric emptying. The prone position, compared to supine, is associated with more quiet sleep and less
active sleep or crying. Quiet sleep, in turn, is associated with improved lung volume, more stable respiration,
less apnea, and improved PAO2 (71,72). Finally, the prone position compared to supine is associated with a
higher PAO2 among healthy preterm infants and, even more significantly, in those with respiratory distress
syndrome (72,73). The evidence suggests that, when possible, the sick infant should be nursed in a prone or
right lateral position.
Parents of preterm infants often complain that their baby's feet turn out. In fact, the legs more often are
externally rotated at the hip. Grenier (74) described hip deformities seen on x-ray of preterm infants after
prolonged nursing in a frog-leg position. The bulk of the diaper in extremely preterm infants exacerbates the
problem. Winging of the scapula also is frequent in the preterm infant. Proper support of the trunk and limbs in
the prone or supine position lessens this extreme rotation and may diminish orthopedic or neuromuscular
complications.
In the prone position, placing the infant on a small folded strip from shoulder to hip, could allow more
physiologic flexion and adduction. In side lying, it may be easier to position the infant in soft flexion. Gentle
containment of the limbs usually can be managed with strips of soft cloth across the upper arm and thigh. Some
movement should be allowed within a controlled range. A posture of physiologic flexion and adduction in the
supine position requires swaddling. Maintenance of postures can be facilitated by nesting the infant in soft rolls,
but the rolls must not reach above the level of the shoulder. Rolls placed at the buttocks don't allow for
adequate leg/hip extension. Each posture should facilitate the infant bringing hands to mouth.
For older, more medically stable preterm infants, infant seats are used as an alternative to continuous lying in
bed. The infant should be swaddled and nested, the angle probably no greater than 30 degrees, and the length
of time should be limited. Oxygen desaturation has been reported in stable preterm infants placed in car seats.
Kangaroo Care
Kangaroo care is a technique that evolved primarily in South America (75). Traditionally, the infant is clad only
in a diaper and placed under the mother's clothing between her breasts, remaining there according to the
mother's comfort, and feeding on demand. The technique provides fairly sustained multimodal stimulation:
tactile, vestibular, proprioceptive, olfactory, and auditory. It appears to be safe for larger preterm infants or
those who are medically stable. Temperature regulation in the infant does not appear to be a problem, but
needs to be carefully monitored on an individual basis. It seems to have the greatest benefit in terms of
facilitating and maintaining lactation and enhancing maternal sense of competency for these infants. The
studies, however, are of insufficient sample size to evaluate whether morbidity, such as intracranial
hemorrhage, is increased. More data are needed among medically stable infants before kangaroo care should be
attempted prior to 32 weeks conceptional age or with infants requiring mechanical ventilation. The increased
tactile stimulation and additional handling easily could be overly stressful for the immature or sick infant.
Chemical Senses
The chemoreceptors include taste and olfaction. Taste receptors are in the taste buds, which are located
primarily in the papillae of the tongue but also are found on the soft palate and epiglottis (29,76). Taste stimuli
(i.e., sweet, sour, bitter, salt) transmit to the brainstem with a primary branch to the hypothalamus. Cortical
regions are involved in learned taste preferences. The olfactory receptors are located in the lining of the
olfactory epithelium in the posterior portion of the nasal passage. The afferent pathway has no cortical
projection area, but is direct to the limbic system. Olfaction plays an important part in gustatory experiences.
Olfaction also is an integral part of infant attachment to the caregiver and may even be mutual (77).
Development of Taste
The chemoreceptors are well developed within the first trimester (29,76). Taste buds appear around 8 to 9
weeks of gestation. The receptors are present at least by week 16
P.117
of gestation, and increase by term to adult levels. During the second half of gestation, morphologic changes
occur that continue after term. Taste receptors are functional before birth. Injection of distinct tastes into the
amniotic fluid of pregnant women between 34 and 39 weeks of gestation alter fetal swallowing behavior, which
increases with the sweeter taste and decreases with bitter taste (29).
Taste discrimination has been measured by differential consumption, autonomic responses, and the presence of
characteristic facial expressions. Plain water evokes an aversive response, which may be a biologically based
protective mechanism. Taste is sufficiently sensitive at term to detect a 0.1 mol/L-concentration of NaCl in water
(78). Full-term neonates, even anencephalic infants, demonstrate differential behavioral responses to sweet,
bitter, sour, and salt (79). In a behavior described as savoring, normal newborns discriminate between
different concentrations of sucrose and even among various sugars (78). Preterm infants (30 to 36 weeks of
gestation) show stronger sucking in response to glucose, compared to plain water, and characteristic behavioral
expressions in response to sour or bitter solutions (80). Behavioral response to formula, or breast milk,
administered to the tip of the tongue has been documented in preterm infants prior to 28 weeks of gestation
(Zorc L. unpublished doctoral dissertation, 2000).
Stimulation of taste receptors has important implications for early feeding and behavioral regulation.
Smotherman and Robinson (81) hypothesized that tastes of milk activate a centrally mediated endogeneous
opioid system in newborn term infants, consistent with that shown in the animal model. This would suggest that,
in normal development, the mechanism to support early feeding extends beyond maintenance of chemical or
caloric balance and becomes feeding to thrive.
Development Of Olfaction
The human olfactory system is composed of four anatomically distinct but integrated subsystems:
The main olfactory
The vomeronasal
The terminal
The trigeminal
Each of these differentiates very early in gestation and is nearly mature prior to term birth (29,82). Epithelia of
the main olfactory system are evident around 5 weeks in the apical part of the nasal cavities. Nerve fibers and
cells form the olfactory nerve, which links the epithelium to the main olfactory bulbs and then to the ventral wall
of the forebrain. The vomeronasal system consists of bipolar sensory cells in the lower nasal septum, with axons
terminating in the accessory olfactory bulb. The terminal system is comprised of free nerve endings in the
anterior part of the nasal septum. The nerve endings of the trigeminal system, which originates from cranial
nerve V, are diffusely distributed in the nasal cavity. The trigeminal system, although activated by touch, may
be the earliest functioning chemoreceptor.
No information exists about the functional onset of human olfaction, but it is presumed to be present prenatally,
having been demonstrated in a rat model. Rat fetuses exposed to citral in the amniotic fluid will selectively
attach postnatally to a nipple of the same scent (83).
Human prenatal olfactory function is inferred from the sophistication present by term, including behavioral
discrimination, preference, and conditioning to olfactory stimuli. For example, 1-week-old infants will reliably
turn their heads away from a noxious smell (84). In response to a series of pleasant or aversive odors, infants
less than 12 hours old will exhibit different facial expressions that are discriminable by adults (79). Infants
younger than 1 week of age reliably prefer the odor of their mother's breast pad to the breast pad of another
mother (85). Neonates who were given a period of familiarization to a novel odor subsequently demonstrated a
preference for that odor, whereas infants exposed to it for the first time did not (77). Finally, classical
conditioning to a novel olfactory stimulus has been demonstrated empirically within the first 48 hours after term
birth (86). Given ten 30-second pairings of citrus odor with stroking, neonates the following day showed
increased activity and head turning in the presence of the citrus but not to a novel odor. By 28 to 32 weeks
gestation, the majority of preterm infants show reliable behavioral response to olfactory input (82).
Disorders
Feeding disorders are reported commonly among preterm infants, particularly those with chronic lung disease.
The infant may even respond aversively to the introduction of food in the mouth. The cause generally is
attributed to frequent stressful procedures around the mouth as well as poor coordination of suck and swallow.
However, marked alteration of the orogustatory environment occurs by nature in a preterm birth. Feeding
disorders also are common among neonates who have sustained brain damage. No studies have attempted to
identify whether deficits in taste or smell are present in infants with feeding disorders. Certainly in adults, loss of
the sense of smell radically affects eating habits.
The amniotic fluid is a complex solution of suspended particulate and dissolved odorants that changes in
chemical composition during maturation of the fetal chemosensory system (29,82,87). Even as early as 18
weeks gestation, more than 120 compounds have been identified in single amniotic fluid samples. The mother
contributes to the chemical variation through hormones and even the types of food consumed. The fetus
contributes to the chemical status through urination, oral mucosa, and lung secretions. More directly, fetal
respiratory movements, sucking, and swallowing cause pulsatile displacement of the amniotic fluid in contact
with the chemoreceptors to probably affect
P.118
adaptation by the receptor cells (82). A link has been proposed between the intrauterine orogustatory
experience and the selective behavioral preference of the newborn to breast milk.
Experience In The Neonatal Intensive Care Unit

The environment in the NICU has not been described previously in terms of gustatory or olfactory content, but it
clearly is not well adapted here. The necessary restriction of oral feeding in sick or extremely premature
neonates is a marked alteration of the fetal orogustatory environment. The chemical composition of initial
feedings (breast milk or formula) is different from amniotic fluid or colostrum. Common changes in formula
composition (brand), concentration (dilute or hypercaloric), and temperature (too cool) introduce unwanted
variability. The addition of noxious oral medications and electrolyte supplements to breast milk or formula
typically activate a gag response in the newborn. Negative experiences temporally associated with feeding can
lead to aversive conditioning. Finally, in contrast to the healthy full-term neonate, preterm infants in the NICU
lack a stable olfactory source that would be provided by sustained body contact of a consistent caretaker.

Stimulation of both smell and taste has significant implications for enhancing care in the NICU. A period of
familiarization with the odor of breast milk or formula, before the oral introduction of food, can facilitate
acquisition of oral feeding skills (77). For a medically fragile infant, the mother's breast pad can be placed
nearby. From another standpoint, familiarization to the odor of medications before ingestion could make the
medication less aversive by the process of habituation. Obviously, like breast milk from the source, bottlefeeding ideally should be at body temperature rather than the typical room temperature.
Small tastes of formula or breast milk before the introduction of the nipple may foster behavioral organization
and facilitate the onset of feeding (87). Most babies who are restricted from oral feeds can safely tolerate a
small drop of breast milk or formula on the lips or tongue tip. The limited research available suggests that gut
priming in the extremely preterm or sick full-term infant probably should not bypass the mouth entirely. Based
on the animal model, surfactant and colostrum may have unsuspected roles in initiation of human feeding (81).
Another important implication of orogustatory stimulation is the effect of oral feeding on activation of the
endogenous opioid system, which raises the threshold to noxious tactile stimuli in the fetal rat model and human
newborn infants (88), although it is no longer elicited in the human infant after 6 weeks of age (89). Sucrose
solution has been applied to the tip of the tongue to decrease the pain response to a heelstick procedure or even
to circumcision in healthy term infants (88,90). However, repeated use of this pathway to modulate pain in the
neonatal period is currently unwarranted since it may have negative repercussions given the link to feeding
behavior.
Auditory System
The auditory system is composed of both peripheral and central components (91,92,93). Sound waves are
conducted through the auditory canal and physically displace the tympanic membrane. Movement of the
membrane is amplified by the ossicles in the middle ear and transmitted to the oval window. This action
displaces fluid in the cochlea. A mechanical disturbance differentially displaces hair cells at a specific place on
the basilar membrane of the cochlea, as a function of both frequency and intensity of the sound. The hair cells
are organized tonotopically, so that those that respond to high-frequency sounds are near the oval window and
those that respond to low-frequency sounds are at the apex of the cochlea. The complex neural impulse thus
generated proceeds to the auditory cortex via the cochlear nucleus, superior olivary nucleus, inferior colliculus,
and medial geniculate body. The primary cortical reception area is the Heschl gyrus in the temporal region.
Approximately 60% of the nerve fibers from each ear transmit to the contralateral hemisphere. The tonotopic
organization is repeated in the cortical structures. The initial development of the central component of the
auditory system is independent of peripheral maturation; however, once the auditory pathway is complete, the
absence of auditory stimulation would cause cortical neuronal degeneration (93).
Development
Development of the auditory system begins around 3 to 6 weeks of gestation (93,94). By 25 weeks gestation all
the major structures of the ear are essentially in place, although the adult dimensions of the external auditory
canal, tympanic membrane, and middle ear cavity will not be attained until 1 year after birth. The ossicles have
evolved from a thickening of mesenchymal tissue and are already of adult proportions, although residual
mesenchyme may diminish auditory thresholds. The cochlear nucleus has reached adult proportions and
differentiated sufficiently to be functional by this time, although microscopically the cochlea still is not mature
even at term. The hair cells are fully present and in a process of differentiation. The frequency-specific place on
the basilar membrane is shifting systematically during this period of development (93). The afferent pathway
from the cochlea to the auditory cortex is complete, and even myelination of the auditory pathway is present.
With regard to function, both cortical auditory-evoked responses and brainstem auditory-evoked responses can
be elicited by 25 to 28 weeks (95,96). Wave morphology is
P.119
different from the full-term infant's, and the latency is prolonged. A blink response to vibroacoustic stimulation
has been obtained in human fetuses of 24 to 25 weeks of gestational age. A more complex behavioral response
to sound occurs at least by 28 weeks, but readily fatigues. The maximum rate of electrophysiologic change
occurs in the cortical auditory-evoked response and brainstem auditory-evoked response between 28 and 34
weeks of gestation. Orienting behavior to soft sound can be elicited by this time.
Maturation of the fetal auditory system is marked by an increase in spectral sensitivity, in both lower and higher
frequencies, and a decrease in auditory threshold (91,92,93). The range of auditory sensitivity initially is fairly
restricted: from 500 to 1,000 Hz in the third trimester compared to around 500 to 4,000 Hz at term and an
adult range of 30 to 20,000 Hz. Changes in auditory threshold are related to maturation of both peripheral and
central components. Auditory thresholds in a preterm infant at 25 weeks of gestation have been obtained with a
65-dB stimulus compared to 25 dB at term.
Evidence for a functional auditory system in the fetus is strong. Specific anatomic sites are present in the cortex
that are responsible for processing complex sounds, such as language. A biological predisposition to respond to
the specific acoustic patterns of speech is present in full-term neonates. For example, they have lower
thresholds for sound within the most important range for speech perception (i.e., 500 to 3,000 Hz) (97). Within
this frequency range, they respond differently to speech and nonspeech stimuli. There are even hemispheric
differences in auditory-evoked potentials that support this language sensitivity (98). Finally, healthy full-term
neonates demonstrate a preference for sound they were exposed to in utero. Research shows that 2- to 4-dayold neonates prefer their mother's voice compared to another female voice and prefer a recording of a story
read by their mother prenatally to a recording of a story read by their mother that was not read to them
prenatally (99,100,101,102,103).
Deficits
Preterm infants are at increased risk for sensorineural hearing loss and developmental language disorders (20).
Language disorders may be receptive or expressive dysfunctions. Receptive language disorders often are
referred to as auditory processing deficits. These deficits primarily include phonemic-based disorders that
involve discrimination between speech sounds, such as ba versus pa, short-term memory deficits, and difficulty
in interpreting the meaning of words implied by grammatical structure. Expressive language problems may
include disorders of speech (as in articulation or fluency), word-finding difficulty, and deficient or disordered
sentence structure (grammar). Language disorders result from direct damage to central structures or can be
incidental to more general brain dysfunction. They occur in children with normal hearing thresholds and
otherwise normal intelligence. They occur more commonly among children who were born preterm (104).
Development of the auditory system during fetal life occurs within a uterine environment that contains rhythmic,
structured, and patterned sound emanating predominantly from the mother. Internal sounds include maternal
respirations, borborygmi, placental and heart rhythms, and the like. Maternal speech transmits both externally
and internally. Prosody (i.e., intonation, rhythm, stress) is probably the most salient aspect of speech available
to the fetus. The intensity of internally recorded sound within the amniotic fluid is approximately 70 to 85 dB,
with a predominance of low frequency (Fig. 9-1) (105). External sound also is transmitted to the fetus, but is
attenuated by the time it reaches the intrauterine cavity, more so at higher frequencies (i.e., 70 dB at 4000 Hz)
than lower frequencies (i.e., 20 dB at 50 Hz) (106). Given these considerations, the fetus probably is minimally
exposed to frequencies above 1,000 dB (107). The available frequencies in utero also parallel cochlear
development (100).
The auditory environment in the womb likely provides the most appropriate substrate for normal development of
the sensory system, but defining the acoustic properties of
P.120
the sound actually transmitted to the inner ear of the fetus is problematic. The fluid-filled womb would alter the
conductive property of the middle ear; however, data on this are unavailable. The best guess generally is that
fetal hearing is limited to bone conduction. Hearing thresholds are elevated in the prematurely born infant, but
no unequivocal data are available on the actual hearing thresholds of the fetus in utero.
Figure 9-1 Comparison of the frequency-specific auditory environment of the fetus in utero and the preterm
infant in the neonatal intensive care unit. SPL, sound pressure level. (From Walker D, Grimwade J, Wood C.
Intrauterine noise: a component of the fetal environment. Am J Obstet Gynecol 1970;109:91, with permission;
and Otho Boone, personal communication, December 1992, with permission.)
After a normal, full-term birth, the auditory environment is quiet by contrast. This may serve to increase the
salience of the human voice. Early speech directed to the neonate is subdued and generally contingent on the
infant's response.
Neonatal Intensive Care Unit Environment

The acoustic environment in the NICU differs from the intrauterine environment in peak intensity, spectral
characteristics, and pattern (Fig. 9-1). Ambient noise is generated by motors, fans, ventilator equipment,
personnel, telephones, alarms, hand-washing areas, trash lids, pagers, intercoms, and carts, to name a few.
The intensity of background noise is around 50 to 60 dB and, therefore, not louder than in utero, although
episodic bursts of higher intensity (i.e. less than 100 dB) occur. It would appear (Fig. 9-1) that speech input for
the neonate is selectively masked by the NICU auditory environment.
The auditory environment is different for babies nursed in open beds versus those nursed in incubators. To
some degree, the incubator may partially attenuate high-frequency room noise (Fig. 9-1), but produces its own
background noise level for the baby within (i.e., 50 to 70 dB). In addition, the infant in an incubator may be
subjected to the closing of porthole doors, an object being placed on the top surface, or rapping on the
incubator to stimulate an infant who is apneic or bradycardic, which may peak at more than 100 dB.
Effect of the Neonatal Intensive Care Unit Environment

Exposure to aberrant noise levels in the NICU may cause sensorineural damage, induce stress, and contribute to
language or auditory processing disorders in the preterm neonate. Schulte and colleagues (20) reported nearly
12% hearing loss in a followup of preterm infants. Damage to the outer hair cells of the cochlea of neonatal
guinea pigs has been reported after exposure to noise of similar intensities and frequencies found in the NICU
(108). These researchers found no correlation between hearing loss and duration of exposure to incubator noise,
but did report an association between hearing loss and a measure of severity of illness. Although not stated,
more severely ill neonates may be exposed to increased NICU noise compared to healthy preterm infants. In
other research, the combination of noise and ototoxic drugs commonly given to sick preterm infants (e.g.,
aminoglycosides, diuretics) was found to have a potentiating effect on hearing loss (109,110,111). The data
further suggest that the immature cochlea may be more susceptible to damage than the mature one. Increased
susceptibility is coincident with the final stages of anatomic development and differentiation of the cochlea
(111). Given these data, incubator manufacturers have lessened the noise level emitted by the incubator motor,
but limits on the environmental source of noise in the NICU itself are few.
In addition to possible sensory nerve damage, loud noise could have physiologic consequences in the newborn
preterm infant in the form of stress, leading to alterations in corticosteroid levels and autonomic changes.
Decreases in oxygen saturation, increases in intracranial pressure, and peripheral vasoconstriction are reported
in preterm infants after exposure to sudden noise (112). Finally, sleep is disrupted by intermittent noise in the
NICU.
Abnormal auditory environmental conditions could contribute to language problems in preterm children. Normal
auditory habituation patterns were impaired in chicks reared in an NICU-sound environment (113). Delayed
cortical auditory-evoked responses among healthy preterm infants have been reported, in addition to deficits in
brainstem response to linguistic stimuli (104,114). The NICU environment, like the fetal environment, has an
important role in the normal development of the auditory system.

Auditory intervention in the NICU includes efforts both to reduce ambient noise and to induce patterned auditory
input. Personnel in the NICU generally have made focused efforts to lower the level of noise in the environment.
Quiet times during each shift are prevailing, and radios are less constant. Evidence would suggest the value of
caring for the smallest and sickest preterm infants inside an incubator, as a shield against environmental noise.
The steady incubator noise may induce sleep in the preterm infant (115). At some point, however, the
continuous white noise in the incubator masks more socially relevant auditory stimuli. Prolonged housing in this
environment becomes disadvantageous at some point, but the timing is unknown.
Another approach to decreasing the level of ambient noise reaching the infant is by occluding the infant's ears.
Zahr and de Traversay (116), using a within-subjects design, found increased quiet sleep and fewer state
changes during the two observation periods. Caution here is warranted. Auditory input could be particularly
critical for hearing-impaired infants. Preterm infants' auditory thresholds generally are not tested until near the
time of discharge. Thus, if attenuation is practiced, the safer device would model that provided by the womb,
where higher frequencies are selectively attenuated. Without a more selective approach, speech sounds also
would be dampened. The strategy is, therefore, potentially harmful. On the other hand, short-term occlusion
might be appropriate
P.121
during a brief, acute phase of illness. Research is needed, with both short- and long-term consequences
carefully evaluated, including speech stimuli and hearing thresholds as outcome measures.
Auditory stimulation, as a single modality, has been studied in the preterm infant. Schmidt and colleagues (115)
reported that auditory stimulation, in the form of a heartbeat, lengthened the duration of the first quiet sleep
period. Quiet sleep is a more stabilized state that reflects central nervous system maturity. Some NICUs have
urged using sound as a protective window for the infantwhen music is played, the infant will not be disturbed.
This approach responds to the potential for conditioning in the preterm infant. Soothing sound is soft, simple,
repetitive, and harmonic, with a limited dynamic range. Nonetheless, speech and nonspeech stimuli differentially
stimulate the cerebral hemispheres (98). Availability of speech sounds may be more critical. In one of the
earliest studies, Katz (117) reported using auditory stimulation in the form of recorded maternal speech, played
to healthy preterm infants of 28 to 32 weeks of gestation, until each reached 36 weeks PCA. Compared to
controls, the intervention group showed better neuromotor development and improved auditory and visual
responses. This study highlights once more that the effects of an intervention may not be limited to the
stimulated sense. Even so, exposure to recorded speech lacks the essential reciprocal nature of normal
communication.
Visual System
The visual system is the most extensively studied sensory system; therefore, the mechanisms are better
understood. The eye is like a window to the brain, as it contains two thirds of the afferent nerve fibers in the
central nervous system. Light energy is transmitted through the cornea, pupil, lens, and optic media to the
retina. There it bypasses the retinal blood vessels, a layer of ganglion cells, and a layer of bipolar cells before it
finally reaches the outer segments of the photoreceptors (i.e., rods and cones). Light is absorbed by the
photoreceptors in a photochemical response that converts the radiant energy to an electrical impulse. The
amount of light energy necessary to stimulate a single photoreceptor cell is extremely smallone quantum
(118). In the absence of a light stimulus, retinal firing still occurs in the form of a tonic discharge. Some
processing occurs even at the level of the retina (118). From the photoreceptors, the impulse travels to the
ganglion cells, the optic nerve, and through the lateral geniculate nucleus to the occipital cortex. Fibers from the
medial portion of each retina decussate, whereas those from the lateral half do not. Thus, information from
either the left or right visual field will fall on the contralateral portion of each retina and be transmitted to the
same hemisphere of the brain. Representation in the cortex is topographic, but upside down and reversed.
Development
The eye is an outgrowth of the brain from the early embryonic stage. By 24 weeks of gestation, gross anatomic
structures are in place and the visual pathway is complete. As shown in Table 9-1, the visual system is
undergoing extensive maturation and differentiation between 24 and 40 weeks of gestation. Corresponding
functional visual responses have been elicited in the preterm infant (10,119,120,121,122,123,124,125).
As early as 24 to 28 weeks of gestation, a visual-evoked response to bright light can be obtained, but it consists
of a
P.122
long-latency negative wave that readily fatigues. A behavioral response to bright light consists of lid tightening,
but the response also fatigues quickly. The refractive error is approximately -5 diopters (D). The optic media is
cloudy.
TABLE 9-1 MATURATION OF THE FETAL EYE IN THE THIRD TRIMESTER
Fetal Eye Components 2628 Weeks of Gestation
Eyelid
Pupil
Lens
Media
Retina
Visual cortex
Fused early in
development, now
reopens.
Tunica vasculosa lentis.
begins to atrophy.
No reflex present.
Second of four-layer
nucleus forming.
Cloudy.
Hyaloid system begins. to
regress.
Rod differentiation begins.
Vascularization just
beginning.
Rapid dendritic growth
and differentiation.
3032 Weeks of
Gestation
3436 Weeks of Gestation
Less translucent.
Fully atrophied by end

of period.
Sluggish reflex.
Second layer complete,
third begins.
Clears.
Hyaloid almost
disappeared.
Complete except for
fovea, cone
differentiation begins.
Nasal portion nearly
complete.
Marked development of
dendritic spines and
synapses.
Few remnants.
Complete reflex.
Some remnants may still

be present.
Cone number in fovea
increases.
Temporal region nearly
fully vascularized.
Morphologically now
similar to full term.
Important functional changes occur around 32 weeks of gestation. The morphology of the visual-evoked
response becomes more complex with the addition of a positive wave, and the latency decreases. The pupillary
reflex is more efficient. A bright light will cause immediate lid closure, and the response sustains. The optic
media has often cleared. The eyes may open spontaneously, and the infant may even briefly fixate. This has
been described as the beginning of attention (119). Attention as such may be best elicited with a large, highcontrast form held closer to the eyes than would be necessary at term, but under similar conditions of low
illumination (i.e., 5 foot-candles [ft-c]).
By 36 weeks, the visual-evoked response resembles that of a full-term infant, but the latency is still longer for
the preterm infant and remains so. Spontaneous eye opening, even in utero, has been observed on ultrasound.
Although alertness still is less sustained than at term, the preterm infant now shows a spontaneous orientation
toward a soft light and can track an object horizontally and vertically. Additionally, the infant prefers a patterned
to a nonpatterned surface, in a manner similar to a full-term infant. The refractive error is near zero.
Relative to the other sensory systems, the visual system is the least mature by term birth, with considerable
development continuing over the next 6 months (126). Having less dense optic media and less macular
pigmentation than an adult, the eye of the newborn infant transmits more short-wavelength light by a factor of
four (78). Newborns are photophobic; thus, visual attention is facilitated under low illumination (i.e.,
approximately 5 ft-c). Acuity estimates are in the range of 20/200 Snellen equivalents. The refractive error is
normally slightly hyperopic (i.e., +1 D).
The newborn can attend to form, object, and face. Specifically, he or she can fixate a high-contrast form (i.e., a
1/16-inch wide line at a distance of 1 foot) and can show preference for patterns along dimensions of brightness
and complexity. She or he will track a bright object horizontally across midline and vertically. Attention to the
human face by a neonate can be explained as a predisposition to respond to contrast (e.g., eyes, open mouth)
or to edge (e.g., hairline), to slow movement (e.g., nodding), and to contingent stimulation (e.g., adult's voice).
In any event, this behavior is powerfully adaptive.
Deficits
It generally is agreed that the visual system of the preterm infant is particularly susceptible to insult. The most
well-known visual problem is retinopathy of prematurity (ROP), which is a proliferative vascular disease of
multifactorial origin. ROP has been linked to oxygen toxicity, but it occurs in preterm infants with cyanotic heart
disease who have never been hyperoxic. ROP is most strongly associated with degree of immaturity of the retina
(127,128,129,130). Visual disordersthicker lenses, poorer visual acuity, higher incidence of astigmatism, high
myopia, strabismus, anisometropia, and color deficits (blue-yellow)other than ROP are more common among
children born prematurely (131,132,133). For example, among a sample of 5-year-old, low-birth-weight
children, 35% lacked stereopsis and 25% had less than 20/20 corrected acuity in both eyes (132). Risk for
visual disorders is inversely related to gestational age.
In addition to these visual problems, the preterm infant also has difficulty processing visual information at a
more cognitive level. Performance on tests of visual attention, visual pattern discrimination, visual recognition
memory, and visuomotor integration repeatedly indicates particular vulnerability for the preterm infant
(9,12,13,134,135).
Intrauterine Environment
The womb generally is dark, but under certain conditions light can transmit to the fetus. A behavioral response
by a fetus to light has been described (136). Transmission through all the tissue is limited to small amounts of
red, or long-wavelength, light. Probably only 2% of incident light reaches the uterus (D. Sliney, personal
communication, June 1992). In later pregnancy, the head of the human fetus is in the vertex position, the neck
is flexed, and the face is posterior, thereby diminishing exposure. It is unlikely that light exposure is a necessary
condition for the fetus, or that periodic exposure to low levels of long-wavelength light is harmful. Aspects of the
light-dark cycle that reach the fetus probably are mediated more by maternal sources such as rest-activity
cycles and hormones than by light directly.
After birth, ambient light increases markedly, although typically the room is kept dim and cycled with dark to
some extent. A prolonged wake period linked to catecholamine release occurs during this transition to
extrauterine life. In dim light, the newborn is more likely to open his or her eyes.
Neonatal Intensive Care Unit Environment

Modern intensive care nurseries are brightly lit environments with ambient light in excess of standard office
lighting for adults (Fig. 9-2). The general range reported has been 30 to 150 ft-c, with peaks of more than 1,500
ft-c from sunlight (137,138). The intensity of ambient illumination for any individual infant is determined by the
location of the crib in a room; the number of overhead light units; the size, location, and compass direction of
windows; the season of the year; and even the prevailing weather conditions (i.e., sunny versus hazy). The
duration of exposure generally is 24 hours a day over the length of hospital stay, which is a function of degree
of immaturity and medical complications. Thus, light exposure is greater for those most vulnerable to visual
problems.
In addition to ambient light, preterm infants routinely are exposed to supplementary sources, such as a bili
light, heat lamp, and indirect ophthalmoscope. The standard double-bank phototherapy unit produces 300 to
400 ft-c
P.123
of illumination. The mini bili light has an intense beam, estimated at more than 10,000 ft-c. Infants' eyes are
routinely patched under phototherapy; however, in some cases the eye pads are inadequate or may slip off. A
commonly used heat lamp produces an intensity of more than 300 ft-c at an infant's face. Exposure time varies,
but typically is longer for younger and sicker infants. The eyes of infants typically are not covered during
procedures.
Figure 9-2A: Levels of light exposure in the intensive care nursery. B: Light exposure and retinal damage in
animals. (From Glass P, Avery GB, Subramanian KN, et al. Effect of bright light in the hospital nursery on the
incidence of retinopathy of prematurity. N Engl J Med 1985;313:401, with permission.)
Finally, an indirect ophthalmoscope is used for the routine eye examinations to rule out ROP. Exposure for 2
minutes, which is the approximate time of retinal examination, at maximum power has been estimated as
equivalent to exposure at 2,000 ft-c for 3 hours (139). Extra precautions for protecting the infant's dilated eyes
from ambient or supplementary sources of light before and after the eye examination are not routine. For the
smallest infants, the addition of a heat lamp during the examination often is necessary to maintain the baby's
temperature.
Phototoxicity in Animals
Animals exposed to similar levels of light have sustained damage to their photoreceptors, pigment epithelium,
and choroid (139,140). Phototoxicity is a consequence of a photochemical effect, but may be exacerbated by
heat. The effectiveness of light in producing retinal damage is proportional to its efficiency in bleaching
rhodopsin. A level of 100 ft-c bleaches rhodopsin to 80% in approximately 10 minutes. Continuous illumination
is more potent than cyclic, but intermittent exposure may be cumulative.
Factors that enhance photochemical damage to the animal retina strikingly parallel the perinatal course of the
preterm infant (137). Retinal damage in animals is facilitated by maintenance of the animal in constant dark
before light exposure, an increase in body temperature, conditions of hyperoxia, hypoxia, or ischemia, and
retinal disease. Finally, light and oxygen may have a synergistic adverse effect on ROP (141). In spite of all
these considerations, safety standards have not yet been determined.
Photobiological Effects of Light in the Neonatal Intensive Care Unit
Mounting data indicate that light has potent biological effects that are not considered routinely in standard NICU
care. An association between light and ROP was first suggested by Terry (142) in 1946. Potential mechanisms to
account for phototoxicity as one of the contributing factors in ROP have been proposed and are consistent with
the oxygen toxicity hypotheses: damage to endothelial cells, alteration of normal retinal metabolism, disruption
of the normal regenerative process of the retina, and generation of free radicals (137,141,143,144,145,146).
There is no evidence that light is a necessary condition for ROP or that maintaining a preterm infant in the dark
will completely prevent it. Seiberth and colleagues (147) found no difference in incidence of ROP for preterm
infants (birth weight less than 1,250 g) who had opaque eye patches day and night from birth to 35 weeks of
gestation compared to an unpatched control group. The daytime ambient light levels were already low (30 to 40
lux, or approximately 3 to 5 ft-c), and the control group had cycled lighting at night.
Three converging lines of evidence lend empirical support to an association between light and ROP. In a
prospective but nonrandomized study, preterm infants for whom the light levels were reduced for the duration
of their hospital stay had a lower incidence of ROP compared to a similar
P.124
group of preterm infants exposed to standard bright levels of nursery light (148). The same effect was found in
both NICUs studied. Similar findings were reported by Hommura and colleagues (149). Reynolds and colleagues
(150) reported no difference in incidence of ROP for preterm infants who wore light-reducing goggles compared
to infants who did not; however, the goggles were discontinued at 31 weeks of gestation. In a separate study,
increased ROP was identified in the region of the retina that is more exposed to lightspecifically, more in the
regions around 3 o'clock and 9 o'clock compared to the superior and inferior regions (151). In addition, preterm
infants with a higher degree of retinal pigment are less likely to develop ROP than infants with less retinal
pigment (152,153). Retinal pigment is protective against phototoxicity.
Photobiological effects are not limited to the retina. Elevated room light or phototherapy may cause degradation
of riboflavin and vitamin A, which are common components of total parenteral nutrition (154). Solutions
containing these vitamins may be covered at the source, but tubing often is not shielded. Light in the visible
spectrum penetrates the skin and thereby may alter more than the bilirubin concentration in the blood.
Riboflavin levels were reduced in vivo in infants undergoing phototherapy (155). Thrombocytopenia (i.e.,
platelets less than 150,000/mm3) was more than tripled among preterm infants exposed to phototherapy light
(156). In vitro experiments demonstrated inhibition in the normal constriction of immature lamb ductal rings
exposed to ambient laboratory light (157). Subsequently, Rosenfeld and colleagues (158) found a significant
reduction in the occurrence of patent ductus arteriosus among preterm infants whose chests were shielded from
exposure to phototherapy light.
Light and Behavior

Bright light in an infant's face is a source of stress (159). Even routine ambient light level affects sleep and wake
states in preterm infants. Lower ambient light is associated with significantly less active rapid eye movement
sleep and significantly more quiet sleep among preterm infants at 35 to 36 weeks PCA (160). Rapid eye
movement is a more destabilized state, with greater fluctuations of oxygen saturation and more frequent
occurrence of apnea and bradycardia. Lower ambient light also is associated with increased eye opening and
awake periods. Light level influences an infant's response to sound. Under bright light, an infant shows an
aversive response to a tone, whereas under a dim light the same tone elicits an orienting response (161).
Dimming the light has a quieting effect on caregivers, who are apt to behave as though someone were sleeping.

The optimal level of NICU lighting has not been determined; however, no study supports the safety of current
bright light levels. Therefore, it would be prudent to limit ambient nursery light to necessary levels and shield
the infant's eyes and chest from ambient as well as supplementary sources. Shading does not mean occluding.
Evidence does not suggest that infants' eyes should be patched beyond what is necessary for phototherapy.
Prolonged patching may be detrimental, both in terms of stimulus deprivation and possible effects on corneal
growth.
Opportunities for spontaneous eye opening under dim (i.e., 5 ft-c) or dark conditions should be provided. Visual
attention is more likely to occur under low light levels. Animal studies suggest that dim-dark cycling may be
beneficial for regeneration after retinal damage.
A day-night cycling regimen in the intermediate care nursery before hospital discharge affects behavior (162).
Infants in the cycled nursery showed improved sleep patterns both in the hospital and after discharge, spent
less time feeding, and gained more weight; however, light, noise, and handling all were reduced at night in the
experimental unit. That does not negate the effect, but the necessity of entrainment of preterm infants to lightdark cycles is not supported. Rest-activity cycles may be more potent. Biological rhythms are much more
complex (163,164).
The question then becomes whether to provide patterned visual stimulation. An infant's ability to respond to a
level of stimulation does not necessarily mean that he or she should be stimulated at that level. For example,
infants are more likely to respond to a louder sound, yet no one would recommend higher-intensity noise just
because the baby hears it better. Likewise, babies attend more to high-contrast black-and-white stimuli than to
pastel, but that does not necessarily mean that the baby should be stimulated with the stronger visual pattern
either. Prolonged or obligatory visual attention is not a preferred behavior. Given that the visual system is the
least mature, the most parsimonious approach would be to provide stimulation of the other senses first. Then
the most appropriate visual stimulus to begin with is probably the human face, which bears no resemblance to
strong black-and-white patterns.
GENERAL PRINCIPLES
A central issue in developmental intervention is whether to conceptualize the preterm infant as an extrauterine
fetus and, therefore, attempt to reproduce the intrauterine environment, or whether by dint of being born, the
system now requires other forms of stimulation to foster the unique development of the preterm infant. To some
extent this is a nonissue; abrupt change to extrauterine life in altricial species most always has been modulated
by the mother. The aversive conditions found in the NICU would not be conducive to development for even the
healthiest full-term infant. Further research is not necessary to determine whether excess handling, noise, and
bright light ought to be reduced. The NICU environment is a potential source of stress and overt damage to the
preterm brain. Research is needed to establish safety limits, rather than to study whether or not aversive
conditions do harm.
P.125
Sensory deprivation also can affect development. Research still is necessary to determine the optimal type,
timing, duration, and level of stimuli; however, basic guiding principles for developmental intervention do exist:
Preterm infants are not a homogeneous group. Thus, determining the appropriate level of stimulation is
based on an understanding of developmental neurophysiology and evaluation of an individual infant's
medical status, neurologic maturation, physiologic stability, and social and physical needs.
Sensory development is not a simple unitary process. The hierarchical organization and integration of
function among the sensory systems provides the conceptual framework for developmental intervention.
Thus, intervention should begin with the most mature system, should support the normal maturational
process, and should not attempt to accelerate development.
The model of optimal stimulation for early development lies in the sources naturally available to the fetus
and infant (i.e., the mother). The optimal NICU model thus would begin with the intrauterine conditions,
then parallel and extend the transition period that occurs immediately after a normal term birth.
A general algorithm therefore seems possible:
Provide, for any infant, protective care, minimal handling, undisturbed rest, dim, and quiet. Provide for
stabilization of autonomic, state, and motor processes through positioning and containment. Model
supportive touch.
Consider the role of olfaction and early bonding. Include positive taste experiences.
If the infant is sufficiently mature (32 weeks gestation) and medically stable, introduce graded tactilevestibular and auditory input. Introduce familiar forms of stimulation at lower intensity first and only if no
major medical changes are occurring simultaneously. Other than a parent's face, wait for visual
stimulation. Avoid any attempt to accelerate development.
Monitor and modify the approach as the infant matures.
Decreasing aversive conditions, such as bright light, noise, and handling in the NICU, and enhancing comfort
through touch, holding, positioning, and containment, benefit the infant and communicate directly to parents.
ACKNOWLEDGMENTS
I thank Cara Coffman and Susan Lydick for substantial contributions to a previous version of this chapter, and R.
D. Walk, my mentor.
REFERENCES
1. Hack M, Breslau N, Weissman B, et al. Effect of very-low-birth-weight and subnormal head size on
cognitive abilities at school age. N Engl J Med 1991;325:231.
2. Hack M, Fanaroff AA. Outcomes of children of extremely low birthweight and gestational age in the 1990s.
Semin Neonatol 2000;5:89106.
3. McCormick MC, Gortmaker SL, Sobol AM. Very-low-birth-weight children: behavior problems and school
difficulty in a national sample. J Pediatr 1990;117:687.
4. Klein N, Hack M, Gallaher J, et al. Preschool performance of children with normal intelligence who were
very-low-birth-weight infants. Pediatrics 1985;75:531.
5. Ross G, Lipper EG, Auld PAM. Educational status and school-related abilities of very-low-birth-weight
premature children. Pediatrics 1991;88:1125.
6. Taylor HG, Klein N, Minich NM, et al. Middle-school-age outcomes in children with very low birthweight.
Child Dev 2000; 71(6):1495.
7. Volpe JJ. Cognitive deficits in premature infants. N Engl J Med 1991;325:276.
8. Sostek AM. Prematurity as well as IVH influence development outcome at five years. In: Friedman S,
Sigman M, eds. The psychological development of low birth weight children. New York: Academic Press,
1992:259.
9. Kopp C, Sigman M, Parmelee A, et al. Neurological organization and visual fixation in infants at 40 weeks
conceptional age. Dev Psychobiol 1975;8:165.
10. Parmelee AH, Sigman M. Development of visual behavior and neurological organization in pre-term and
full-term infants. In: Minnesota symposium on child psychology, vol. 10. Minnesota: University of Minnesota
Press, 1976:119.
11. Sostek AM, Quinn PO, Davitt MX. Behavior, development and neurologic status of premature and full term
infants with varying medical complications. In: Field TM, Sostek A, Goldberg S, et al, eds. Infants born at risk.
New York: Spectrum, 1979:281.
12. Caron A, Caron R. Processing of relational information as an index of infant risk. In: Friedman S, Sigman
M, eds. Preterm birth and psychological development. New York: Academic Press, 1981: 219.
13. Rose SA. Enhancing visual recognition memory in preterm infants. Dev Psychol 1980;16:85.
14. Garcia-Coll C. Behavioral responsivity in preterm infants. Clin Perinatol 1990;17:113.
15. Sigman M. Early development of preterm and fullterm infants: exploratory behavior in eight-month olds.
Child Dev 1976;47: 606.
16. Gunnar M, Hertsgaard L, Larson M, et al. Cortisol and behavioral responses to repeated stressors in the
human newborn. Dev Psychobiol 1991;24:487.
17. Gunnar MR. Reactivity of the hypothalmic-pituitary-adrenocortical system to stressors in normal infants
and children. Pediatrics 1992;90:491.
18. Gorski PA. Developmental intervention during neonatal hospitalization. Pediatr Clin North Am
1991;38:1469.
19. Weisel TN, Hubel DH. Single cell response in striate cortex of kittens deprived of vision in one eye. J
Neurophysiol 1963;26:1003.
20. Schulte FJ, Stennert E, Wulbrand H, et al. The ontogeny of sensory perception in preterm infants. Eur J
Pediatr 1977;126:211.
21. Dobbing J. Later development of the brain and its vulnerability. In: Davis JA, Dobbing J, eds. Scientific
foundations of paediatrics. London: Heinemann, 1974:565.
22. Field T. Supplemental stimulation of preterm infants. Early Hum Dev 1980;4:301.
23. Cornell EH, Gottfried AW. Intervention with premature human infants. Child Dev 1976;47:32.
24. Lawson KR, Daum C, Turkewitz G. Environmental characteristics of the neonatal intensive care unit. Child
Dev 1977;48:1633.
25. Gorski PA. Premature infant behavioral and physiological responses to caregiving interventions in the
intensive care nursery. In: Call JD, Galenson E, Tyson RL, eds. Frontiers of infant psychiatry. New York: Basic
Books, 1983:256.
26. Korones S. Iatrogenic problems in intensive care. In: Moore TD, ed. Report of the sixty-ninth Ross
conference on pediatric research. Columbus, OH: Ross Laboratories, 1976:94.
27. Korner AF. Preventive intervention with high-risk newborns: theoretical, conceptual, and methodological
perspectives. In: Osofsky JD, ed. Handbook of infant development, 2nd ed. New York: John Wiley and Sons,
1987:1006.
28. Avery GB, Glass P. The gentle nursery: developmental intervention in the NICU. J Perinatol 1989;9:204.
P.126
29. Bradley RM, Mistretta CM. Fetal sensory receptors. Physiol Rev 1975;55:352.
30. Gottlieb G. The psychobiological approach to developmental issues. In: Mussen PH, ed. Handbook of child
psychology, vol. II, 2nd ed. New York: John Wiley and Sons, 1983:1.
31. Turkewitz G, Kenny PA. The role of developmental limitations of sensory input on sensory/perceptual
organization. J Dev Behav Pediatr 1985;6:302.
32. Humphrey T. Correlation between appearance of human fetal reflexes and development of the nervous
system. Prog Brain Res 1964;4:93.
33. Hooker D. The prenatal origin of behavior. New York: Hafner, 1969.
34. Fitzgerald M, Shaw A, MacIntosh N. Postnatal development of the cutaneous flexor reflex: comparative
study of preterm infants and newborn rat pups. Dev Med Child Neurol 1988;30:520.
35. Harlow H, Harlow M. The effects of rearing conditions on behavior. Bull Menninger Clin 1962;26:213.
36. Thoman EB, Ingersoll EW, Acebo C. Premature infants seek rhythmic stimulation, and the experience
facilitates neurobehavioral development. J Dev Behav Pediatr 1991:12:11.
37. Klaus MH, Kennell JH. Maternal-infant bonding. St. Louis: CV Mosby, 1976.
38. Tribotti SJ. Effects of gentle touch on the premature infant. In: Gunzenhauser N, ed. Advances in touch:
new implications in human development. Skillman, NJ: Johnson & Johnson Consumer Products, 1990:80.
39. Eyler FD, Woods NS, Behnke M, et al. Changes over a decade: adult-infant interaction in the NICU, 1992
(unpublished manuscript).
40. Adam K, Oswald I. Sleep helps healing. BMJ 1984;289:1400.
41. Sassin JF, Parker DC, Mace JW, et al. Human growth hormone release: relation to slow-wave sleep and
sleep-waking cycles. Science 1969;165:513.
42. Long JG, Philip AGS, Lucey JF. Excessive handling as a cause of hypoxemia. Pediatrics 1980;65:203.
43. Murdoch DR, Darlow BA. Handling during neonatal intensive care. Arch Dis Child 1984;59:957.
44. Peabody JL, Lewis K. Consequences of newborn intensive care. In: Gottfried AW, Gaiter JL, eds. Infant
stress under intensive care: environmental neonatology. Baltimore: University Park Press, 1985:201.
45. Perlman JM, Volpe JJ. Suctioning in the preterm infant: effects on cerebral blood flow velocity, intracranial
pressure, and arterial blood pressure. Pediatrics 1983;72:329.
46. Speidel BD. Adverse effects of routine procedures on preterm infants. Lancet 1978;2:864.
47. Volpe JJ. Intraventricular hemorrhage and brain injury in the premature infant: diagnosis, prognosis, and
prevention. Clin Perinatol 1989;16:387.
48. Schanberg S, Field T. Maternal deprivation and supplemental stimulation. In: Field T, McCabe P,
Schneiderman N, eds. Stress and coping across development. Hillsdale, NJ: Erlbaum, 1988:3.
49. Kuhn CM, Schanberg SM, Field T, et al. Tactile-kinesthetic stimulation effects on sympathetic and
adrenocortical function in preterm infants. J Pediatr 1991;119:434.
50. Als H, Lawhon G, Brown E, et al. Individualized behavioral and environmental care for the very-low-birthweight preterm infant at high risk for bronchopulmonary dysplasia: neonatal intensive care unit and
developmental outcome. Pediatrics 1986;78:1123.
51. Jay S. The effects of gentle human touch on mechanically ventilated very short gestation infants. Ph.D.
Thesis, University of Pittsburgh, Pittsburgh, PA, 1982.
52. Field TM, Schanberg SM, Scafidi F, et al. Tactile/kinesthetic stimulation effects on preterm neonates.
53. Scafidi FA, Field TM, Schanberg SM, et al. Massage stimulates growth in preterm infants: a replication.
Infant Behav Dev 1990;13:167.
54. Harrison LL, Leeper JD, Yoon M. Effects of early parent touch on preterm infants' heart rates and arterial
oxygen saturation levels. J Adv Nurs 1990;15:877.
55. Hack M, Estabecek M, Robertson S. Development of sucking rhythm in preterm infants. Early Hum Dev
1985;11:133.
56. Birnholz J, Stephens J, Faria M. Fetal movement patterns: a possible means of defining neurologic
developmental milestones in utero. AJR 1978;130:537.
57. Bernbaum JC, Pereira GR, Watkins JB, et al. Nonnutritive sucking during gavage feeding enhances growth
and maturation in premature infants. Pediatrics 1983;71:41.
58. Field T, Ignatoff E, Stringer S, et al. Nonnutritive sucking during tube feedings: effects on preterm
neonates in an intensive care unit. Pediatrics 1982;70:381.
59. Woodson R, Hamilton C. Effects of nonnutritive sucking on heart rate in pre-term infants. Dev Psychobiol
1988;21(3):207213.
60. Dixon S, Syder J, Holve R, et al. Behavioral effects of circumcision with and without anesthesia. J Dev
Behav Pediatr 1984; 5:246.
61. Field T, Goldson E. Pacifying effects of nonnutritive sucking on term and preterm neonates during
heelstick procedures. Pediatrics 1984;74:1012.
62. Geldard FA. The human senses. New York: John Wiley and Sons, 1967.
63. Mason WA. Wanting and knowing: a biological perspective on maternal deprivation. In: Thoman EB, ed.
Origins of infant's social response. Hillsdale, NJ: Erlbaum, 1979:225.
64. Neal MV. Vestibular stimulation and developmental behavior of the small premature infant. Nurs Res Rep
1968;3:1.
65. Korner AF. The use of waterbeds in the care of preterm infants. J Perinatol 1986;6:142.
66. Cordero L, Clark DL, Schott L. Effects of vestibular stimulation on sleep states in premature infants. Am J
67. Kramer LI, Pierpont ME. Rocking waterbeds and auditory stimuli to enhance growth of preterm infants. J
Pediatr 1976;88:297.
68. Pelletier JM, Short MA, Nelson DL. Immediate effects of waterbed flotation on approach and avoidance
behaviors of premature infants. In: Ottenbacher KJ, Short-DeGraff MA, eds. Vestibular processing dysfunction
in children. Binghamton, NY: Haworth Press, 1985:81.
69. Gatts JD, Fernbach SA, Wallace HD, Singra TS. Reducing crying and irritability in neonates using a
continuous controlled learning environment. J Perinatol 1995;15(3):215221.
70. Yu VYH. Effect of body position on gastric emptying in the neonate. Arch Dis Child 1975;50:500.
71. Henderson-Smart DJ, Read DJ. Depression of intercostal and abdominal muscle activity and vulnerability
to asphyxia during active sleep in the newborn. In: Guilleminault C, Dement W, eds. Sleep apnea syndromes.
New York: Alan R. Liss, 1978:93.
72. Martin RJ, Herrell N, Rubin D, et al. Effect of supine and prone positions on arterial oxygen tension in the
preterm infant. Pediatrics 1979;63:528.
73. Wagaman MJ, Shutack JG, Moomijian AS, et al. The effects of different body positions on pulmonary
function in neonates recovering from respiratory disease. Pediatr Res 1978;12: 571(abstract).
74. Grenier A. Prvention des dformations prcoches de hanche chez les nouveau-ns cerveau ls:
maladie de Little sans ciseaux? Ann Pediatr (Paris) 1988;35:423.
75. Anderson GC. Current knowledge about skin-skin (kangaroo) care for preterm infants. J Perinatol
1991;11:216.
76. Mistretta CM, Bradley RM. Development of the sense of taste. In: Blass EM, ed. Handbook of behavioral
neurobiology. Vol. 8: Developmental psychobiology and developmental neurobiology. New York: Plenum
Press, 1986:205.
77. Porter RH, Balogh RD, Makin JW. Olfactory influences on mother-infant interaction. In: Rovee-Collier C,
Lipsitt LP, eds. Advances in infancy research. Camden, NJ: Ablex, 1988:39.
78. Werner JS, Lipsitt LP. The infancy of human sensory systems. In: Gollin ES, ed. Developmental plasticity:
behavioral and biological aspects of variations in development. New York: Academic Press, 1981:35.
79. Steiner JE. Human facial expressions in response to taste and smell stimulation. Adv Child Dev Behav
1979;13:257.
80. Tatzer E, Schubert MT, Timischl W, et al. Discrimination of taste and preference for sweet in premature
babies. Early Hum Dev 1985;12:23.
81. Smotherman WP, Robinson SR. Milk as the proximal mechanism for behavioral change in the newborn.
Acta Paediatr Suppl 1994;397:64.
82. Schaal B, Orgeur P, Rognon C. Odor Sensing in the human fetus: Anatomical, functional, and
chemoecological bases. In: Lecanuet J-P, Fifer WP, Krasnegor NA, et al., eds. Fetal development: A
psychobiological perspective. Hillsdale, NJ: Erlbaum, 1995:205.
83. Pedersen PE, Greer CA, Shepherd GM. Early development of olfactory function. In: Blass EM, ed.
Handbook of behavioral neurobiology. Vol. 8: Developmental psychobiology and developmental neurobiology.
New York: Plenum Press, 1986:163.
P.127
84. Rieser J, Yonas A, Wikner K. Radial localization of odors by human newborns. Child Dev 1976;47:856.
85. Macfarlane JA. Olfaction in the development of social preferences in the human neonate. In: Parent-infant
interaction: Ciba Foundation Symposium 33. Amsterdam: Elsevier, 1975:103.
86. Sullivan RM, Taborsky-Barba S, Mendoza R, et al. Olfactory classical conditioning in neonates. Pediatrics
1991;87:511.
87. Smotherman WP, Robinson SR. Dimensions of fetal investigation. In: Smotherman WP, Robinson SR, eds.
Behavior of the fetus. Caldwell, NJ: Telford, 1988:19.
88. Blass EM, Hoffmeyer LB. Sucrose as an analgesic for newborn infants. Pediatrics 1991;87:215.
89. Barr RG, Quek VS, Cousineau D, et al. Effects of intra-oral sucrose on crying, mouthing and hand-mouth
contact in newborn and six-week-old infants. Dev Med Child Neurol 1994;36:608.
90. Blass EM, Shah A. Pain-reducing properties of sucrose in human newborns. Chem Senses 1995;20:29.
91. Aslin RN, Pisoni DB, Jusczyk PW. Auditory development and speech perception in infancy. In: Mussen PH,
ed. Handbook of child psychology, vol. II, 2nd ed. New York: John Wiley and Sons, 1983:573.
92. Hecox K. Electrophysiological correlates of human auditory development. In: Cohen LB, Salapatek P, eds.
Infant perception: from sensation to cognition. Perception of space, speech, and sound, vol. II. New York:
Academic Press, 1975:151.
93. Rubel EW. Auditory system development. In: Gottlieb G, Krasnegor N, eds. Measurement of audition and
vision in the first year of postnatal life: a methodological overview. Camden, NJ: Ablex, 1985:53.
94. Parmelee HP, Sigman MD. Perinatal brain development and behavior. In: Mussen PH, ed. Handbook of
child psychology, vol. II, 2nd ed. New York: John Wiley and Sons, 1983:95.
95. Birnholz JC, Benacerraf BR. The development of human fetal hearing. Science 1983;222:516.
96. Querleu D, Renard X, Boutteville C, et al. Hearing by the human fetus? Semin Perinatol 1989;13:409.
97. Berg, KM, Smith M. Behavioral thresholds for tones during infancy. J Exp Child Psychol 1983;35:409.
98. Molfese D, Freeman R, Palermo D. Ontogeny of brain lateralization for speech and non-speech stimuli.
Brain Lang 1975;2:356.
99. Fifer W, Moon C. Psychobiology of newborn auditory preferences. Semin Perinatol 1989;13:430.
100. Fifer WP, Moon C. Auditory experience in the fetus. In: Smotherman WP, Robinson SR, eds. Behavior of
the fetus. Caldwell, NJ: Telford, 1988:175.
101. DeCasper AJ, Fifer WP. Of human bonding: newborns prefer their mothers' voices. Science
1980;208:1174.
102. DeCasper AJ, Spence MJ. Prenatal maternal speech influences on newborn's perception of speech
sounds. Infant Behav Dev 1986; 9:133.
103. Spence M, DeCasper A. Newborns prefer a familiar story over an unfamiliar one. Infant Behav Dev
1987;10:133.
104. Kurtzberg D, Stapells DR, Wallace IF. Event-related potential assessment of auditory system integrity:
implications for language development. In: Vietze PM, Vaughan HG, eds. Early identification of infants with
developmental disabilities. Philadelphia: Grune & Stratton, 1988:160.
105. Gerherdt K. Characteristics of the fetal sheep sound environment. Semin Perinatol 1989;13:362.
106. Armitage SE, Baldwin BA, Vince MA. The fetal sound of sheep. Science 1980;208:1174.
107. Walker D, Grimwade J, Wood C. Intrauterine noise: a component of the fetal environment. Am J Obstet
Gynecol 1970;109:91.
108. Douek E, Dodson HC, Bannister LH, et al. Effects of incubator noise on the cochlea of the newborn.
Lancet 1976;2:1110.
109. Falk SA. Combined effects of noise and ototoxic drug. Environ Health Perspect 1972;2:5.
110. Walton JP, Hendricks-Munoz K. Profile and stability of sensorineural hearing loss in persistent pulmonary
hypertension of the newborn. J Speech Hear Res 1991;34:1362.
111. Carlier E, Pujol R. Supra-normal sensitivity to ototoxic antibiotic of the developing rat cochlea. Arch
Otorhinolaryngol 1980;226: 129.
112. Long JG, Lucey JF, Philip AGS. Noise and hypoxemia in the intensive care nursery. Pediatrics
1980;65:143.
113. Philbin MK, Ballweg DD, Gray L. The effect of an intensive care unit sound environment on the
development of habituation in healthy avian neonates. Dev Psychobiol 1994;27:11.
114. Salamy A, Mendelson T, Tooley WH, et al. Differential development of brainstem potentials in healthy
and high-risk infants. Science 1980;210:553.
115. Schmidt K, Rose SA, Bridger WH. Effect of heartbeat sound on the cardiac and behavioral responsiveness
to tactual stimulation in sleeping preterm infants. Dev Psychol 1980;16:175.
116. Zahr LK, de Traversay J. Premature infant responses to noise reduction by earmuffs: effects on
behavioral and physiologic measures. J Perinatol 1995;15:448.
117. Katz V. Auditory stimulation and developmental behavior of the premature infant. Nurs Res 1971;20:196.
118. Gregory RL. Eye and brain: the psychology of seeing, 4th ed. Princeton, NJ: Princeton University Press,
1990.
119. Hack M, Mostow A, Miranda S. Development of attention in preterm infants. Pediatrics 1976;58:669.
120. Dreyfus-Brisac C. Neurophysiological studies in human premature and fullterm newborns. Biol Psychiatry
1975;10:485.
121. Mann I. Development of the human eye. New York: Grune & Stratton, 1964.
122. Purpura DP. Morphogenesis of visual cortex in the preterm infant. In: Brazier MAB, ed. Growth and
development of the brain: nutritional, genetic, and environmental factors. International Brain Research
Organization monograph series. New York: Raven Press, 1975:1.
123. Dubowitz LM, Dubowitz V, Morante A, et al. Visual function in the preterm and fullterm newborn infant.
Dev Med Child Neurol 1980;22:465.
124. Miranda SB. Visual abilities and pattern preferences of premature infants and full-term neonates. J Exp
Child Psychol 1970;10: 189.
125. Senecal J, Defawe G, Roussey M, et al. Le comportement visuel du premature. Arch Fr Pediatr
1979;36:454.
126. Abramov I, Gordon J, Hendrickson A, et al. Light and the developing visual system. In: Marshall J, ed.
Vision and visual dysfunction. Boca Raton, FL: CRC Press, 1991.
127. James L, Lanman J. History of oxygen therapy and retrolental fibroplasia. Pediatrics 1976;57:590.
128. Lucey J, Dangman B. A reexamination of the role of oxygen in retrolental fibroplasia. Pediatrics
1984;73:82.
129. Johns KJ, Johns JA, Feman SS, et al. Reinopathy of prematurity in infants with cyanotic congenital heart
disease. Am J Dis Child 1991;145:200.
130. Inder TE, Clemett, RS, Austin NC, et al. High iron status in very-low-birth-weight infants is associated
with an increased risk of retinopathy of prematurity. J Pediatr 1997;131:541.
131. Fledelius T. Prematurity and the eye. Acta Ophthalmol 1976; 128:3.
132. Hoyt C. Long-term visual effects of short-term binocular occlusion of at-risk neonates. Arch Ophthalmol
1980;98:1967.
133. Dobson V, Quinn GE, Abramov I, et al. Color vision measured with pseudoisochromatic plates at five-anda-half-years in eyes of children from the CRYO-ROP study. Invest Ophthalmol Vis Sci 1996;37:2467.
134. Sigman M, Parmelee A. Visual preferences of four month old premature and fullterm infants. Child Dev
1974;45:959.
135. Siegel L. The prediction of possible learning disabilities in preterm and fullterm children. In: Field T,
Sostek A, eds. Infants born at risk: physiological, perceptual, and cognitive processes. New York: Grune &
Stratton, 1983:295.
136. Brazelton TB, Field TM. Introduction. In: Gunzenhauser N, ed. Advances in touch: new implications in
human development. Skillman, NJ: Johnson & Johnson Consumer Products, 1990:xiii.
137. Glass P. Light and the developing retina. Doc Ophthalmol 1990;74:195.
138. Landry RJ, Scheidt PC, Hammond RW. Ambient light and phototherapy conditions of eight neonatal care
units: a summary report. Pediatrics 1985;75:434.
139. Lanum J. The damaging effects of light on the retina: empirical findings, theoretical and practical
implications. Surv Ophthalmol 1978;22:221.
140. Williams TP, Baker BN, eds. The effects of constant light on visual processes. New York: Plenum Press,
1980.
141. Ham WT, Mueller HA, Ruffolo JJ. Mechanisms underlying the production of photochemical lesions in the
mammalian retina. Curr Eye Res 1984;3:165.
142. Terry L. Retrolental fibroplasia. J Pediatr 1946;29:770.
143. Dorey CK, Delori FC, Akeo K. Growth of cultured RPE and endothelial cells is inhibited by blue light but
not green or red light. Curr Eye Res 1990;9:549.
P.128
144. Riley PA, Slater TF. Pathogenesis of retrolental fibroplasia. Lancet 1969;2:265.
145. Stefansson E, Wolbarsht ML, Landers MB. In vivo O2 consumption in rhesus monkeys in light and dark.
Exp Eye Res 1983;37:251.
146. Zuckerman R, Weiter JJ. Oxygen transport in the bullfrog retina. Exp Eye Res 1980;30:117.
147. Seiberth V, Linderkamp O, Knorz MC, et al. A controlled clinical trial of light and retinopathy of
prematurity. Am J Ophthalmol 1994;118:492.
148. Glass P, Avery GB, Subramanian KN, et al. Effect of bright light in the hospital nursery on the incidence
of retinopathy of prematurity. N Engl J Med 1985;313:401.
149. Hommura S, Usuki Y, Takei K, et al. Ophthalmic care of very low birthweight infants, report 4: clinical
studies of the influence of light on the incidence of ROP. Nippon Ganka Gakkai Zasshi 1988;92:456.
150. Reynolds JD, Hardy RJ, Kennedy KA, et al. Lack of efficacy of light reduction in preventing retinopathy of
prematurity. N Engl J Med 1998;338:1572.
151. Fielder AR, Robinson J, Shaw DE, et al. Light and retinopathy of prematurity: does retinal location offer a
clue? Pediatrics 1992; 89:648.
152. Monos T, Rosen SD, Karplus M, et al. Fundus pigmentation in retinopathy of prematurity. Pediatrics
1996;97:343.
153. Schaffer D, Palmer E, Plotsky D, et al, on behalf of the CRYO-ROP Cooperative Group. Prognostic factors
in the natural course of retinopathy of prematurity. Ophthalmology 1993; 100:230.
154. Bhatia J, Mims L, Roesel R. The effect of phototherapy on amino acid solutions containing multivitamins.
J Pediatr 1980;96:284.
155. Sisson T. Advances in phototherapy of neonatal hyperbilirubinemia. In: Helene C, Charlier M, MontenayGarestier T, et al, eds. Trends in photobiology. New York: Plenum Press, 1982:339.
156. Maurer H, Fratkin M, McWilliams N, et al. Effects of phototherapy on platelet counts in low-birthweight
infants and on platelet production and life span in rabbits. Pediatrics 1976;57:506.
157. Clyman RI, Rudolph AM. Patent ductus arteriosus: a new light on an old problem. Pediatr Res
1978;12:92.
158. Rosenfeld W, Sadhev S, Brunot V, et al. Phototherapy effect on the incidence of patent ductus arteriosus
in premature infants: prevention with chest shielding. Pediatrics 1986;78:10.
159. Shogan MG, Schumann LL. The effect of environmental lighting on the oxygen saturation of preterm
infants in the NICU. Neonat Netw 1993;12:7.
160. Glass P, Sostek A. Sleep organization in preterm infants: the effect of nursery illumination. Presented at
the International Conference of Infancy Studies (poster session), New York, April 21, 1984.
161. Haith MM. Rules that babies look by. Hillsdale, NJ: Erlbaum, 1980.
162. Mann NP, Haddow R, Stokes L, et al. Effect of night and day on preterm infants in a newborn nursery:
randomised trial. BMJ 1986;293:1265.
163. Glotzbach SF, Rowlett EA, Edgar DM, et al. Light variability in the modern neonatal nursery:
chronobiologic issues. Med Hypotheses 1993;41(3):217-224.
164. Mirmiran M, Kok JH. Circadian rhythms in early human development. Early Hum Dev 1991;26:121.
Chapter 10
Prenatal Diagnosis in the Molecular Age
Indications, Procedures, and Laboratory
Techniques
Arie Drugan
Nelson B. Isada
Mark I. Evans
The modern era of molecular and biochemical genetics commenced with the
observations of Sir Archibald Garrod at the beginning of the twentieth century. He
proposed that four diseasesnamely, alkaptonuria, albinism, cystinuria, and pentosuria
resulted from inherited disorders of chemical metabolism. He also suggested that
these disorders, which he called inborn errors of metabolism, represented only a small
fraction of every human's chemical individuality that had gone awry (1).
Advances in biochemistry have confirmed Garrod's concepts by characterizing the
structural protein abnormality or enzymatic defect of many disorders. Other advances
in molecular genetics have allowed precise identification of the defect in the
deoxyribonucleic acid (DNA) message, sometimes before the protein defect itself is
known (2). This knowledge has direct and immediate applications in the field of
prenatal diagnosis (3). This chapter discusses gene organization; mutations and
polymorphism analysis; molecular diagnostic techniques; DNA cloning; an approach to
disorders diagnosable by molecular genetics; biochemical disorders not amenable to
DNA technology or better studied by protein chemistry techniques; and carrier
screening.
GENE ORGANIZATION
The Watson-Crick double-helix model of DNA organization is well known (4). DNA
conveys information encoded by a series of four nucleotidesadenine (A), thymine (T),
cytosine (C), and guanine (G)that are connected sequentially on two strands. The two
strands complement each other, with nucleotide base pairs (bp) being formed by
hydrogen bonding between adenine-thymine and guanine-cytosine. Eukaryotic DNA is
located in the nucleus and organized into structures called chromosomes. During
interphase, chromosomes are not visible by light microscopy. They can be observed
only when the genetic content has doubled and the chromosomes condense before
mitosis. Chromosomal material is organized into euchromatin and heterochromatin.
Euchromatin is vigorously transcribed into ribonucleic acid (RNA). Heterochromatin is
relatively inactive. An example of heterochromatin is the inactivated X chromosome.
An unexpected discovery made in the 1970s was that some regions of the eukaryotic
chromosome do not code for any known protein (5). Specifically, these noncoding
regions (introns), were noted to be interspersed within coding regions (exons) (6,7).
Exons carry information to direct the assembly of amino acids into a protein, whereas
introns do not. Messenger RNA (mRNA) acts as an intermediate molecule to convey
information encoded in the DNA by a process called translation. Posttranslational
modification of mRNA takes place such that introns are removed and exons are joined
together before amino acid sequences are formed. After additional biochemical
modifications, the mRNA passes out of the nucleus into the cytoplasm, where proteinsynthesizing organelles are located.
Approximately 60% of the human genome is comprised of regions of unique nucleotide
sequences that presumably
P.131
code for proteins (8). It is estimated that there are 50,000 to 100,000 expressed genes
and proteins active in humans; however, expressed genes comprise less than 10% of
total genomic DNA. A significant portion of the human genome, perhaps approximately
40%, contains repetitive DNA sequences (9). Various terms are used for the different
classes of repetitive DNA sequences found in humans. Highly repetitive sequences are
found in the chromosome region adjacent to the centromere. These are simple
sequences that are repeated thousands of times, are present in more than 104 copies,
and comprise approximately 20% of the genome.
Other repetitive, simple sequences are several hundred base pairs long, and are
separated easily by centrifuging slightly fragmented DNA through a cesium chloride
density gradient. DNA separated by this process is called satellite DNA, because the
centrifuged DNA forms a main band and several satellite bands above and below the
main band. In humans, four satellite bands comprise approximately 6% of the total
DNA, each band representing tandem repeat sequences. Blocks of satellite DNA are
readily localized by in situ hybridization to regions around the centromeres of
metaphase chromosomes. Satellite DNA should not be confused with satellites, a
cytogenetic term referring to the segment of an acrocentric chromosome distal to short
arm and separated by a constriction.
Other classes of repetitive DNA include moderately repetitive DNA, which is gene size in
length, repeated from 10 to 1,000 times, and comprises approximately 20% of genome
(i.e., one-half of the 40% that is repetitive DNA); tandem repeat sequences, which are
stretches of DNA in which a short nucleotide sequence is repeated 20 to 100 times, the
exact number varying from person to person; alphoid DNA, which is a chromosomespecific, repeated, monomeric 170-bp unit located in centromeric regions; and Alu
sequences, which are highly repetitive 300-bp sequences that are not clustered around
centromeres, but are more evenly distributed throughout the genome and interspersed
within longer stretches of unique or moderately repetitive DNA. Most contain a single
cleavage site near the middle for the restriction enzyme Alu I, derived from the
bacterium Arthrobacter luteus (see Restriction Fragment Length Polymorphism Analysis
below). Almost 1 million Alu sequences are present in the human genome, accounting
for 3% to 6% of the total DNA. Each individual human has a unique amount of
repetitive DNA. This genetic fingerprint is used for paternity testing and forensic
analysis.
A discovery involving repeat sequences demonstrates an association between changes

in length of triplet repeats and human disease. A significant increase in length of the
normal sequence of three repeated nucleotides is associated with marked tendency to
clinical morbidity. This was first described in the fragile X syndrome, a common form of
heritable male mental retardation, where a polymorphic sequence of -CGG- repeats on
the X chromosome is increased from a mean of 29 repeats to more than 200 repeats in
affected males. Triplet expansion of -GCT- is associated with congenital myotonic
dystrophy. Triplet expansion of -CAG- is found in spinal-bulbar muscular atrophy,
spinocerebellar ataxia, and Huntington disease (10,11,12).
Regulation of gene expression occurs at many levels. Substrate concentrations both
enhance and repress specific enzymes, especially in bacteria. In eukaryotes and
prokaryotes, a region separate from a given gene but still involved in regulation of
expression is called promoter; it binds RNA polymerase and initiates transcription.
Similar functional areas have been found that bind a large group of transcription
factors. In eukaryotes, such an area or box contains a common or consensus nucleotide
sequence, CAAT, and is called a CAAT box. A conserved area associated with RNA
polymerase positioning contains repeated sequences of the nucleotides thymine and
adenine and is called a TATA box.
Extranuclear DNA found in mitochondria (mtDNA) is inherited independently of nuclear
DNA, is 16.5 kb in length, contains no introns, and is organized as a double-stranded
circle. mtDNA is subject to a relatively high rate of mutation compared to nuclear DNA
and is inherited maternally, the paternal spermatic mitochondria being excluded from
the oocyte at the time of fertilization. A variety of neonatal and pediatric disorders,
predominantly neuromuscular, are associated with mtDNA deletions (13,14,15).
ELEMENTS OF MOLECULAR GENETICS

The deciphering of DNA structure and the genetic code has advanced understanding of
the molecular basis of genetic disease. Alterations in DNA sequence may affect a gene
structure or function, and may result in an altered gene product that may (or may not)
be expressed as genetic disease. An alteration in the sequence of nucleotides within a
gene is termed a mutation.
Many types of mutations can affect a gene. The DNA within a gene contains information
for the final sequence of a protein and signals for the correct expression and processing
of mRNA. If the actual coding region is altered, then the resultant protein may be
changed. These alterations can be in the form of deletions that may be many kilobases
in length or as small as a single base, inversions or translocations that produce no net
nucleotide changes but potential or actual protein changes, or single-base substitution.
Even a change at the junction of a coding and noncoding region can result in abnormal
mRNA formation. Defects in the promoter region may result in too little or too much
expression of mRNA, which will be reflected in abnormal protein synthesis. A deletion of
all or part of the gene will almost always result in the disruption of normal gene
expression.
Two examples of molecular pathology can be seen in Ashkenazi Jews (16). In the
severe infantile type of Tay-Sachs disease, a mutation at an exon-intron splice site in
the -subunit has been identified (i.e., G to C transversion); another mutation is a fourbase insertion in exon 11 of the
P.132
-subunit causing a frameshift mutation and marked reduction in mRNA (17,18).
Another example is the F508 mutation, the most common (approximately 70%)
mutation causing cystic fibrosis (CF) in people of northern European ancestry
(19,20,21). The remaining 30% of CF cases in people of northern European ancestry
are caused by a heterogeneous assortment of other mutations (i.e., W1282X).
Commercial testing is available and can detect more than 90% of mutations in the
northern European population (22). However, other ethnic groups carry their own
particular repertoire of mutations and detection rate may be as low as 50% (23).
A variety of approaches have been used to detect mutations. Optimally, determination
of DNA structure and sequence, followed by elucidation of gene structure and
organization in the normal allele, are completed before beginning a search for specific
defects. However, only 5% to 10% of clinically significant mutations are a result of
gross alterations in gene structure that are detectable by Southern blot analysis of
genomic DNA, leaving unknown the remaining 90% to 95%. The problem is
compounded by normal variation in the nucleotide sequence (polymorphism). Thus,
when variations from the normal sequence are found, additional analysis is required
before these changes can be construed as being a disease-causing mutation.
Because DNA is present in each cell nucleus, any nucleated cell theoretically is suitable
for DNA analysis, regardless of whether the gene in question is being transcribed and
expressed. Thus, leukocytes, amniocytes, and chorionic villi all are candidate cells for
DNA analysis, using Restriction Fragment Length Polymorphism (RFLP). Other methods
also used for DNA diagnosis include Southern blot, oligonucleotide probes, and
polymerase chain reaction (PCR). Northern blotting is used for RNA analysis.
RFLP analysis uses bacterial enzymes that recognize and cleave DNA at specific sites.
Presumably, these enzymes evolved as a defense mechanism against hostile, invading
DNA, as might occur with bacteriophages. Because these cleavage sites are quite
specific, that is, are restricted to specific palindromic sequences 4 to 10 nucleotides in
length, these enzymes are called restriction endonucleases. When these enzymes are
added to eukaryotic DNA, the resultant mixture contains a variety of DNA fragments of
different sizes, which can be separated by gel electrophoresis and transferred for
analysis by Southern blotting. Each enzyme cuts an individual's DNA according to the
positions of the cleavage sites, with every person having his or her own unique pattern
of cleaved DNA fragments. Thus, people are polymorphic for the resulting lengths of
DNA fragments. Many of these recognition-site polymorphisms are neutral and
represent normal inherited variability. These characteristics have given rise to the term
restriction fragment length polymorphisms, which refers to the polymorphic patterns
observed in specific nucleotide sequences that are cleaved by bacterial restriction
enzymes (Fig. 10-1). The resultant mixture of DNA fragments can be separated and
further characterized by gel electrophoresis, Southern blotting, and oligonucleotide
probes (24). DNA alterations that affect an RFLP site either by creating a new site for
endonuclease cleavage or eliminating a previously existing one can be detected on
Southern blot as a result of changes in the size of the DNA fragment associated with
this site (25). If, by chance, either a mutation or normal sequence corresponds to an
RFLP site, this situation can be exploited for allele identification by using linkage
analysis.
Figure 10-1 Restriction fragment length polymorphisms. Lane 1, A1, and A2 present;
lane 2, A1-A3 present; lane 3, A1-A4 present; A1-A5, hypothetical polymorphisms.
An initial use of such mapping involved the Huntington disease locus (26). Another
early application was for prenatal identification of the sickle-cell mutation in the chain
of hemoglobin (27,28). For example, the restriction enzyme Dde I, derived from the
bacterium Desulfovibrio desulfuricans aestuarii, recognizes the nucleotide sequence CTNAG- (where N indicates that any nucleotide may occupy that position) that occurs
within the hemoglobin A (-CTGAG-) and hemoglobin C (-CTAAG-) gene, but not within
hemoglobin S (-CTGTG-). In hemoglobin S, the nucleotide thymine is substituted for
adenine, which is not recognized by Dde I and thus is not cleaved by Dde I. This results
in a much larger RFLP fragment that can be recognized on Southern blot. Initially, this
technology used unamplified DNA and Southern blot transfer. Target gene sequences
can now be preamplified a million-fold by PCR (see next page) and then cut by
restriction endonucleases, which greatly facilitates target sequence recognition. This
approach is potentially useful in prenatal diagnosis, particularly when the quantity of
clinical material is limited.
When the precise nucleotide mutations of the abnormal gene causing the disease are
unknown, linkage analysis is a very powerful technique that can be used for diagnosis
by association of the diseased gene to a known gene or polymorphic site (29). This
implies that the closer together two genetic traits are on a chromosome, the more likely
they are to segregate together during meiosis and the less likely crossing-over occurs
between them. When genes are in such close proximity that crossing-over rarely
occurs, such as for
P.133
hemophilia A and color blindness, the two sites are said to be linked (30). Any
polymorphism that is linked with the trait of interest is termed informative.
Unfortunately, family studies are not always informative, because linked molecular or
clinical polymorphisms are not always present.
The likelihood that two given traits are linked can be described mathematically by the
logarithm of the odds (LOD) score, developed in 1955 by Morton (31). It can be derived
from recombination observed between clinical or biochemical traits from pedigree
analysis, or from molecular polymorphisms (32). The probability of recombination
during meiosis between two loci is quantified by the recombination fraction (i.e., theta
or q), the maximum being 0.5. The LOD score is derived from various values of q.
Viewed simplistically, the higher the LOD score, the higher the likelihood of linkage.
Because this number is used on a logarithmic scale, each integer increase reflects a
tenfold increase in likelihood of linkage. Thus, a LOD score of 4 suggests that there is
linkage between two polymorphisms, the odds of random association being 10,000:1. A
LOD score of zero suggests there is no linkage and that the two traits are on different
chromosomes or are far apart on the same chromosome. Tight linkage indicates little or
no recombination and suggests an actual physical proximity of two polymorphisms,
measured in physical map distances, with the common unit of genetic distance reported
as centimorgans (cM). Linkage disequilibrium describes closely linked genes that occur
more frequently than would be expected from random distribution, suggesting
nonrandom mating or some survival advantage from natural selection. Examples of
linkage disequilibrium include the carrier states for sickle cell disease and thalassemia,
where those affected have increased resistance to certain types of malarial infections.
Other methods used for DNA analysis include Southern blot, oligonucleotide probes,
and PCR. Northern blot is used for RNA analysis. Southern blot is a standard method for
DNA analysis in both the clinical and basic science settings. In the Southern blot
technique, named after Edwin Southern, double-stranded DNA is digested by a
restriction endonuclease chosen because of its ability to detect a DNA polymorphism
(33). After endonuclease digestion, the resulting DNA fragments are separated using
gel electrophoresis. The DNA in the gel is denatured to generate single-stranded DNA
molecules. DNA fragments are transferred from the gel to nylon filter paper (blotting),
and specific filter-bound DNA fragments then can be detected by hybridization. A
radiolabeled DNA or RNA probe is used that has sequence homology to the DNA
fragment of interest, usually 200 to 2,000 bases long. Subsequent autoradiography
produces a radiographic film with banding patterns that indicate the hybridization
locations on the filter that reflect the fragment sizes of the DNA sequences homologous
to that particular probe (Fig. 10-2).
Allele-specific oligonucleotide hybridization has proven to be a valuable technique that
measures the specific binding of short (18 to 20), labeled oligonucleotide probes, that
match exactly either the wild-type) normal (or the mutant DNA sequence, under
stringent washing conditions. Only the probes that exactly complement the immobilized
DNA will remain bound and thus generate a signal seen on autoradiography. Conner
(34) originally described this technique for the detection of sickle cell -globin allele.
This technique greatly facilitates the evaluation of genetic disorders in which the gene
has to be screened for numerous mutations such as thalassemia, Tay-Sachs, Gaucher,
or CF.
Figure 10-2 Southern blot for a hypothetical autosomal recessive disorder. Lane 1,
unaffected; lane 2, affected; lane 3, carrier.
PCR has revolutionized the field of molecular genetics (35,36). It's discoverer, Kary
Mullis, won the Nobel Prize in Chemistry in 1993. This procedure allows in vitro
amplification of minute amounts of DNA to generate sufficient quantities of signal to
make detection by more traditional methods possible. PCR makes use of Taq I, a
relatively heat-stable bacterial enzyme derived from Thermus aquaticus, a
thermoacidophilic bacterium. If the target nucleotide sequence is known, a specific set
of oligonucleotides, called primers, can be synthesized to encompass the target
sequence. The target DNA, oligonucleotide primers, Taq I polymerase, and free
nucleotides are placed in solution. This reaction mixture is further heated to allow
already denatured DNA to anneal with the oligonucleotides, between which the
polymerase synthesizes complementary strands (Fig. 10-3). Repeated cycles of heating
and cooling result in cyclic primer sequence synthesis, leading to annealing and
amplification of the target sequence, because
P.134
each set of DNA strands gives rise to two additional sets of sequence templates in each
cycle of the reaction. This process can be automated to allow 20 to 30 cycles, which can
produce more than a million-fold duplication of the target sequence within hours.
Modifications of this process can be performed to allow:
Analysis of RNA.
Analysis of multiple DNA areas (multiplex PCR). This technique allows detection of
more than 97% of deletions in Duchenne muscular dystrophy and all those of
patients at risk for Becker muscular dystrophy (37).
Selective amplification of one strand instead of both (i.e., asymmetric PCR).
Simultaneous use of one primer set within another to increase specificity (i.e.,
nested PCR).
Simultaneous use of two different primer sets, one of which selects for a normal
sequence and the other for a mutant sequence (i.e., competitive oligonucleotide
priming).
Simultaneous use of a known amount of a second, easily identified target DNA to

measure the amount of original DNA (i.e., semiquantitative PCR).
Figure 10-3 Polymerase chain reaction.
Many technical difficulties must be addressed to eliminate both false-positive and falsenegative PCR results. Problems with reagent or reactant contamination can lead to falsepositive results and require that appropriate control methods be performed
simultaneously to verify positive PCR results. Other problems, such as primer instead of
target amplification and nonspecific amplification, must be recognized and avoided.
Northern blotting is used for RNA analysis. This requires prompt specimen processing
and committed laboratory reagents and instruments because of ubiquitous
ribonucleases, present even on finger surfaces. The general principles of the technique
are similar to those for Southern blotting. Examination of the size and amount of a
mRNA transcript is a useful initial step in evaluating the expression of mutated genes in
cells or tissues. Fibroblasts and lymphocytes are good sources of mRNA. Hepatic or
muscle tissue is also useful, if available. Placental tissue can be used if maternal cell
contamination can be avoided. Of the cell's total RNA, only 1% to 2% is mRNA, which is
highly unstable at room temperature because of tissue RNA-ases. The remainder of the
RNA is mainly ribosomal RNA (rRNA) and transfer RNA (tRNA). Once isolated, the mRNA
is denatured, separated by agarose gel electrophoresis, transferred to a membrane
filter, and analyzed by hybridization of a specific fluorescent or radiolabeled probe.
For most diseases studied at the molecular level, 5% to 10% of patients have no
detectable mRNA for the gene product in question; 10% to 20% have reduced but
detectable amounts of normal mRNA; and approximately 5% have some alteration in
mRNA size. The approximately 50% remaining have normal amounts of normal-size
mRNA. Given this information, it is possible to deduce the general type of mutation at
the DNA level, such as large or total gene deletions, which are suggested by a total
absence of mRNA. Mutations in promoter regions are suggested by reduced amounts of
normal mRNA. Mutations at exon-intron junctions are suggested by changes in mRNA
sizenormal-size mRNA but abnormal function of protein hints at point mutations.
In some instances, the quantity of RNA is insufficient to be detected in the previously
mentioned methods. In these cases, reverse transcriptase PCR methodology permits
identification and isolation of small quantities of mRNA and thereby analyze genes in a
more fastidious manner. Grompe et al. (38) were unable to detect by Northern analysis
the mRNA in an ornithine transcarbamylase-deficient patient. Nonetheless, the mRNA
was isolated and successfully amplified (after synthesis of a single-stranded
complementary DNA [cDNA] by using the RNA-directed DNA polymerase, reverse
transcriptase), and the mutation responsible for the disease was identified. This
technique also allows isolation of the shorter cDNA fragments corresponding to the
coding region of the gene of interest. As in Menkes' (kinky hair) disease, a
neurodegenerative disorder associated with a disturbance of copper metabolism, exon
splicing is the characteristic result of the splice-junction mutations seen in this rather
large gene (39). By using reverse transcriptase PCR, one can discern which exons are
lacking by the size of fragment seen on agarose gel or by sequence analysis.
The basic steps in the study of a gene start with the isolation of a DNA molecule
complementary to its mRNA, called cDNA, that contains only exonic sequences. The
first
P.135
step in cDNA cloning is the isolation of mRNA from a particular cell or tissue that
contains a significant amount of the desired mRNA. A retroviral enzyme, reverse
transcriptase, is used to synthesize cDNA from the mRNA. Because the cell or tissue
mRNA contains transcripts of many genes, the resultant pool of cDNA will be
heterogeneous and must be sorted out. The cDNA strands are inserted into a vector to
form a cDNA library. A second type of chromosome library is composed of fragments of
native genomic DNA and contains introns and other noncoding regions. Vectors include
viruses that can replicate within bacteria, such as bacteriophage lambda, or
autonomous, self-replicating, circular DNA molecules found in bacteria, called plasmids.
Another vector that combines properties of plasmids and bacteriophage lambda are
called cosmids. Cosmids are plasmid vectors into which larger fragments of DNA can be
cloned. The term cosmid is derived from the presence of internal cohesive end sites
(cos) that have been inserted into a plasmid. Cos are nucleotide sequences from
bacteriophage lambda between which DNA sequences are normally expressed as
capsule proteins. In cosmids, these sequences can be replaced with other nucleotide
sequences, which then can be expressed and concentrated in vitro.
An alternative approach to cDNA cloning uses mRNA itself instead of cDNA. Using mRNA
is advantageous because the mRNA pool, although heterogeneous, is not nearly so

complex as a corresponding pool of genomic DNA fragments that contains a mixture of
introns, nonexpressed genes, and DNA from noncoding regions. Methods of cDNA
cloning to detect the presence of a specific cDNA segment include screening a cDNA
library with synthetic oligonucleotide probes and screening a bacteriophage lambda
library with antibodies directed against the protein of interest. Using the first approach,
a set of radiolabeled oligonucleotide probes is synthesized with sequences
complementary to those predicted for the mRNA from the amino acid sequence found in
the protein. These oligonucleotides are used to probe a cDNA library that contains
bacterial colonies infected with bacteriophage lambda. With the second approach,
antibodies against the protein of interest are used to detect a corresponding protein
expressed by cDNA inserted into bacteriophage lambda.
cDNA for many genes have been cloned using these approaches. Nucleotide sequencing
is the first step in the analysis of a cDNA. With this information, the investigator can
identify potential sites of restriction endonuclease cleavage, can compare the deduced
amino acid sequence with that of the protein of interest, and can identify regions of
homology with other proteins, which may suggest evolutionary relationships and
previously unrecognized functions. In addition to providing nucleotide and amino acid
sequence information, these cDNA can be used as hybridization probes to characterize
and quantitate corresponding mRNA in cells and tissues from patients with a variety of
diseases.
The process by which DNA analysis leads to an identifiable protein abnormality was
previously called reverse genetics; it is now called positional cloning (40,41).
Several hundred genes associated with human disease have been identified. In some
diseases, such as Cystic Fibrosis, the triple-nucleotide/single-amino-acid deletion in the
CFTR protein has been recognized. For other diseases, the gene and protein are better
characterized, such as for neurofibromatosis type 1 (42). For most human diseases,
however, the genes and the defect in protein synthesis have yet to be defined.
PRENATAL DIAGNOSIS
Whether by cytogenetic, biochemical, or molecular methods, prenatal diagnosis of an
affected fetus requires obtaining either fetal tissue (i.e., blood) or other tissue for
analysis that is representative of the fetus (e.g., amniocytes or placenta). Invasive
procedures for prenatal diagnosis of fetal disease are available throughout gestation,
from the first trimester onward. Even earlier, assisted reproduction technologies (ART)
enable diagnosis (or exclusion) of several disorders on the 4- to 8-cell embryo before
implantation.
Invasive procedures for diagnosis in pregnancy of fetal genetic disorders have been
available for almost four decades, since the introduction of techniques for culturing and
karyotyping of amniotic fluid fibroblasts in the mid-1960s (43). The first diagnosis of a
fetal chromosome anomaly by amniocentesis (44) was followed shortly by the diagnosis
of an enzyme deficiency in amniotic fluid cells (45). Thereafter, collaborative studies
established the safety and accuracy of midtrimester amniocentesis, so that this
technique became a routine part of prenatal care in high-risk patients and the gold
standard against which other procedures for prenatal diagnosis are compared (46,47).
Despite its proven efficacy, a major disadvantage of amniocentesis is the availability of
results late in the second trimester, generally 18 to 20 weeks of gestation. The
emotional and physical implications of termination of pregnancy this late in gestation
are obvious. Improvement in ultrasonography machinery and increasing expertise in
ultrasound-guided procedures enabled physicians in the late 1980s to attempt prenatal
diagnosis in the first trimester, introducing chorionic villus sampling (CVS) and early
amniocentesis. These technical developments are backed and reinforced by increasing
preference on the part of patients for first trimester prenatal diagnosis (48). Chorionic
villus sampling is usually performed between 11 and 13 weeks of gestation so that
results are available by the end of the first trimester. The accuracy and safety of CVS
are quite comparable to those of mid trimester amniocentesis (49,50) and the early
results allow patients privacy in reproductive decisions and an earlier and safer
termination of pregnancy if so opted. An alternative to CVS was offered by early
amniocentesis performed between 10 and 14 weeks of gestation, but this technique has
been largely abandoned (51).
Over the years, a change in the pattern of indications for prenatal diagnosis has been
observed. The most common
P.136
indication for genetic counseling and prenatal diagnosis is the need to evaluate the
karyotype of the fetus, this being indicated in more than 70% of cases of advanced
maternal age (defined as 35 years or older at birth). Other classical indications to
evaluate fetal karyotype include a previous, affected offspring and a balanced structural
rearrangement of parental chromosomes, the latter being clinically evident as recurrent
pregnancy loss (Table 10-1). In recent years, increased use of biochemical serum
screening and of ultrasonographic screening for fetal chromosome anomalies have
caused more young patients, previously considered to be at low risk for fetal
aneuploidy, to opt for invasive prenatal testing. The combination of double, triple, or
quadruple serum screening (-fetoprotein [AFP], human chorionic gonadotropin
[hCG], and unconjugated estriol [uE3], with or without inhibin A) and maternal age will
select for prenatal testing a sub-group of patients among whom 65% to 75% of
chromosomally abnormal conceptions will be contained. Using a risk cutoff for fetal
aneuploidy equal to that of age 35 years, some 5% of young pregnant patients will
have a positive screening test, and 1 in 50 amniocenteses performed for this indication
will diagnose a chromosomally abnormal conception (52). In the second trimester,
sonographic markers for fetal chromosome anomalies are observed in 3% to 5% of
pregnancies (53) (Table 10-2) and are another indication for fetal karyotyping. The
most worrisome of these findings are abnormalities of fetal neck, indicating the need for
evaluation of fetal chromosomes even in association with normal biochemical serum
screening in young patients (54).
TABLE 10-1 INDICATIONS FOR PRENATAL DIAGNOSIS
Increased risk of chromosome anomalies

Advanced maternal age
Previous offspring with chromosome anomalies
Parental balanced translocation or inversion
Ultrasonogram diagnosis of fetal malformations or anomalies
Abnormal biochemical screening in maternal serum
Previous offspring with neural tube defect
Parents carriers of a mendelian genetic trait
Molecular DNA diagnosis (e.g., cystic fibrosis, sickle cell anemia and fragile X
syndrome)
Enzymatic activity in villi or amniocytes (e.g., Tay-Sachs and Refsum disease)
Precursor levels in cell free amniotic fluid (e.g., 17-hydoxycortico- steroids
progesterone in congenital adrenal hyperplasia; 7-dehydrocholesterol in
Smith-Lemli-Opitz syndrome)
TABLE 10-2 RELATIVE RISK FOR ANEUPLOIDY ASSOCIATED WITH ISOLATED
SONOGRAPHIC MARKERS
Sonographic Marker
Prevalence
Choroid plexus cysts

Nuchal Thickening or
hygroma
>4 mm
>5 mm
>6 mm
Left ventricular echogenic
focus
Hyperechoic bowel
Pyelectasis
1.25%
4%5%
Fetal biometry
Short CRL
Short Femur Length
Short Humerus
Short Femur and
5%
0.6%0.8%
2%
7%
4%5%
4%5%
2.4%
Relative Risk*
18
28
36
4
1416
3.33.9
3
2.7
4.1
11.5
Humerus
* Risk for trisomy 21 as calculated in relation to maternal age alone or in
combination with biochemical screening.
Risk specific for trisomy 18 and possibly trisomy 21.
The most effective screening test for Down syndrome is probably the integrated test,
based on estimation of the nuchal translucency on ultrasonography, hCG, and
pregnancy-associated plasma protein A (PAPP-A) in the first trimester, combined with
AFP, HCG, estriol, and inhibin A in the second trimester of pregnancy. The integrated
risk assessment is reported to have a 94% detection rate and a 5% false-positive rate
(55). The results of biochemical or ultrasonography screening also can be used to
modify the risk of aneuploidy in the population previously considered at risk, reducing
the number of invasive diagnostic procedures in patients of advanced maternal age by
more than half (56,57). Furthermore, at-risk patients who previously declined
amniocentesis may be influenced to accept invasive prenatal diagnosis following a
positive screen result (57). However, a major problem of this integrated approach is
that results obtained in the first trimester are withheld from the patient, and the
advantages of first trimester diagnosis are lost.
The need for rapid karyotyping may arise when fetal anomalies are suspected near the
statutory limit for termination
P.137
of affected pregnancies, that is, after an abnormal result on biochemical or
ultrasonography evaluation. In those cases, the diagnostic options are late CVS (58)
or cordocentesis and karyotyping of fetal blood lymphocytes. Other tests that may be
performed on blood obtained by cordocentesis include hematologic parameters, acidbase balance, and immunologic status of the fetus (59).
Birth of a child with an inherited genetic disorder caused by malfunction in a single gene
implies a 25% to 50% risk of recurrence in subsequent gestations, depending on the
specific mode of inheritance of that disease. This risk and the associated genetic burden
of the disease may be considered so high that many couples will opt to avoid further
reproduction unless prenatal diagnosis is available. Fetal karyotype is uninformative in
these cases. Prenatal diagnosis of mendelian disease is performed using biochemical or
molecular techniques on fetal or placental tissue. Biochemical assays include
assessment of gene products such as enzymes, receptors, and transport proteins, and
metabolites such as amino acids, organic acids, vitamins, and hormones. When the
underlying biochemical defect is known and is expressed in accessible fetal tissue or
cells, prenatal diagnosis can be achieved by enzyme analysis of material obtained by
CVS, amniocentesis, or cordocentesis. Because variability caused by different mutations
and different genomic backgrounds exists among families, additional testing of
leukocytes or cultured skin fibroblasts from presumably unaffected parents and siblings
can provide valuable information. In addition to the benefit in interpretation of prenatal
results, such studies may provide a reliable means for identification of other carriers
among members of the extended family.
Prenatal diagnosis is now available for many inherited metabolic disorders. For an
autosomal recessive disease, biochemical assays can be used should discriminate
among homozygous affected, heterozygous unaffected, and homozygous normal
fetuses. Assays for detection of autosomal dominant diseases, such as some of the
porphyrias, usually are capable of identifying affected homozygotes, but sometimes fail
to differentiate conclusively affected heterozygotes from unaffected fetuses.
Heterozygote detection in X-linked disorders is difficult because of the random X
inactivation occurring in every pregnancy with a female fetus. Depending on the ratio of
an active mutant X to the normal X in tissues involved in the pathogenesis of the
disease, a female heterozygous for an X-linked disorder may be clinically normal, or
may have mild or even severe disease manifestations (60). To complicate matters
further, measured enzymatic activities also vary depending on the ratio of mutant to
normal X chromosomes that are active in the analyzed specimenchorionic villi, for
example. Occasionally, the activity levels in chorionic villi will not correlate with clinical
expression. Males, conversely, have only one X chromosome and are either hemizygous
affected with deficient enzyme activity or hemizygous normal with activity in the normal
range. Thus, prenatal biochemical assessment of X-linked disorders is less complicated
if the fetus is male.
The use of direct and cultured fetal specimens for prenatal evaluation of metabolic
disorders ideally requires the availability of normal control preparations. Except for
trophoblasts and amniotic fluid cells that can be maintained in culture, availability of
fresh controls is often a problem, and in most instances long-term frozen controls with
partial loss of activity must be used. There are other potential pitfalls that seem specific
for each of these tissue, cell, and fluid types. All samples should be analyzed as soon as
possible, except those requiring initial tissue culture. Chorionic villi, fetal tissue
biopsies, cell pellets, amniotic fluid supernatant, and fetal serum or plasma that are not
used for tissue culture can be kept frozen and shipped on dry ice. Cell and tissue
cultures, however, should be shipped at room temperature. Whenever possible,
appropriate controls matched by gestational age should accompany the samples to be
analyzed. Extraction and analysis of labile enzymes is especially difficult, because test
results are very sensitive with respect to the duration of homogenization or sonication.
Using fresh chorionic villi or amniocytes or freshly harvested trophoblasts helps
preserve the activity of such labile enzymes (61).
Fetal liver and muscle biopsy should be considered only in the absence of other
alternatives, because the risk for pregnancy loss associated with these invasive
procedures is significantly higher. Fetal muscle biopsy has been used in rare cases of
Duchenne muscular dystrophy (DMD), when molecular analysis of trophoblasts,
amniocytes, or fetal leukocytes is nondiagnostic and family studies are uninformative.
An in utero fetal muscle biopsy can be performed in the middle of the second trimester
to assess dystrophin levels in myoblasts by in situ hybridization (62). Absence of
dystrophin suggests an affected fetus.
Fetal liver biopsy also can be performed for certain rare enzyme deficiencies. For
example, in one type of glycogenosis, glucose-6-phosphatase is decreased; this enzyme
is expressed only in fetal liver and kidney. In the absence of direct DNA techniques, the
only option available for prenatal diagnosis is fetal liver biopsy in which glucose-6phosphatase activity can be measured. Fetal liver biopsy also is applicable in rare cases
of ornithine transcarbamylase deficiency where family studies are uninformative and
known deletions cannot be detected (63). In the future, it is probable that most of
these procedures will be considered obsolete, and prenatal diagnosis of most genetic
disorders will be performed by molecular analysis, including some mitochondrial
disorders that may be amenable to prenatal diagnosis (64).
Prenatal Diagnosis in the First Trimester

Chorionic Villus Sampling
The first attempts at first trimester prenatal diagnosis were made in the early 1970s for
fetal gender determination. Prenatal diagnosis did not gain popularity in the Western
world until 1983, when laboratory techniques to obtain adequate karyotypes from
chorionic villi were developed
P.138
by Simoni and Brambati in Milan (65). They also reported the first diagnosis of trisomy
21 by the direct method (66). Over the years, the quality of chromosome preparation
from CVS material has improved considerably, approaching the banding quality
obtained from amniocytes or blood karyotypes (67). The clinical procedure also has
been refined, with the use of real-time ultrasonographic guidance and malleable
catheters for villi aspiration.
Two types of cells are observed in chorionic villi. The outer layer consists of
cytotrophoblast, which divide spontaneously, and is used for direct evaluation of
metaphases. The inner mesenchymal core is used to initiate long-term cultures and is
usually considered as more representative of fetal karyotype. Fetal karyotype is
obtained from direct analysis, from long-term culture, or from both in 99.6% of
successful sampling of villi (67). Results of direct analysis are equivocal in 0.5% to 2%
of cases, but questions raised usually are resolved by long-term CVS or amniotic cell
cultures (68). Most commonly, an abnormal direct result that proves to be normal on
long-term culture will not be confirmed in karyotypes obtained from amniocytes or fetal
lymphocytes. The unusual situation is a normal direct result followed by an abnormal
result in long-term culture, with the abnormality being confirmed in fetal tissue in onehalf of cases (67). Maternal cell contamination has been observed in 1.9% of long-term
cultures, but usually is not observed in direct preparations and does not contribute to
diagnostic error in any case (68).
Chromosomal mosaicism in CVS material affects 1.2% to 2.5% of cases (average
1.3%), and is more common in direct preparations than in long-term culture (69).
Mosaicism was restricted to extraembryonic tissue in 70% to 80% of cases. Despite
mosaicism being confined to the placenta, followup of these cases documented a
significantly elevated fetal loss rate (7.5% to 16.7%), mostly in the second and third
trimesters, suggesting that such placental mosaicism is not entirely benign (70,71).
Intrauterine growth restriction (IUGR) also appears to be more common in this situation
(72). Mosaic trisomy 3 is one of the most common types of mosaicism observed in
placental cells (69). An adverse impact on pregnancy outcome seems to be most
common in association with confined placental mosaicism of chromosomes 13, 16, and
22 (73).
The diagnosis of mosaicism in CVS presents difficulties in genetic counseling because it
implies uncertainty with respect to fetal phenotype and genotype. Using both direct
preparation and long-term culture might increase the accuracy of diagnosis by CVS.
When further evaluation is needed, level II ultrasonography and amniocentesis is
generally adequate for followup. The use of early amniocentesis in these cases may
allow earlier clarification of fetal karyotype without a significant increase in procedure
related fetal loss rate (71). In our experience, however, the frequency of mosaicism in
amniotic fluid cultures is not significantly different from that observed in CVS (0.35%
vs. 0.56%, respectively)(74). Cordocentesis may also be used in such cases to verify
mosaicism in fetal blood (75). Even if fetal blood karyotype is normal, however, there
still remains a small chance that mosaicism is confined to specific fetal tissues, as
observed in trisomy 20 mosaicism.
CVS can be offered to almost every patient who needs prenatal diagnosis in the first
trimester. The most common indication for CVS is evaluation of fetal karyotype, which
is indicated for patients with advanced maternal age (70% to 80% of cases), a previous
child with chromosome anomalies, or a parent carrier of a balanced translocation or
inversion. Mendelian genetic disease can be diagnosed in at-risk cases by enzyme
level analysis in fresh or cultured villi (i.e., disorders such as Tay-Sachs disease or the
mucopolysaccharidoses)(76). However, it is crucial to obtain for analysis chorionic villi
that are only of fetal origin, with maternal cells either completely absent or extremely
rare. For most biochemical prenatal tests in the first trimester, the recommended
practice is to use fresh chorionic villi for preliminary evaluation followed by subsequent
analysis of cultured trophoblast for confirmation of the diagnosis. One exception is with
nonketotic hyperglycinemia (NKH). Although the glycine-to-serine ratio in amniotic fluid
is elevated in this disease, there is a significant overlap with normal values. The
potential for prenatal diagnosis of this disorder would rely exclusively on the results
obtained in fresh chorionic villi, because the glycine cleavage system is not expressed in
amniotic fluid cells or trophoblasts, but is detectable in fresh tissue (77). When the
enzyme in question is very labile (e.g., sialidase), or when its normal activity in
chorionic villi is extremely low, as for -iduronidase, the use of frozen controls may
cause false-negative diagnoses. Specific problems also may be encountered because of
different distribution of enzymes and isozymes (78). The characteristic presence of high
levels of arylsulfatase C activity in chorionic villi hampers the differential detection of
arylsulfatase A in metachromatic leukodystrophy and arylsulfatase B in
mucopolysaccharidosis VI.
Chorionic villus sampling is particularly suitable for prenatal molecular diagnosis of
mendelian genetic disease. The amount of DNA obtained from even a few villi is much
larger than that contained from 40 mL of amniotic fluid. The molecular techniques
described throughout this chapter and in a large number of other sources have
particular application to CVS material.
Technical Aspects of Chorionic Villus Sampling

CVS procedures are commonly performed between 11 and 13 weeks of gestation by the
transcervical, transabdominal, or transvaginal route (Fig. 10-4). After 13 weeks, the
procedure is best performed using a transabdominal approach (69,79). Fetal viability
and gestational age must always be confirmed before the procedure. Placental location
will usually dictate which approach to employ. Transcervical CVS is used when the
placenta is low lying, whereas fundal placentas are most commonly an indication for
transabdominal needle aspiration of the villi. Transcervical or transvaginal CVS should
be
P.139
avoided in patients who have an active vaginal or cervical infection (e.g., herpes); the
transabdominal approach also should be avoided in patients who have an interposed
bowel or marked uterine retroversion (68). In those cases, the procedure can be
performed transvaginally with a needle guided by transabdominal or transvaginal
ultrasonography (80).
Figure 10-4 In transcervical chorionic villus sampling (CVS), the CVS catheter
(arrows) is guided through the cervical canal and into the placenta.
Despite the preference of some American and European centers to use the
transabdominal approach for CVS (81,82), our experience is that transabdominal
sample size usually is lower than that obtained transcervically (83). Proficiency in both
types of procedures is necessary, however. Tailoring the type of procedure to placental
location is expected to reduce complication rates after CVS (83,84). It appears also that
transabdominal CVS is more suitable for testing multifetal pregnancies and can be
applied before fetal reduction (85). After the procedure, fetal heart activity should be
verified by ultrasonography. Maternal vaginal bleeding should also be assessed.
The safety of CVS must be judged in view of the natural pregnancy loss rate in the first
trimester. It is estimated that 3% to 5% of pregnancies are miscarried after
documentation of fetal heart beats on ultrasound at 8 to 11 weeks of gestation; the
likelihood of spontaneous abortion also increases with age (86). Both the Canadian and
the American collaborative studies documented an excess loss of pregnancy rate in the
CVS group of 0.6% to 0.8%, which was not significantly different than the loss rate in
the amniocentesis group (48,49). Fundal placental location, three catheter insertions,
and obtaining small amounts of villi are significantly associated with pregnancy loss
after CVS. These factors may reflect technical difficulty during the procedure (68).
Concerns over an increased risk of limb reduction defects (LRDs) following CVS were
raised in the early 1990s (87,88). Evaluation of more than 135,000 cases from
experienced centers worldwide reveals that the incidence of LRD or of any other defect
is identical to that of the background risk in this population (89). Furthermore, closer
assessment of cases of LRD reported after CVS has shown that several patients had
unaccounted for familial factors (90).
Because some of these cases were reported with CVS procedures done earlier than the
recommended gestational age, it is advisable to postpone CVS until 10 weeks of
gestation are completed. However, there are special circumstances under which earlier
CVS may be appropriate. In populations at high risk for genetic disorders there is
commonly a tendency for diagnosis as early in gestation as possible. For example, in
Orthodox Judaism, abortions are permitted, but only until 40 days past conception (i.e.,
54 days after the last menstrual period or 8 weeks). CVS in these patients has been
performed at 7 weeks to get answers in time (91). The ethics of using a procedure
with known higher risks to meet the theological needs of patients is a fascinating
subject that is beyond scope of this chapter.
Another concern regarding CVS is the potential for fetomaternal transfusion. Transiently
rising maternal serum AFP levels after CVS, suggest some transfer of fetal blood into
maternal circulation, this being in correlation with sample size (92). The calculated
mean volume of transfused fetal blood was 5.4 mL. Others have reported the volume of
fetomaternal transfusion after CVS to reach 21% of fetoplacental blood volume (69).
Thus, Rhesus (Rh) immunoprophylaxis (anti-D 300 g) should be administered to Rhnegative patients. Prior Rh isoimmunization should be considered a relative
contraindication to CVS.
Early Amniocentesis
Early amniocentesis refers to aspiration of amniotic fluid less than 15 weeks from the
last menstrual period. The procedure is technically similar to later amniocentesis
procedures; using continuous ultrasonography guidance and aseptic technique, a 22gauge needle is inserted into a pocket of fluid, and about 1 mL of fluid per week of
gestation is aspirated into a 20-mL syringe.
With improving ultrasonography technology and increasing experience with
ultrasonography-guided needle manipulations, early amniocentesis late in the first
trimester appeared, in the early to mid 1990s, to be an attractive alternative to CVS.
However, women who had early genetic amniocentesis were more likely to have more
postprocedure amniotic fluid leakage (2.9% vs. 0.2%), vaginal bleeding (1.9% vs.
0.2%), or fetal loss (2.2% vs. 0.2%) than were women undergoing amniocentesis at 16
to 19 weeks (93). Moreover, in approximately 5% of early procedures, aspiration of
fluid can be hampered by tenting of membranes; rotating the needle or the use of a
stylet longer than the needle allows the procedure to succeed in these situation (51,94).
The total unintentional loss rate after early amniocentesis is estimated as 1.4% to 4.2%
(93,95). It appears that the principal determinant of total fetal loss after any procedure
is gestational age. There seems to be a trend in all series toward increased loss rates
after amniocentesis performed at 11 or 12 weeks of gestation (96,97,98).
Early amniocentesis has a lower rate of pseudomosaicism and maternal cell
contamination than that observed in CVS analysis. Culture failure rates of early
amniocytes have been reported as 0.32% to 1.6% (95,96,99,100). In our experience,
culture failure rate is 1 in 700 after midtrimester amniocentesis, but approximately 1%
after early procedures
P.140
(nearly 5% when amniocentesis was performed before 12 weeks of gestation). It also
should be noted that failure of culture at amniocentesis has been reported more
commonly in association with fetal chromosome anomalies (97). The mean culture time
may also be somewhat longer after early amniocentesis than after midtrimester
procedures (98).
AFP peaks in amniotic fluid at 12 to 13 weeks of gestation and then gradually decline,
similar to fetal serum (99). Analysis of AFP and acetylcholinesterase (AChE) in early
amniotic fluid specimens may permit the diagnosis of fetal structural anomalies. High
amniotic fluid AFP levels have been observed with fetal neural tube defects (NTDs) or
omphalocele even in these early samples (100), and low amniotic fluid AFP values
accompanied some conceptions diagnosed as aneuploid, as seen in samples of amniotic
fluid obtained at midtrimester. The interpretation of AChE results in early amniotic fluid
specimens is more complex. Acetyl cholinesterase usually is analyzed on gel
electrophoresis as a bimodal result, either positive or negative. In early amniotic fluid
samples, a faint, inconclusive band is frequently observed, this being associated with
fetal anomalies only in a minority of cases (101). A normal sonogram in this situation is
also reassuring in regard to fetal neural or abdominal wall defects, but a higher rate of
adverse outcomes has been reported in these pregnancies (102).
First-Trimester Prenatal Diagnosis: Chorionic Villus

Sampling or Early Amniocentesis?
A prospective, randomized evaluation of CVS and first trimester amniocentesis suggests
that the two techniques are comparable in terms of accuracy and successful results
(103). It is apparent also that the two procedures are equally efficient in providing a
sample for evaluation (CVS 99.3%, early amniocentesis [EA] 100%) and in giving a
nonmosaic cytogenetic result (CVS 97.5%, EA 97.9%)(104). Spontaneous pregnancy
loss after early amniocentesis is, however, significantly higher than after CVS (4.9% vs.
2.1%, respectively). Similar results were concluded from analysis of the Cochrane
database: more spontaneous miscarriages (relative risk [RR] 1.92, confidence interval
[CI] 1.14 to 3.23) and more cases of neonatal talipes were observed in the EA group
(105). Concerns have also been raised that removal of amniotic fluid so early in
gestation may affect fetal lung development and function (106). Moreover, results of EA
are available about 4 weeks later than after CVS. Thus, in cases that are diagnosed as
abnormal, one advantage of first trimester CVS (that patients can have a less traumatic
and less risky termination of pregnancy) is lost when early amniocentesis is employed.
Our opinion is that CVS is the diagnostic procedure of choice before 14 weeks of
gestation for karyotyping, enzymatic, and DNA molecular analysis. Conversely, EA
(after 12 gestational weeks) may be preferable in certain biochemical disorders and in
situations in which CVS may not truly represent fetal tissue (i.e., contamination with
maternal cells or with tissue from a twin pregnancy with fused placenta).
Prenatal Diagnosis in the Second and Third
Trimesters
Midtrimester Amniocentesis
Midtrimester amniocentesis is the oldest, most commonly performed procedure for
prenatal diagnosis. It also is considered the gold standard to which other procedures for
prenatal diagnosis are compared. Indications for genetic amniocentesis include
increased risk for metabolic disorders, for chromosome anomalies or for structural
anomalies that may be associated with elevated AFP (see Table 10-1).
Some metabolic genetic disorders may be diagnosed by measurement of precursor
levels in cell-free fluid or by enzyme activity in cultured amniocytes. Elevated
concentrations of amino acids and organic acids in amniotic fluid serve as preliminary
indications for several inherited disorders, such as amino and organic acidopathies or
urea cycle defects.
Amniotic fluid supernatants should be divided into multiple vials to avoid the loss of
activity that occurs with repeated freezing and thawing. Determinations of amniotic
fluid concentrations of specific metabolites, as well as enzymes and other proteins,
usually serve as supporting evidence in prenatal diagnoses. The variability in enzyme
activities or in the levels of other proteins and metabolites frequently observed in
cultured amniotic fluid cells and trophoblasts can be minimized by a careful choice of
control cells. Preferably, the final diagnosis should rely on molecular analysis of specific
mutations or on demonstration of the underlying biochemical defect in fetal cells or
tissue (measuring the actual gene products responsible for the metabolic block).
In the cytogenetics laboratory, amniocytes are removed from amniotic fluid by
centrifugation and cultured in flasks or on cover glasses to grow in monolayers. Dividing
cells are arrested in metaphase, when chromosomes are maximally condensed, using
agents such as Colchimide that prevent spindle formation. The cells are then harvested
and placed in hypotonic saline, which causes intracellular swelling and better spreading
of the chromosomes during slide preparation. After fixation, the chromosomes are
stained with Giemsa or quinacrine for microscopic analysis. The use of triple gas
incubators, specific growth media (e.g., Chang) that enhance cellular proliferation, and
in situ culture on cover glass have shortened considerably the sampling-harvesting
interval; in most laboratories, the results are available within 1 to 2 weeks. Additional
improvement is obtained by computerized cytoanalyzers that expedite recognition of
metaphase spreads and obviate the need for darkroom and photography.
Cytogenetic analysis of amniotic fluid cells reflects fetal status accurately in more than
99% of cases, but mosaicism sometimes may confuse the interpretation of results. In
our experience, the frequency of results needing further investigation is similar in
cultures from amniocentesis and from CVS (74). Differentiation between cytogenetic
P.141
abnormalities that truly reflect fetal chromosome aberrations from those that are the
result of laboratory artifacts may be difficult. One or more hypermodal cells that are
limited to one colony or one culture flask are identified in 2% to 3% of all amniocytes
cultures and are usually associated with a normal phenotype (107). True fetal
mosaicism should be considered when hypermodal cells are identified in different

culture flasks or in separate colonies in the same culture flask. The frequency of
hypermodal cells in one or more colonies from the same flask is 0.7%. Hypermodal
cells with the same abnormality originating from multiple culture flasks are observed in
0.2% of amniotic cultures (108). Even in the latter situation, however, the abnormality
in culture may not represent true fetal mosaicism, as observed by Gosden and
colleagues (75). Fetal blood sampling for karyotype should help to avoid termination of
pregnancy of some normal fetuses in these cases.
The use of fluorescence in situ hybridization (FISH) is a valuable adjunct to standard
cytogenetics (109). With this technique, fluorescent probes identify chromosomespecific repetitive DNA, unique sequence regions, or site-specific cosmids in uncultured
cells. Fetal cells obtained at the time of amniocentesis or chorionic villus sampling can
be analyzed with commercially available probes for chromosomes 21, 18, 13, X, and Y,
the result being available in 24 hours. In addition, FISH analysis can be used to identify
microdeletions that cannot be observed with standard cytogenetics, that is,
microdeletions that may be found in Prader-Willi, Angelman, Williams, DiGeorge/
velocardiofacial or Smith-Magenis syndromes (110). It must be emphasized that FISH is
still an adjunct to standard cytogenetics and does not detect or identify some cases of
mosaicism, translocations, marker chromosomes, or other rare aneuploidies; standard
cytogenetics must still be performed. FISH is properly used on patients for whom rapid
diagnosis is essential, for example, those with abnormalities seen on ultrasonography,
or who are suspicious either at term or at the cusp of their legal options for termination,
or who have had multifetal pregnancies prior to reduction. We have found it to be very
reliable and efficient in allowing rapid diagnosis (111).
Technical Aspects of Amniocentesis

Genetic counseling is a prerequisite to amniocentesis as well as to any other procedure
for invasive prenatal diagnosis (112). Pedigree information obtained during the
counseling session will allow a more accurate ascertainment of genetic risks in patients
referred for other indication (e.g., advanced maternal age or abnormal serum
screening). Only experienced physicians trained in the procedure should perform
amniocentesis. In general, a detailed ultrasonographic examination is performed to
evaluate gestational age, placental location, and amount of amniotic fluid, and to
exclude fetal anomalies. After sterile preparation of the skin, a draped sterile
ultrasonography transducer is used to locate a suitable pocket of fluid. Then, under
continuous ultrasonographic guidance, a 20- to 22-gauge, 3.5-inch-long spinal needle is
inserted in a single, smooth motion into the pocket of fluid. When the needle, which is
visualized on ultrasonography as a bright spot (Fig. 10-5), is placed satisfactorily into
the pocket of amniotic fluid, the stylet is removed. Twelve to 20 mL (depending on the
laboratory) of amniotic fluid are then gently aspirated, transferred into sterile tubes,
and transported at room temperature to the laboratory for processing.
Figure 10-5 In amniocentesis, the needle (arrows) is inserted through the uterine
wall and into the amniotic cavity.
The risk of Rh isosensitization in Rh-negative women with Rh-positive fetuses has been
estimated to increase by 1% after amniocentesis (113). Thus, the patient's blood type
and antibody status should be known before amniocentesis, and unsensitized Rhnegative women should receive Rh immunoprophylaxis after the procedure.
Amniocentesis is a safe procedure in experienced hands. The procedure-related
pregnancy loss rate is 0.2% to 0.5% over and above the spontaneous loss rate at 16
weeks of gestation (the latter is estimated at 2% to 3%) (114). Pregnancy loss rates
seem to be associated with the number of failed needle insertions at the same session
and with vaginal bleeding after amniocentesis. Within reason, gestational age at the
time of amniocentesis, volume of fluid removed, and amniocentesis repeated at a
different session after a failed attempt do not seem to correlate with increased risk of
pregnancy loss. Transplacental amniocentesis does not appear to increase the rate of
fetal loss (115) provided the umbilical cord is avoided. However, the risk of Rh
isosensitization is apparently increased by transplacental passage of the needle. With
ultrasonography-guided amniocentesis, fetal injury by the needle should be very rare.
Leakage of amniotic fluid is a relatively frequent complication, affecting approximately
1% to 2% of patients after amniocentesis, but it usually is of minor long-term
consequence; in most cases, it resolves with bed rest for 48 to 72 hours (116). Even
patients with complete absence of fluid after amniocentesis may reaccumulate amniotic
fluid and go on to have normal outcomes. Thus, as long as there is no evidence of
infection, expectant management (for at least several days) seems prudent. Prolonged
amniotic fluid leakage, however, may lead to
P.142
severe oligohydramnios, which can result in fetal pressure deformities (e.g.,
arthrogryposis) and in pulmonary hypoplasia (117).
The risk of severe amnionitis endangering maternal health appears to be very low,
around 0.1% (114). However, injury to maternal bowel or blood vessels and rare cases
of mortality have also been reported (118).
Late Chorionic Villus Sampling

The technique of transabdominal CVS also has been applied successfully in the second
and third trimesters. Nicolaides and associates first reported six successful human
placental biopsies at 14 to 37 weeks of gestation, and suggested that CVS should not
be confined to the first trimester (119). Chieri and Aldini performed transabdominal
placental biopsy and amniocentesis in 220 patients at midtrimester; 210 of the
procedures were indicated for advanced maternal age (120). A sample adequate for
analysis (>2 mg of cleaned villi) was obtained in 90.9% of cases. The success rate was
94% with an anterior or fundal placenta, and 83.8% with a posterior placental location.
In the latter situation, the approach to the placenta was transamniotic. Cytogenetic
results were obtained in 95% of samples. No discrepancy could be found between
cytogenetic results obtained in villi or in amniotic fluid, and when fetal anomalies were
observed on CVS, they were acted on without waiting for corroboration from the
amniotic fluid culture (120). Several studies since (121,122) have documented a
sampling and diagnostic success rate of 98% to 99%, with a very low rate (0.3% to
1.8%) of procedure-related pregnancy loss, even with oligohydramnion.
The collaborative results of 2,058 late CVS procedures performed at 24 centers were
reported by Holzgreve and colleagues (58). A fetal karyotype was obtained in 96% of
procedures. The frequency of abnormal chromosome results was 21% when fetal
anomalies were observed on ultrasound, and 6.2% with normal ultrasonographic
findings. The pregnancy loss rate, excluding terminations, was 10.3% in the group with
abnormal ultrasonography findings, and 2.3% in the normal ultrasonography group.
Holzgreve and associates further presented their own data on 301 CVS procedures in
the second and third trimester, 225 (74.7%) of which were performed for abnormal
ultrasonography scans (123). Fetal karyotype was obtained in 99% of cases. The rate
of chromosome anomalies was 20% in the group with abnormal ultrasonographic scans,
and increased to 38% when the ultrasonography findings included abnormalities of
amniotic fluid volume.
It was apparent from these studies that despite the decrease in mitotic index with
placental aging, direct results could be obtained even with small amounts of placental
tissue. The quality of karyotypes obtained from analysis of second and third trimester
samples was similar to that obtained in the first trimester (124). Moreover, the risk of
maternal cell contamination in late CVS samples actually may be lower than that
observed in the first trimester because of less direct contact between villi and decidua
at later gestational age (125). Consequently, placental biopsies in the second and third
trimesters are technically feasible procedures for prenatal diagnosis, with results
apparently as accurate as those of amniocentesis and probably associated with a
smaller risk of pregnancy loss than percutaneous umbilical blood sampling. The major
advantage of late CVS is the possibility of obtaining rapid results in situations where
such information is needed for decisions about pregnancy termination or fetal therapy.
Such situations include the ultrasonography diagnosis of fetal anomalies late in the
second trimester, close to the legal limit in gestational age after which termination of
pregnancy is no longer possible. Late CVS also offers a distinct advantage over
cordocentesis in cases complicated by oligohydramnion. Prenatal availability of fetal
karyotype in pregnancies complicated by severe IUGR or fetal anomalies may influence
the mode of delivery, the management of intrapartum fetal distress, which is a common
phenomenon in fetuses with chromosome anomalies, or the decision for surgical
intervention within the first few hours after birth.
Cordocentesis
Freda and Adamson originally attempted to access the vascular system of the fetus for
treatment of Rh isoimmunization by hysterotomy and fetal exposure (126). This
method soon was abandoned because of the unacceptably high risk for the mother and
fetus. Subsequently, the development of fiberoptics allowed the introduction of
fetoscopy to visualize and sample vessels on the chorionic plate or the umbilical cord
(127). Although the risk of maternal compromise with this method was relatively small,
the residual high rate of pregnancy loss associated with fetoscopy (up to 11.3%) was
considered a major disadvantage (128).
Daffos introduced ultrasonography-guided percutaneous umbilical blood sampling
(PUBS) in 1983 for the diagnosis of fetal infections (129). The procedure gained rapid
and wide acceptance. In experienced hands, the risk of fetal loss is relatively small,
between 1% to 2.3%, although numbers as high as 5.4% have been reported
(130,131). Other complications, usually associated with excessive needle
manipulations, include hematoma of the umbilical cord and placental abruption,
chorioamnionitis (0.6%). and preterm delivery (9%)(131). Fetal exsanguination from
the puncture site is a relatively rare complication but has been reported (132). Maternal
complications are negligible, although one case of life-threatening amnionitis has been
reported (133). It appears that the risks are higher when the mother is obese, the
placenta is posterior, and the sampling is performed relatively early in gestation (i.e.,
before 19 weeks)(130).
Table 10-3 lists the indications for cordocentesis. With the development of better
molecular tests, the use of cordocentesis has dramatically decreased over the last 10
years. Parental counseling before cordocentesis should include the risk of that
pregnancy being affected by the conditions considered and the yield of information
obtained through fetal blood sampling in such a situation. The risk and
P.143
potential complications of the procedure itself also should be discussed. Before
cordocentesis, a detailed ultrasonography examination should evaluate gestational age,
placental location, and fetal anomalies. Fetal blood can be obtained by puncture of the
fetal heart, the intrahepatic part of the umbilical vein, or by puncture of an umbilical
vessel close to its placental insertion, the latter being by far the most common site for
cordocentesis. When the placenta is located on the anterior or lateral wall of the uterus,
the needle is introduced through the placenta into the umbilical cord (Fig. 10-6). In
cases with a posterior placenta, the needle is introduced through the amniotic fluid and
the cord is punctured close to its placental insertion. Different guidance techniques (i.e.,
fixed-needle guides vs. freehand), needles of lengths varying from 8 to 15 cm, gauges
varying from 20 g to 27 g, and differing patient preparation protocols are used by
various centers. Nicolaides and colleagues advocate an outpatient setting in the
ultrasonography department, without need for maternal fasting, sedation, tocolytics,
antibiotics, or fetal paralysis for the procedure (134).
TABLE 10-3 INDICATONS FOR CORDOCENTESIS
INDICATIONS FOR CORDOCENTESIS

Prenatal diagnosis of inherited blood disorders
Hemoglobinopathies (e.g., homozygous thalassemia, sickle-cell disease
Coagulopathies (e.g., hemophillia A and B, von Willebrand disease)
Prenatal diagnosis of metabolic disorders
Fetal Infections
Toxoplasmosis: specific IgM or DNA hybridization
Rubella: specific IgM
Cytomegalovirus: specific IgM and blood cultures
Varicella zoster virus: specific IgM
Human parvovirus (B19): viral DNA
Human immunodeficiency virus: specific IgM
Rapid fetal karyotyping
Late booking or failed amniocentesis
Suspected fetal mosaicism on amniocentesis
Abnormal maternal serum screening
Ultrasound diagnosis of fetal malformations
Evaluation of the small-for-gestational-age fetus
Acidbase status
Oxygenation
Assessment and treatment of fetal anemia
Diagnosis and treatment of fetal thrombocytopenia
Adapted from ref. 104.
Molecular diagnoses are now available for many of the mutations resulting in
thalassemia, sickle cell disease, and hemophilia A and B, enabling diagnosis in the first
trimester by CVS, with blood sampling performed only in noninformative cases or in
cases with ambiguous results. Cordocentesis is, however, necessary for the diagnosis
and management of congenital alloimmune thrombocytopenia (135). In alloimmune
thrombocytopenia, cordocentesis allows the determination of fetal platelet phenotype
and count. A low fetal platelet count in this situation can be treated by weekly infusion
of platelets until delivery (136).
In Rh isoimmunization, fetal blood sampling is performed for immediate confirmation of
fetal antigenic status, obviating the need for further intervention in the Rh-negative
fetus. If the fetus is Rh-positive, cordocentesis enables a more accurate assessment of
fetal anemia and an immediate rise in fetal erythrocyte count on correction by
intravascular transfusion. From case-control studies, it appears that at all gestational

ages and at all levels of disease severity, intravascular correction of fetal anemia is
more efficient and less risky to the mother and fetus than the intraperitoneal approach
(137). Moreover, in cases with cardiac decompensation in which the fetus may be
compromised by the volume overload needed to correct the severe anemia, better
results were obtained by intravascular exchange transfusion (138). It should be noted,
however, that cordocentesis might enhance maternal sensitization more than does
amniocentesis, especially if blood sampling is performed by a transplacental approach
(139). Rh immunoprophylaxis should be offered to all Rh-negative, nonsensitized
patients with an Rh-positive fetus undergoing cordocentesis.
The diagnosis of fetal infection is based commonly on the demonstration of the agentspecific immunoglobulin (Ig) M in fetal blood, because the large molecule IgM does not
cross the placenta. Fetal blood sampling should be scheduled to allow enough time from
initial exposure for IgM to appear after immunocompetence develops in the fetus. For
first trimester exposures, the best time for cordocentesis is probably after 20 weeks'
gestation. In a few specific cases, in utero treatment also is available. Thus, after
toxoplasmosis infection in the mother, and demonstration of IgM specific for
toxoplasmosis in fetal blood, antibiotic treatment with spiramycin reduced significantly
the risk of congenital toxoplasmosis, as well as the risk of late sequelae (140).
Cordocentesis has also
P.144
been used for repeated blood transfusions in utero to hydropic fetuses with hemolytic
anemia caused by parvovirus B19 infection (141).
Figure 10-6 is, the umbilical cord insertion in the placenta (small arrows) must be
located. The needle tip (bright spot) is placed in the umbilical vein (large arrow).
Severe, early onset IUGR commonly is associated with fetal chromosome anomalies.
Cordocentesis allows for rapid fetal karyotyping, which can be available within 48 to 72
hours. Other abnormalities observed in blood samples from IUGR fetuses with normal
chromosomes include hypoxemia, hypercapnia, lactic acidemia, leukopenia,

thrombocytopenia, and disturbed carbohydrate, lipid, and protein metabolism (142). A
statistically significant correlation was reported between increased umbilical systolic-todiastolic ratio (>3.5) and fetal hypoxia and acidemia in cord blood (143), these
relationships being even more pronounced with absent or reversed diastolic flow. Thus,
when a fetus is severely growth restricted, fetal pressure of oxygen (PO2) and pH may
provide additional and crucial information to the biophysical score and umbilical blood
flow studies, allowing us to decide between the risk of premature delivery and the risk
of leaving the fetus to grow in a hostile intrauterine environment.
Preimplantation Genetic Diagnosis

Preimplantation genetic diagnosis (PGD) was developed for patients at high risk of
transmitting an inherited disease to their offspring who, for whatever reason, do not
want to have even a first trimester abortion. Patients have to go through in vitro
fertilization (IVF) procedures to produce preimplantation embryos in vitro. One or two
cells are biopsied from the cleavage-stage embryos, at the 6- to 8-cell stage (day 3 of
development), and single-cell diagnosis with PCR or FISH is performed to determine
which embryos are free of the genetic disease. This technique has been applied for a
wide variety of disorders, including Tay-Sachs disease (144), -thalassemia (145),
cystic fibrosis (146), aneuploidy (147), and Marfan syndrome (148). Overall obstetric
outcome appears normal (149).
Because only normal embryos are transferred to the uterus, PGD virtually eliminates
the need for termination of an affected pregnancy. However, most centers still
recommend that prenatal diagnosis be done, as errors have been made.
PGD appears to be the perfect solution for those couples who are at high risk of
transmitting genetic diseases. However, for a multitude of reasons, relatively few
centers worldwide offer the procedure, and fewer than 3,000 cycles have been
performed in more than 10 years (150).
Single-Cell Diagnosis
Because it is difficult to karyotype single blastomeres, FISH has been used to analyze
chromosomes in embryos. Several studies have shown that human embryos exhibit a
high level of chromosome abnormalities, mainly mosaicism and chaotic chromosome
makeup (151,152). Mosaic embryos can lead to problems in terms of certain PGD
diagnoses because the blastomere biopsied may not be representative of the rest of the
embryo; in some circumstances, this can be minimized by analyzing two independent
blastomeres for PGD (152,153).
The most widespread application of FISH in prenatal diagnosis is for the rapid detection
(1 to 2 working days) of the common numerical chromosome abnormalities using
chromosome-specific probes applied to single blastomeres or to interphase cells from
amniocentesis and chorionic villi samples (154,155). Most commonly, probes specific
for chromosomes 13, 18, 21, X, and Y are used because, depending on the indications
for invasive tests, numerical abnormalities involving these chromosomes account for
70% to 95% of the chromosome aberrations identified prenatally (154,155,156). The

usefulness of interphase FISH in prenatal diagnosis and the test performance
characteristics were recently assessed (156). The detection rate was nearly 100%, with
a false-positive rate of only 0.003% and a false-negative rate of 0.024%. FISH results
are now commonly acted on if, for example, a common trisomy is found, without
waiting for the examination of a complete karyotype.
The diagnostic capability of interphase FISH is obviously limited by the choice of probe
used; structural chromosome aberrations and numerical chromosome abnormalities
other than those tested for will be missed. In addition, the incomplete hybridization of
FISH probes or a reduction in signal size because of heteromorphism (154,157) may
lead to false-negative prenatal FISH results. In cases of ambiguous FISH results,
however, full karyotype should remain the gold standard.
FISH has also been used for the PGD of structural chromosome abnormalities
(158,159). This has only been possible in recent years, as probes for the acrocentric
chromosomes and specific locus probes were only recently developed. PGD of
robertsonian translocations is a good example; minimum dual-color FISH with one
probe for each chromosome is possible, but three-color FISH, with two probes for
different areas of the most important chromosome, is better (160). PGD for reciprocal
translocations is more difficult because specific probes for the region in question are
required.
Another limiting factor in PGD for chromosome abnormalities is the high incidence of
abnormalities, including mosaicism, seen in preimplantation embryos (160). For some
translocations, 70% to 100% of embryos may be abnormal, consistent with the high
reproduction failure of such couples. PGD offers these couples a potential major
advantage in only transferring a chromosomally balanced embryo.
The techniques developed for PGD have also been applied to aneuploidy screening for
patients going through IVF procedures (PGD-AS)(161,162). PGD-AS was initially
performed for women of older reproductive age (older than age 35 years), with
recurrent IVF failure (more than three attempts with no pregnancy) and after recurrent
miscarriages when the couple have normal karyotypes (150). Several reports of an
increased pregnancy rate have been published (161,162,163), but a randomized
controlled clinical trial has not been performed.
P.145
Comparative genomic hybridization (164,165) has also now been successfully applied in
PGD-AS, with one normal live birth (165). Because the time required for comparative
genomic hybridization is 72 hours, embryos have to be frozen as they cannot be kept in
culture for that length of time.
Whether by cytogenetics, biochemical, or molecular methods, prenatal diagnosis of an
affected fetus requires obtaining for analysis either fetal tissue (i.e., blastomeres or
blood) or other tissue that is representative of the fetus (e.g., amniocytes or placenta).
Invasive procedures for prenatal diagnosis of fetal disease are available throughout
gestation, from the first trimester onward. Assisted reproductive technologies have
allowed preimplantation diagnosis (or exclusion) of several disorders. Despite its
limitations, as the technology for genetics and assisted reproductive technologies

improve, we believe there will be dramatic increases in the use of this modality.
REFERENCES
1. Garrod AE. The Croonian lectures. Lancet 1908;2:1.
2. Caskey CT. Disease diagnosis by recombinant DNA methods. Science

1987;236:12231229.
3. King CR. Prenatal diagnosis of genetic disease with molecular genetic technology.
Obstet Gynecol Surv 1988;43:493508.
4. Watson JD, Crick FHC. Molecular structure of nucleic acids: a structure for
deoxyribose nucleic acid. Nature 1953;171:737738.
5. Berget SM, Moore C, Sharp PA. Spliced RNA segments at the 5[T407]8-terminus of
late adenovirus 2 in RNA. Proc Natl Acad Sci U S A 1977;74:31713175.
6. Gilbert W. Why genes in pieces? Nature 1978;271:501.
7. Gilbert W. Genes-in-pieces revisited. Science 1985;228:823824.
8. Cooper DN, Smith BA, Cooke HJ, et al. An estimate of unique DNA sequence
heterozygosity in the human genome. Hum Genet 1985;69:201208.
9. Miller DA, Choi YC, Miller OJ. Chromosome localization of highly repetitive human
DNAs and amplified ribosomal DNA with restriction enzymes. Science 1983;219:395
397.
10. Caskey CT, Pizzun A, Ying-Hui Fu, et al. Triplet repeat mutations in human
disease. Science 1992;256:784789.
11. Richards RI, Sutherland GR. Heritable unstable DNA sequences. Nat Genet
1992;1:79.
12. ACMG/ASHG Statement. Laboratory guidelines for Huntington disease genetic

testing. Am J Hum Genet 1998;62:1243.
13. Poulton J, Marchington DR. Segregation of mitochondrial DNA (mtDNA) in human

oocytes and in animal models of mtDNA disease: clinical implications. Reproduction
2002;123:751755.
14. Smeitink J, van den Heuvel L, diMauro S. The genetics and pathology of oxidative
phosphorylation. Nature 2001;2:342352.
15. DiMauro S, Andreu AL. Mutations in mtDNA: are we scraping the bottom of the
barrel? Brain Pathol 2000;10:431441.
16. Petersen GM, Rotter JI, Cantor RM, et al. The Tay-Sachs disease gene in North
American Jewish populations: geographic variation and origin. Am J Hum Genet
1983;35:12581269.
17. Myerowitz R, Costigan C. The major defect in Ashkenazi Jews with Tay-Sachs
disease is an insertion in the gene for the alpha-chain for beta-hexosaminidase. J Biol
Chem 1988;263:1858718589.
18. Ohno K, Suzuki K. A splicing defect due to an exon-intron junctional mutation

results in abnormal beta-hexosaminidase X-chain in RNAs in Ashkenazi Jewish
patients with Tay-Sachs disease. Biochem Biophys Res Commun 1988;153:463469.
19. Cutting GR, Kasch LM, Rosenstein BJ, et al. Two cystic fibrosis patients with mild
pulmonary disease and nonsense mutations in each CFTR gene. N Engl J Med
1990;323:16851689.
20. Cutting GR, Kasch LM, Rosenstein BJ, et al. A cluster of cystic fibrosis mutations
in the first nucleotide-binding fold of the cystic fibrosis conductance regulator protein.
Nature 1990;346: 366369.
21. Kerem E, Corez M, Kerem BS, et al. The relationship between genotype and
phenotype in cystic fibrosis: analysis of the most common mutation (delta F508). N
Engl J Med 1990;323: 15171522.
22. Stern RC. The diagnosis of cystic fibrosis. N Engl J Med 1997;336:487491.
23. Kerem B, Chiba-Falek O, Kerem E. Cystic fibrosis in Jews: frequency and

mutation distribution. Genet Test 1997;1:3539.
24. Antonarakis SE, Phillips JA III, Kazazian HH Jr. Genetic diseases: diagnosis by
restriction endonuclease analysis. J Pediatr 1982;100:845856.
25. Gusella JF. DNA polymorphism and human disease. Annu Rev Biochem
1986;55:831854.
26. Gusella JF, Wexler NS, Conneally PM, et al. A polymorphic DNA marker
genetically linked to Huntington disease. Nature 1983;306:234238.
27. Antonarakis SE, Kazazian HH Jr, Orkin SH. DNA polymorphism and molecular
pathology of the human globin gene clusters. Hum Genet 1985;60:114.
28. Embury SH, Scharf SJ, Saiki RK et al: Rapid prenatal diagnosis of sickle cell
anemia by a new method of DNA analysis. N Engl J Med 1987;316:656661.
29. Smith CAB. The development of human linkage analysis. Ann Hum Genet
1986;50:293296.
30. Haldane JBS, Smith CAB. A new estimate of the linkage between the genes for
color-blindness and hemophilia in man. Ann Genet 1947;14:10.
31. Morton NE. Sequential tests for the detection of linkage. Am J Hum Genet
1955;7:277310.
32. Risch N. Genetic linkage: interpreting LOD scores. Science 1992;255:803804.
33. Southern EM. Detection of specific sequences among DNA fragments separated
by electrophoresis. J Mol Biol 1975;98:503517.
34. Conner BJ, Reyes AA, Morin C, et al. Detection of sickle cell beta S-globin allele
by hybridization with synthetic oligonucleotides. Proc Natl Acad Sci U S A
1983;80:278282.
35. Ehrlich HA, Gelfand D, Sninsky JJ. Recent advances in the polymerase chain
reaction. Science 1991;252:16431651.
36. Saiki RK, Gelfand DH, Stoffel S, et al. Primer-directed enzymatic amplification of
DNA with a thermostable DNA polymerase. Science 1988;239:487491.
37. Beggs AH, Koenig M, Boyce FM, et al. Detection of 98% of DMD/BMD gene
deletions by polymerase chain reaction. Hum Genet 1990;86:4548.
38. Grompe M, Muzny DM, Caskey CT. Scanning detection of mutations in human
ornithine transcarbamoylase by chemical mismatch cleavage. Proc Natl Acad Sci U S
A 1989;86:58885892.
39. Das S, Levinson B, Shitney S, et al. Diverse mutations in patients with Menkes
disease often lead to exon skipping. Am J Hum Genet 1994;55:883889.
40. Collins FS. Positional cloning: let's not call it reverse anymore. Nat Genet
1992;1:36.
41. Ruddle FH. The William Allan Memorial Award address: reverse genetics and
beyond. Am J Hum Genet 1984;36:944953.
42. Xu G, O'Connell P, Viskochil D, et al. The neurofibromatosis type 1 gene encodes

a protein related to GAP. Cell 1990;62: 599608.
43. Steel MW, Breg WR. Chromosome analysis of human amniotic fluid cells. Lancet
1966;1:383387.
44. Jacobson JB, Barter RH. Intrauterine diagnosis and management of genetic
defects. Am J Obstet Gynecol 1967;99:795801.
45. Nadler HL. Antenatal detection of hereditary disorders. Pediatrics 1968;42:912

918.
46. Medical Research Council. Diagnosis of genetic disease by amniocentesis during

second trimester of pregnancy. Ottawa, Ontario: Author, 1977.
47. National Institute of Child Health and Human Development Amniocentesis

Registry. The safety and accuracy of midtrimester amniocentesis. DHEW publication
no (NIH) 78190. Washington, DC: United States Department of Health, Education
and Welfare, 1978.
48. Evans MI, Drugan A, Koppitch FC, et al. Genetic diagnosis in the first trimester:
the norm for the 90s. Am J Obstet Gynecol 1989;160:13321336.
49. Canadian Collaborative CVS-Amniocentesis Clinical Trial Group. Multicenter
randomized clinical trial of chorionic villus sampling and amniocentesis. Lancet

1989;1:118.
P.146
50. Rhoads GG, Jackson LG, Schlesselman SE, et al. The safety and efficacy of
chorionic villus sampling for early prenatal diagnosis of cytogenetic abnormalities. N
Engl J Med 1989;320: 609617.
51. Hanson FW, Happ RL, Tennant FR, et al. Ultrasonography-guided early
amniocentesis in singleton pregnancies. Am J Obstet Gynecol 1990;162:13761381.
52. Drugan A, Reichler A, Bronshtein M, et al. Abnormal biochemical serum screening

versus second trimester ultrasound-detected minor anomalies as predictors of
aneuploidy in low-risk patients. Fetal Diagn Ther 1996;11(5):301305.
53. Drugan A, Johnson MP, Evans MI. Ultrasound screening for fetal chromosome
anomalies. Am J Med Genet 2000;90:98107.
54. Zimmer EZ, Drugan A, Ofir C, et al. Ultrasound anomalies of the fetal neck:
implications for the risk of aneuploidy and structural anomalies. Prenat Diagn
1997;17:10551058.
55. Wald NJ, Hackshaw AK. Advances in antenatal screening for Down's syndrome.
Baillieres Clin Obstet Gynaecol 2000;14: 563580.
56. Haddow JE, Palomaki GE, Knight GJ, et al. Reducing the need for amniocentesis
in women 35 years of age or older with serum markers for screening. N Engl J Med
1994;330:11141118.
57. Beekhuis JR, De Wolf BT, Mantingh A, et al. The influence of serum screening on
the amniocentesis rate in women of advanced maternal age. Prenat Diagn
1994;14:199202.
58. Holzgreve W, Miny P, Schloo R, et al. Late CVS international registry:

compilation of data from 24 centers. Prenat Diagn 1990;10:159167.
59. Hoskins IA. Cordocentesis in isoimmunization and fetal physiologic measurement,

infection and karyotyping. Curr Opin Obstet Gynecol 1991;3:266271.
60. Puck JM, Willard HF. X inactivation in females with X-linked disease [Editorial]. N
Engl J Med 1998;338:325328.
61. Ben-Yoseph Y, Evans MI, Bottoms SF, et al. Lysosomal enzyme activities in fresh
and frozen chorionic villi and in cultured trophoblasts. Clin Chim Acta 1986;161:307
313.
62. Evans MI, Krivchenia EL, Johnson MP, et al. In utero fetal muscle biopsy alters
diagnosis and carrier risks in Duchenne and Becker muscular dystrophy. Fetal Diagn
Ther 1995;10(2):7175.
63. Holzgreve W, Golbus MS. Prenatal diagnosis of ornithine transcarbamylase

deficiency utilizing fetal liver biopsy. Am J Hum Genet 1984;320328.
64. Faivre L, Cormier-Daire V, Chretien D, et al. Determination of enzyme activities

for prenatal diagnosis of respiratory chain deficiency. Prenat Diagn 2000;20:732737.
65. Simoni G, Brambati B, Danesino C, et al. Efficient direct chromosome analyses

and enzyme determinations from chorionic villi samples in the first trimester of
pregnancy. Hum Genet 1983;63:349357.
66. Brambati B, Simoni G. Letter to the editor. Lancet 1989;1:586.
67. Ledbetter DH, Martin AO, Verlinsky Y, et al. Cytogenetic results of chorionic villus
sampling: high success rate and diagnostic accuracy in the United States
collaborative study. Am J Obstet Gynecol 1990;162:495501.
68. Simpson JL. Chorionic villus sampling. Semin Perinatol 1990;14:446455.
69. McGowan KD, Blackemore KJ. Amniocentesis and chorionic villus sampling. Curr
Opin Obstet Gynecol 1991;3:221229.
70. Johnson A, Wapner RJ, Davis GH, et al. Mosaicism in chorionic villus sampling: an
association with poor perinatal outcome. Obstet Gynecol 1990;75:573577.
71. Sundberg K, Lundsteen C, Philip J. Early filtration amniocentesis for further

investigation of mosaicism diagnosed by chorionic villus sampling. Prenat Diagn
1996;16:11211127.
72. Kalousek DK, Dill FJ. Chromosomal mosaicism confined to the placenta in human
conceptions. Science 1983;221:665667.
73. Leschot NJ, Schuring Blum GH, Van Prooijen-Knegt AC, et al. The outcome of
pregnancies with confined placental chromosome mosaicism in cytotrophoblast cells.
Prenat Diagn 1996; 16:705712.
74. Wright DJ, Brindley BA, Koppitch FC, et al. Interpretation of chorionic villus
sampling laboratory results is just as reliable as amniocentesis. Obstet Gynecol
1989;74:739744.
75. Gosden C, Rodeck CH, Nicolaides KH. Fetal blood sampling in the investigation of
chromosome mosaicism in amniotic fluid cell culture. Lancet 1988;1:613617.
76. Evans MI, Moore C, Kolodny F, et al. Lysosomal enzymes in chorionic villi,
cultured amniocytes, and cultured skin fibroblasts. Clin Chim Acta 1986;157:109113.
77. Hayasaka K, Tada K, Fueki N, et al. Feasibility of prenatal diagnosis of nonketotic

hyperglycinemia: existence of the glycine cleavage system in placenta. J Pediatr
1987;110:124126.
78. Giles L, Cooper A, Fowler B, et al. Aryl sulphatase isozymes of chorionic villi:
implications for prenatal diagnosis. Prenat Diagn 1987;245252.
79. Podobnick M, Ciglar S, Singer Z, et al. Transabdominal chorionic villus sampling

in the second and third trimesters of high risk pregnancies. Prenat Diagn
1997;17:125133.
80. Sidransky E, Black SH, Soenksen DM, et al. Transvaginal chorionic villus
sampling. Prenat Diagn 1990;10:583586.
81. Brambati B, Oldrini A, Lanzani A. Transabdominal villus sampling: a free hand

ultrasound guided technique. Am J Obstet Gynecol 1987;157:134137.
82. Smidt-Jensen S, Hahnemann N. Transabdominal fine needle biopsy from

chorionic villi in the first trimester. Prenat Diagn 1984;4:163169.
83. Evans MI, Quigg MH, Koppitch FC, et al. First trimester prenatal diagnosis. In:
Evans MI, Fletcher JC, Dixler AO, et al, eds. Fetal diagnosis and therapy: science,
ethics and the law. Philadelphia: JB Lippincott, 1989:17.
84. Brambati B, Lanzani A, Tului L. Transabdominal and transcervical chorionic villus

sampling: efficiency and risk evaluation of 2411 cases. Am J Med Genet
1990;35:160164.
85. Eddleman KA, Stone JL, Lynch L, et al. Chorionic villus sampling before multifetal
pregnancy reduction. Am J Obstet Gynecol 2000;185:772774.
86. Simpson JL. Incidence and timing of pregnancy losses: relevance to evaluating
safety of early prenatal diagnosis. Am J Med Genet 1990;35:165173.
87. Firth HV, Boyd PA, Chamberlain P, et al. Severe limb abnormalities after chorionic
villus sampling at 5666 days' gestation. Lancet 1991;337:762763.
88. Burton BK, Schulz CJ, Burd LI. Limb anomalies associated with chorionic villus
sampling. Obstet Gynecol 1992;79:726730.
89. Kuliev A, Jackson L, Froster U, et al: Chorionic villus sampling safety. Report of
World Health Organization/EURO meeting. Am J Obstet Gynecol 1996;174:807811.
90. Schloo R, Miny P, Holzgreve W, et al. Distal limb deficiency following chorionic
villus sampling? Am J Med Genet 1992;42:404413.
91. Wapner RJ, Evans MI, Davis G, et al: Procedural risks versus theology: chorionic
villus sampling for Orthodox Jews at less than 8 weeks' gestation. Am J Obstet
Gynecol 2002;186(6):11331136.
92. Shulman LP, Meyers CM, Simpson JL, et al. Fetomaternal transfusion depends on
amount of chorionic villi aspirated but not on method of chorionic villus sampling. Am
J Obstet Gynecol 1990;162:11851188.
93. Brumfield CG, Lin S, Conner W, et al. Pregnancy outcome following genetic
amniocentesis at 1114 versus 1619 weeks' gestation. Obstet Gynecol
1996;88:114118.
94. Dombrowsky MP, Isada NB, Johnson MP, et al. Modified stylet technique for
tenting of amniotic membranes. Obstet Gynecol 1996;87:455456.
95. Stripparo I, Buscaglia M, Longatii L, et al. Genetic amniocentesis: 505 cases

performed before the sixteenth week of gestation. Prenat Diagn 1990;10:359364.
96. Elejalde BR, de Elejalde MM, Acuna JM, et al. Prospective study of amniocentesis
performed between weeks 9 and 16 of gestation: its feasibility, risks, complications
and use in early genetic amniocentesis. Am J Med Genet 1990;35:188196.
97. Reid R, Sepuvelda W, Kyle PM, et al. Amniotic fluid culture failure: clinical
significance and association with aneuploidy. Obstet Gynecol 1996;87:588592.
98. Diaz Vega M, De La Cueva P, Leal C, et al. Early amniocentesis at 1012 weeks
gestation. Prenat Diagn 1996;16:307312.
99. Drugan A, Syner FN, Greb A, et al. Amniotic fluid alpha-fetoprotein and
acetylcholinesterase in early genetic amniocentesis. Obstet Gynecol 1988;72:3538.
100. Crandall BF, Chua C. Detecting neural tube defects by amniocentesis between
11 and 15 weeks' gestation. Prenat Diagn 1995;15:339343.
P.147
101. Drugan A, Syner FN, Belsky RL, et al. Amniotic fluid acetylcholinesterase:
implications of an inconclusive result. Am J Obstet Gynecol 1988;159:469474.
102. Brown CL, Colden KA, Hume RF, et al: Faint and positive amniotic fluid
acetylcholinesterase with a normal sonogram. Am J Obstet Gynecol 1996;175:1000
1003.
103. Byrne D, Marks K, Azar G, et al. Randomized study of early amniocentesis

versus chorionic villus sampling: a technical and cytogenetic comparison of 650
patients. Ultrasound Obstet Gynecol 1991;1:235240.
104. Nicolaides KH, Brizot ML, Patel F, et al. Comparison of chorionic villus sampling
and early amniocentesis for karyotyping in 1492 singleton pregnancies. Fetal Diagn
Ther 1996;11:915.
105. Alfirevic Z. Early amniocentesis versus transabdominal CVS for prenatal

diagnosis Cochrane Database Syst Rev 2000;(2):CD000077.
106. Yuksel B, Greenough A, Naik S, et al. Perinatal lung function and invasive
antenatal procedures. Thorax 1997;52:181184.
107. Simpson JL. Amniocentesis: what it can tell you and what it can't. Contemp
108. Hsu LYF, Kaffe S, Perlis ET. Trisomy 20 mosaicism in prenatal diagnosis: a
review and update. Prenat Diagn 1987;7:581596.
109. Schwartz S. Efficacy and applicability of interphase fluorescent in situ

hybridization for prenatal diagnosis [Invited Editorial]. Am J Hum Genet
1993;52:851853.
110. Ligon AH, Beaudet AL, Sheffer LG. Simultaneous multilocus FISH analysis for
detection of microdeletions in the diagnostic evaluation of developmental delay and
mental retardation. Am J Hum Genet 1997;61:5159.
111. Evans MI, Henry GP, Miller WA, et al. International collaborative assessment of
146,000 prenatal karyotypes: expected limitations if only chromosome specific
probes and fluorescent in situ hybridization are used. Hum Reprod 2000;15(1):228
230.
112. Cohn GM, Gould M, Miller RC, et al. The importance of genetic counseling before
amniocentesis. J Perinatol 1996;16:352357.
113. Murray JC, Karp LE, Williamson RA, et al. Rh isoimmunization as related to
amniocentesis. Am J Hum Genet 1983;16:527534.
114. Drugan A, Johnson MP, Evans MI. Amniocentesis. In: Evans MI, ed.
Reproductive risks and prenatal diagnosis. Norwalk, CT: Appleton & Lange, 1992:191.
115. Bombard AT, Power JF, Carter S, et al. Procedure related fetal losses in
transplacental versus nontransplacental genetic amniocentesis. Am J Obstet Gynecol
1995;172:868872.
116. Crane JP, Rohland BM. Clinical significance of amniotic fluid leakage after
genetic amniocentesis. Prenat Diagn 1986;6:2531.
117. Nimrod C, Varela-Gittings F, Machin G, et al. The effect of very prolonged

membrane rupture on fetal development. Am J Obstet Gynecol 1984;148:540543.
118. Chervenak JL, Kardon NB. Advancing maternal age: the actual risks. Female
Patient 1991;16(11):1724.
119. Nicolaides KH, Soothill PH, Rodeck CH, et al. Prenatal diagnosis: why confine
chorionic villus (placental) biopsy to the first trimester? Lancet 1986;1:543544.
120. Chieri PR, Aldini AJR. Feasibility of placental biopsy in the second trimester for
fetal diagnosis. Am J Obstet Gynecol 1989;160:581583.
121. Ko TM, Tseng LH, Hwa HL, et al. Prenatal diagnosis by transabdominal chorionic
villus sampling in the second and third trimesters. Arch Gynecol Obstet
1995;256:193197.
122. Podobnik M, Ciglar S, Singer Z, et al. Transabdominal chorionic villus sampling

in the second and third trimesters of high-risk pregnancies. Prenat Diagn
1997;17:125133.
123. Holzgreve W, Miny P, Gerlach B, et al. Benefits of placental biopsies for rapid
karyotyping in the second and third trimesters (late chorionic villus sampling) in high
risk pregnancies. Am J Obstet Gynecol 1990;162:11881192.
124. Smidt-Jensen S, Lundsteen C, Lind AM, et al. Transabdominal chorionic villus

sampling in the second and third trimester of pregnancy: chromosome quality,
reporting time and feto-maternal bleeding. Prenat Diagn 1993;13:957969.
125. Ganshirt-Ahlert D, Pohlschmidt M, Gal A, et al. Transabdominal placental biopsy

in the second and third trimester of pregnancy: what is the risk of maternal
contamination in DNA diagnosis? Obstet Gynecol 1990;75:320323.
126. Freda VJ, Adamson KJ. Exchange transfusion in utero. Am J Obstet Gynecol
1964;89:817821.
127. Rodeck CH, Cambell S. Umbilical cord insertion as source of pure fetal blood for
prenatal diagnosis. Lancet 1979;1:12441245.
128. Ward RHT, Modell B, Fairweather DVI. Obstetric outcome and problems of
midtrimester fetal blood sampling for antenatal diagnosis. Br J Obstet Gynaecol
1981;88:10731080.
129. Daffos F, Cappella-Pavlovsky M, Forestier F. Fetal blood sampling via the

umbilical cord using a needle guided by ultrasound: report of 66 cases. Prenat Diagn
1983;3:271277.
130. Buscaglia M, Ghisoni L, Bellotti M, et al. Percutaneous umbilical blood sampling:

indication changes and procedure loss rates in a nine years' experience. Fetal Diagn
Ther 1996;11:106113.
131. Bernaschek G, Yildiz A, Kolankaya A, et al. Complications of cordocentesis in

high risk pregnancies: effects on fetal loss and premature deliveries. Prenat Diagn
1995;15:9951000.
132. Seligman SP, Young BK. Tachycardia as the sole fetal heart rate abnormality
after funipuncture. Obstet Gynecol 1996;87:833834.
133. Wilkins I, Mezrow G, Lynch L, et al. Amnionitis and life threatening respiratory
distress after percutaneous umbilical blood sampling. Am J Obstet Gynecol
1989;160:427428.
134. Nicolaides KH, Soothill PW, Rodeck CH, et al. Ultrasound guided sampling of
umbilical cord and placental blood to access fetal well being. Lancet 1986;1:1065
1067.
135. Bussel JB, Berkowitz RL, McFarland JG, et al. Antenatal treatment of neonatal
thrombocytopenia. N Engl J Med 1988;319:13741378.
136. Murphy MF, Pullon HWH, Metcalfe P, et al. Management of fetal allo-immune
thrombocytopenia by weekly in utero platelet transfusions. Vox Sang 1990;58:4549.
137. Harman CR, Bowman JM, Manning FA, et al. Intrauterine transfusion:
intraperitoneal versus intravascular approach: a case control comparison. Am J
Obstet Gynecol 1990;162:10531059.
138. Poissonier MH, Brossard Y, Demedeiros N, et al. Two hundred intrauterine

exchange transfusions in severe blood incompatibilities. Am J Obstet Gynecol
1989;161:709713.
139. Weiner CP, Grant S, Hudson J, et al. Effect of diagnostic and therapeutic
cordocentesis on maternal serum alpha-fetoprotein concentration. Am J Obstet
Gynecol 1989;161:706708.
140. Daffos F, Forestier F, Capella-Pavlovsky M, et al. Prenatal management of 746

pregnancies at risk for congenital toxoplasmosis. N Engl J Med 1988;318:271275.
141. Peters MT, Nicolaides KH. Cordocentesis for the diagnosis and treatment of
human fetal parvovirus infection. Obstet Gynecol 1990;75:501504.
142. Weiner CP. The relationship between the umbilical artery systolic/diastolic ratio
and umbilical blood gas measurements in specimens obtained by cordocentesis. Am J
Obstet Gynecol 1990;162:11981202.
143. Ribbert LSM, Sniders RJM, Nicolaides KH, et al. Relationship of fetal biophysical
profile and blood gas values at cordocentesis in severely growth retarded fetuses. Am
J Obstet Gynecol 1990;163:569571.
144. Gibbons WE, Gitlin SA, Lansendorf SE, et al. Preimplantation genetic diagnosis
for Tay-Sachs disease: successful pregnancy after pre-embryo biopsy and gene
amplification by polymerase chain reaction. Fertil Steril 1995;63(4):723728.
145. Ray PF, Kaeda JS, Bingham J, et al. Preimplantation genetic diagnosis of thalassaemia major. Lancet 1996;347:1696.
146. Strom CM, Verlinsky Y, Milayeva-Rechitsky S, et al. Preconception genetic

diagnosis for cystic fibrosis by polar body removal and DNA analysis. Lancet
1995;336:306307.
147. Verlinsky Y, Cieslak J, Friedine M, et al. Pregnancies following preconception

diagnosis of common aneuploidies by fluorescence in situ hybridization. Hum Reprod
1995;10:19231927.
148. Harton GL, Tsipouras P, Sisson ME, et al. Preimplantation genetic testing for
Marfan's syndrome. Mol Hum Reprod 1996;2(9):713715.
149. Soussis I, Harper JC, Handyside AH, et al. Obstetric outcome of pregnancies
resulting from embryos biopsied for pre-implantation diagnosis of inherited disease.
Br J Obstet Gynaecol 1996;103:784788.
150. ESHRE PGD Consortium Steering Committee. ESHRE Preimplantation Genetic

Diagnosis (PGD) Consortium: data collection III (May 2001). Hum Reprod
2000;17:233246.
151. Harper JC, Coonen E, Handyside AH, et al. Mosaicism of autosomes and sex
chromosomes in morphologically normal, monospermic pre-implantation human
embryos. Prenat Diagn 1995;15:4149.
P.148
152. Delhanty JDA, Harper JC, Ao A, et al. Multicolor FISH detects frequent
chromosomal mosaicism and chaotic division in normal pre-implantation embryos
from fertile patients. Hum Genet 1997;99:755760.
153. De Vos A, Van Steireghem A. Aspects of biopsy procedures prior to preimplantation genetic diagnosis. Prenat Diagn 2001: 21:767780.
154. Tepperberg J, Pettenati MJ, Rao PN, et al. Prenatal diagnosis using interphase
fluorescence in situ hybridization (FISH): 2-year multi-center retrospective study and
review of the literature. Prenat Diagn 2001;21:293301.
155. Cheong Leung W, Chitayat D, Deaward G, et al. Role of amniotic fluid interphase
fluorescence in situ hybridization (FISH) analysis in patient management. Prenat
Diagn 2001;21:327332.
156. Evans MI, Henry GP, Miller WA, et al. International collaborative assessment of
146,000 prenatal karyotypes expected limitations if only chromosome-specific probes
and fluorescent in situ hybridization are used. Hum Reprod 1999;14:12131216.
157. Weremowicz S, Sandstrom DJ, Morton CC, et al. Fluorescence in situ

hybridization (FISH) for rapid detection of aneuploidy: experience with 911 prenatal
cases. Prenat Diagn 2001; 21:262269.
158. Fridstrom M, Ahrlund-Richter L, Iwarsson E, et al. Clinical outcome of treatment

cycles using PGD for structural chromosomal abnormalities. Prenat Diagn
2001;21:781781.
159. Munne S, Sandalinas M, Escudero T, et al. Outcome of pre-implantation genetic

diagnosis of translocations. Fertil Steril 2000;73:12091218.
160. Munne S, Magli C, Cohen J, et al. Positive outcomes after pre-implantation

diagnosis of aneuploidy in human embryos. Hum Reprod 1999;14:21912199.
161. Magli MC, Sandalinas M, Escudero T, et al. Double locus analysis chromosome
21 for pre-implantation genetic diagnosis a of aneuploidy. Prenat Diagn
2001;21:10801085.
162. Conn CH, Harper JC, Winston RML, et al. Infertile couples with robertsonian
translocations: pre-implantation genetic analysis of embryos reveals chaotic cleavage

divisions. Hum Genet 1998;102:117123.
163. Verlinsky Y, Cieslak J, Ivakhnenko V, et al. Preimplantation diagnosis of

common aneuploidies by the first- and second-polar body FISH analysis. J Assist
Reprod Genet 1998; 15:285289.
164. Vouliare L, Slater J, Williamson R, et al. Chromosome analysis of blastomeres

from human embryos by using CGH. Hum Genet 2000;105:210217.
165. Wilton L, Williamson R, McBain J, et al. Birth of a healthy infant after preimplantation confirmation of euploidy by comparative genomic hybridization. N Engl J
Med 2001;345:15371541.
Chapter 11
Feto-Maternal Interactions: Placental Physiology, the
In Utero Environment, and Fetal Determinants of Adult
Disease
Gabriella Pridjian
The human placenta has become a highly evolved, sophisticated interface between mother and fetus. As the
gatekeeper for maternofetal interactions, its functions are diverse and essential. The in utero environment not
only influences fetal and newborn growth and development, but possibly the development of adult disease.
HUMAN PLACENTATION
Based on the modified classification of Grosser, which separates placentas by the number of layers interfacing
the maternal and fetal circulations, the human placenta is hemomonochorial, with only the syncytiotrophoblast,
fetal connective tissue, and fetal capillary endothelium forming the barrier between the two circulations (Fig. 111)(1). The human placenta, with this most intimate interface, is similar to that of the guinea pig and the
monkey. Placentation in other animals is different, and there is great variation in mammalian placentation. For
example, the ovine placenta is epitheliochorial and almost impermeable to diffusional transfer of the ketone
body -hydroxybutyrate, but human placenta is permeable to this ketone body, which has been implicated with
the fetal distress accompanying diabetic ketoacidosis (2). The human placenta is discoid and made up of 8 to 10
cotyledons. Fetal blood is supplied to the placenta by two umbilical arteries and drained by one umbilical vein.
On the fetal surface of the placenta, the umbilical arteries, crossing over fetal veins, decrease in caliber and
increase in divisions as they travel toward the placental edges and dive deeply into the placental disc to supply
individual cotyledons. Within the substance of the placenta, the caliber of the arteries decreases until only fetal
capillaries exist at the level of the terminal villi. The fetal capillaries are dilated, providing a broad surface area
for maternofetal transfer. The maternal blood supply to the placenta originates from the uterine artery, which
divides into spiral arteries and percolates through the intervillous space, bathing the terminal villi.
PLACENTAL TRANSFER
The fetus depends almost exclusively on the placenta for nutritional, respiratory, and excretory functions. The
placenta, growing steadily as gestation progresses, parallels fetal growth. Studies of placental growth and
physiology in disease states suggest that placental growth and size are determined by the fetus and modulated
by maternal factors. Normal placental-to-fetal weight ratios are approximately 1:6. As the placenta grows,
villous processes increase in number as fetal vasculature expands, and by the third trimester, a large surface
area is available to the maternal and fetal circulations.
Most placental transport is transcellular. Although the placenta often is thought of as a separating membrane,
it is actually a series of membranes. The most efficient areas for maternofetal exchange are the epithelial plates,
which consist of thinly stretched, attenuated villous tissue separating maternal blood in the intervillous space
from fetal
P.150
blood in the fetal sinusoids. To cross epithelial plates from the maternal to the fetal side, a substance must
traverse:
The brush border membrane of the syncytiotrophoblast
The cellular plasma of this cell
The basement membrane of the syncytiotrophoblast
The maternal side of the fetal capillary endothelial cell
The fetal side of the same endothelial cell
Figure 11-1 Electron micrograph of the placental barrier of the human hemomonochorial full-term placenta.
Notice the numerous endocytotic vesicles in various stages of formation located on the maternal-side brush
border membrane of the syncytiotrophoblast. b, basement membrane; FC, fetal capillary; MBS, maternal blood
space or intervillous space; Tr, syncytiotrophoblast with microvillous brush border membrane. Bar = 0.5 mm.
(From Thornberg KL, Faber JJ. Placental physiology. New York: Raven Press, 1983:19, with permission.)
The microvillous brush border membrane of the syncytiotrophoblast appears to be the membrane most involved
in regulation of transport, especially of active or carrier mediated transport. Certain diffusible substances
traverse the trophoblast and endothelial cell intact for release on the fetal side, some substances may be
partially or completely metabolized by the placenta, and others may be involved in intricate transport systems
(Fig. 11-2).
Simple Diffusion
Many nutrients, metabolites, and excretory products cross the placenta by diffusion. Diffusion of substances in
the placenta depends on multiple factors, as summarized in Table 11-1.
The amount of a nutrient delivered to the placenta is directly proportional to its concentration in the maternal
bloodstream, which depends on nutritional intake and gastrointestinal absorption. Famine, maternal
gastrointestinal diseases that interfere with absorption, or maternal pulmonary diseases that interfere with
alveolar exchange may significantly affect blood concentrations and the transfer and accrual of fetal fuels. A lack
of fetal fuels produces fetal and placental growth restriction. Abnormalities in maternal homeostatic mechanisms
may produce either an insufficiency or an abundance of nutrients. For example, in poorly controlled diabetes,
maternal hyperglycemia, hyperaminoacidemia, and hypertriglyceridemia allow un restrained nutrient delivery to
the fetus with excessive growth of fetal organs, body fat, and the placenta (3).
Figure 11-2 Patterns of maternofetal transport. A:Minimal or no placental uptake and no fetal transfer (e.g.,
succinylcholine, highly charged quaternary ammonium compounds). B: Placental uptake, degradation, and no
fetal transfer (e.g., insulin). C:Placental uptake and transfer predominantly unmodified to the fetus (e.g., hydroxybutyrate, bilirubin).D:Placental uptake, partial use, and transfer to the fetus (e.g., oxygen, glucose,
amino acids, free fatty acids). E: Uptake, partial metabolism, and transfer to the fetus (e.g., cyclosporine). F:
Uptake, modification, and transfer to the fetus (e.g., 25-hydroxyvitamin D3, of which most undergoes 1hydroxylation in the placenta to form 1,25-dihydroxyvitamin D3). G: Carrier-coupled uptake with release of the
ligand to the fetal side and regeneration of the carrier on the maternal side (e.g., transferrin- iron complex).
Delivery of a nutrient to the placenta is directly proportional to blood flow in the intervillous space. Maternal
blood volume gradually increases to 30% to 40% higher than prepregnancy volume, with 40% directed to the
uterus and placenta. Maternal cardiac disease with lower cardiac output may result in fetal and placental growth
restriction. Even in healthy women, maternal position influences blood flow to the uterus. Normal pregnant
women have an 18% lower cardiac output in the standing position compared with lying on their sides, perhaps
explaining why women who stand at work throughout their pregnancy have newborns of lower birth-weight.
Fetal factors influencing diffusion are those that affect nutrient delivery to the fetal side of the placenta. The
concentration of a substance in the umbilical artery depends on the amount of prior placental transfer,
absorption from swallowed amniotic fluid, and fetal metabolism. Fetal blood flow to the uterus depends on fetal
cardiac output
P.151
and placental vascular tone. Normally, fetal vessels on the chorionic plate are maximally dilated, providing the
least resistance to flow.
TABLE 11-1 FACTORS AFFECTING THE PLACENTAL TRANSFER OF A DIFFUSIBLE SUBSTANCE
Maternal Factors
Placental Factors
Fetal Factor
Amount Delivered to the

Intervillous Space
Transfer Physiology
Amount Delivered to the Fetal

Capillaries
Blood concentration
Exogenous and endogenous
supplies
Homeostatic mechanisms
Arteriovenous mixing in the
intervillous space
Flow rate in intervillous space
Area of diffusing membrane(s)
Blood concentration
Diffusion resistance
Characteristics of transferred
material (size, charge, polarity,
shape)
Characteristics of membrane
(physiochemical composition,
Hemodynamic factors in mother fluidity)
Local circulatory factors
Diffusion pressure across each
Shunting
Fetal metabolic production or

gastrointestinal absorption
Prior placental transfer
Flow rate in the fetal capillaries
Hemodynamic factors in fetus
Local circulatory factors
Shunting
placental cell membrane

Maternofetal concentration
gradients
Placental cellular production
or use
Maternofetal blood flow
characteristics; intervillous flow
Numerous placental factors influence diffusion. Overall, transfer is governed by the quantity of epithelial plates,
which are specialized regions of enhanced diffusion where the interhemal barrier is less than a few micrometers.
The human placenta has an intervillous pool flow system in which fetal capillaries in terminal villi are bathed in a
maternal blood reservoir continuously filled by arteries and drained by veins (Fig. 11-3). Concurrent and
countercurrent flows exist in areas of uneven distribution of flow (i.e., shunting), where a portion of the villus is
well supplied by maternal blood but poorly supplied by fetal blood; in other areas, the opposite occurs.
Figure 11-3 Areas of concurrent and countercurrent flow exist in the intervillous pool flow system of the
human placenta.
Stereochemical characteristics of a substance are major factors in transferability. Small, compact, nonpolar, lipophilic substances are transferred most efficiently. The placenta is relatively impermeable to large, polar
molecules that do not have specific transport systems or carrier proteins or are unable to take advantage of an
analogous transport system to aid in their transfer. For example, -fetoprotein (AFP), a 70-kd fetal protein, does
not transfer to the maternal side in appreciable amounts despite large quantities in fetal blood. Maternal AFP is
derived from transplacental transfer from fetal blood and transmembrane (i.e., chorioamnion) transfer from
amniotic fluid. The fetal blood AFP level at 17 weeks of gestation is approximately 3 mg/mL when the maternal
blood level is approximately 0.1 mg/mL, resulting in a fetomaternal gradient of approximately 30,000:1. The low
level of fetomaternal transplacental transfer allows detection of elevated maternal serum levels from
transmembrane (i.e., amniochorion) transfer of abnormally high amniotic fluid AFP, which provides the basis of
maternal serum AFP screening for neural tube defects. False-positive elevations of maternal serum AFP (i.e.,
high maternal serum value with a structurally normal fetus) suggest placental microabruptions, or loss of
integrity of the maternofetal barrier, and forecast a higher rate of fetal morbidity.
Membrane characteristics regulate transport. The fluidity of the membrane, determined by the degree and
character of membrane-incorporated phospholipids, influences transfer of certain substances. Diseases, such as
diabetes, may influence membrane fluidity (4).
The major driving force in favor of transfer by diffusion is the concentration gradient across the placenta; the
resistance to diffusion is dictated by the nature of the molecule. The principles of diffusion of molecules that are
generally applicable to biological membranes hold true in the placenta, although specifics remain to be defined.
Availability of a substance for diffusional transfer across the placenta is not always related to blood levels of that
substance,
P.152
because many metabolites, nutrients, and drugs that are poorly water-soluble are protein bound.
Although proteins aid in delivery of these substances to the placenta, they may actually hinder transfer of the
substances. It is the free, unboundor solublefraction of a substance that is available for transfer. Conversely,
high-affinity carrier proteins on the receiving side of the placenta drive diffusional transfer to their side by
decreasing a ligand's free fraction and increasing its maternofetal gradi ent. Oxygen, for example, is 98% bound
to hemoglobin. It is the transplacental difference in the partial pressure of dissolved oxygen (PO2) that
determines the diffusion pressure. The more oxygen-avid fetal hemoglobin counterbalances the resistance to
transfer from the maternal circulation. The O2 content (i.e., dissolved and hemoglobin-bound O2) of the blood on
each side of the placental membrane is determined principally by different affinities of maternal and fetal
hemoglobin for oxygen. In humans, the fetal oxyhemoglobin dissociation curve is displaced leftward of the
maternal curve, facilitating a much greater uptake of oxygen by fetal blood at the placental capillary level than
would be possible otherwise (see Appendix B). At any given PO2, a much higher O2 content is achieved in fetal
blood than in maternal blood. The O2 content in the umbilical vein (14.5 mL/dL) is as high as that of the uterine
artery (15.8 mL/dL), despite an umbilical venous PO2 of only 27 mm Hg (Table 11-2). Relatively high fetal blood
O2 content confers on the fetus the ability to deliver sufficient oxygen to peripheral tissue despite low PO2. Low
PO2 may be essential to fetal physiologic adaptation to maintain high pulmonary vascular resistance and to keep
the ductus arteriosus open.
The excretion of bilirubin provides an example of fetomaternal interaction using specific permeability properties
of the placenta to accomplish a given objective (5). Before birth, elimination of bilirubin from the fetus is by
diffusional transfer through the placenta to the mother. The placenta is extremely permeable to unconjugated
bilirubin but relatively impermeable to bilirubin glucuronide (i.e., conjugated bilirubin). In the fetus, because of
minimal bilirubin glucuronyltransferase, hepatic conjugation of bilirubin is suppressed. Because fetal bilirubin is
predominantly unconjugated and highly lipid soluble, it diffuses freely from the fetal to the maternal side. After
transfer to the mother, it is efficiently conjugated and excreted (Fig. 11-4).
TABLE 11-2 NORMAL OXYGEN VALUES IN MATERNAL AND FETAL BLOOD
Uterine Artery
Uterine Vein
Umbilical Vein
Umbilical Artery
PO2 (torr)
Oxygen Measurements
95
40
27
15
Hemoglobin O2 saturation
98
76
68
30
(%)
O2 content (mL/dL)
15.8
12.2
14.5
6.4
Hemoglobin (g/dL)
12.0
12.0
16.0
16.0
From Longo L. Disorders of placental transfer. In: Assali NS, ed. Pathophysiology of gestation. New
York: Academic Press, 1972;2:11, with permission.
Facilitated Diffusion
Most substances cross the placenta by simple diffusion. Maternal glucose, the principal substrate for oxidative
metabolism in the fetus, is a water-soluble, polar molecule that crosses the placenta by facilitated diffusion,
which is a gradient-dependent, receptor-mediated, saturable process. In the human placenta, preferential
transfer of D-glucose (over L-glucose) exists. Transfer stereospecificity implies a carrier-mediated process that
provides the fetus with the appropriate isomer for metabolism. The presence of glucose transporter genes in the
placenta, which code for glucose transporter proteins, confirms indirect experimental evidence for the existence
of a membrane-bound D-glucose carrier protein (6,7). Under physiologic and pathologic human conditions, the
carrier protein for glucose is not saturated, and the amount transferred to the fetus is directly related to the
amount supplied to the placenta (8).
Active Transport
To provide appropriate fuels for fetal growth, specific energy-requiring transport mechanisms in the microvillous
surface aid in transfer of substances that are not readily lipid soluble and are required in large amounts by the
fetus.
Most amino acids cross the placenta by an active transport mechanism (9,10,11). Active amino acid uptake has
two major purposes: transfer to the fetus and placental production of peptide hormones.
Transfer from maternal to fetal circulation is especially important for the essential amino acids required for fetal
growth, including the essential adult amino acids histidine, isoleucine, leucine, lysine, methionine, phenylalanine,
threonine, tryptophan, and valine; and the proposed fetal essential amino acids cysteine, tyrosine, histidine, and
taurine. Early in development, before maturation of fetal metabolic systems, all amino acids are essential to the
fetus. Fetal amino acid levels are 1.5- to 5-fold higher than maternal levels, confirming a transport process
against a concentration gradient.
Placental transfer of amino acids is stereospecific, with the natural l-form preferred. Transport of amino acids by
animal cells is mediated by specific carrier systems that have overlapping substrate reactivities. In human villous
P.153
tissue fragments, three carrier systems exist for neutral amino acids (12). System A is sodium dependent,
reversible at low pH, and most reactive with amino acids that have short, polar, or linear side chains (e.g.,
alanine, glycine). System L is sodium independent and most reactive with large, apolar, branched-chain, and
aromatic amino acids (e.g., leucine, isoleucine, tyrosine, tryptophan, valine, phenylalanine, methionine,
glutamine). The ASC system is sodium dependent and is involved in transport of alanine, serine, and cysteine
(ASC). Evidence suggests that a B system exists in placenta for taurine transport (13). Taurine, although
produced by the maternal liver from cysteine and methionine, is essential for fetal neurologic development but is
not produced by the fetus.
Figure 11-4 Antepartum excretion of bilirubin. Fetal bilirubin is transferred from fetal serum albumin through
the placenta to maternal serum. It then is conjugated with glucuronic acid by the maternal liver and excreted
into the bile. Fetal glucuronidation is suppressed. The placenta is relatively impermeable to the glucuronide.
Certain drugs cross the placenta by active transport. Zidovudine, used for treating HIV, has been found in the
perfused human placental model to cross from the maternal to the fetal side by energy-dependent transport
(14). Because zidovudine is a thymidine analog, it may take advantage of placental thymidine transport
systems. Zidovudine levels are higher in cord blood than in maternal blood, suggesting transport against a
concentration gradient and an active transport mechanism. Maternal administration of zidovudine and other
antiretroviral agents has become accepted therapy in HIV-positive mothers for prevention of vertical
transmission to the fetus. In a collaborative, prospective study of pregnant, HIV-seropositive women, zidovudine
administered orally in the prenatal period and intravenously during labor was shown to decrease the vertical
transmission to the fetus from 25.5% to 8.3% (15).
Receptor-Mediated Endocytosis
Although many large protein molecules cross the placenta by pinocytosis in extremely small quantities, specific
receptor-mediated processes expedite transfer of certain larger substances that are required by the fetus. The
receptor-rich microvillous brush border of the syncytiotrophoblast and the numerous coated micropinocytotic
vesicles found just beneath it provide anatomic evidence for receptor-mediated endocytosis (16). The receptors
involved in this process, found on the surface of the syncytiotrophoblast, are thought to extend through the
glycocalyx layer of the cell membrane and bind to the protein clathrin to form a membrane complex. After the
ligands are bound to their receptors, aggregation and internalization occur to form a cytoplasmic-coated vesicle
(Fig. 11-5). Destiny of the contents of the vesicles depends on the ligand.
Maternal immunoglobulin (Ig) molecules are transferred to the fetus by receptor-mediated endocytosis. IgG
subclasses 1 and 3 and IgA are known to cross the placenta. Once internalized, the intact Ig molecules within
the vesicles are delivered from the cytoplasm of syncytiotrophoblast through the capillary endothelial cell and
into the fetal circulation (17,18). Antenatal fetal transfer of mater nal IgG antibodies may interfere with
antibody-based
P.154
diag nostic testing in the fetus, necessitating analysis of the fetal-specific IgM antibody. Developmentally, the
transfer of maternal IgG to the fetus is probably protective and beneficial, but this transfer backfires in some
situations, such as immune fetal hydrops (i.e., erythroblastosis fetalis) and alloimmune fetal thrombocytopenia.
By receptor-mediated endocytosis, anti-D or another blood group antibody crosses the placenta to cause fetal
hemolytic anemia, and anti-PlA1 crosses the placenta to cause fetal thrombocytopenia (19).
Figure 11-5 Receptor-mediated endocytosis. The placental syncytiotrophoblast with the microvillous maternal
border (Fig. 11-1) has (A) specific receptors, located in the microvillous projections (B) clustering in intervening
pits on exposure to specific ligand in the maternal blood stream. Endocytosis occurs. C: The receptor-ligand
complexes and associated cell wall invert to form (D) an endocytic vesicle that is internalized. The destiny of
the vesicle depends on the ligand.
Transfer of transferrin-iron complex into the placental syncytiotrophoblast occurs through receptor-mediated
endocytosis. Brush border membrane transferrin-specific receptors on the maternal side of the
syncytiotrophoblast bind transferrin-iron complex, and aggregate and internalize it to form vesicles of transferriniron complexes. In the cytoplasm, the complexes dissociate to form apotransferrin and ferrous iron.
Apotransferrin is recycled to the maternal circulation, and ferrous iron is stored transiently as ferritin and
released to the fetal circulation to be made into a complex with fetal transferrin. No maternal transferrin or
placental ferritin is transferred to the fetus (20). Maternofetal iron transport is independent of maternal levels.
Uptake of low-density lipoprotein (LDL) cholesterol from maternal blood for progesterone synthesis by the pla
cental trophoblast is accomplished through receptor-mediated endocytosis. Specific receptors that have a high
affinity for LDL but not for high-density lipoprotein (HDL) are located on the microvillous brush border of the
syncytiotrophoblast. LDL binds to its receptor and is actively internalized. Within the cytoplasm, LDL vesicles
fuse with lysosomes, where enzyme hydrolysis of cholesterol esters releases cholesterol for mitochondrial
synthesis of progesterone.
Other Mechanisms of Transfer

A variety of other mechanisms of transfer probably are functional in the human placenta. Large protein
molecules may cross the placental membranes by a slow, non-receptor- mediated process of pinocytosis. Ions
may cross placental membranes with the aid of ion pumps. Evidence suggests that small molecules and ions
may cross through intercellular channels.
PLACENTAL METABOLISM
The placenta is a highly metabolic organ. Oxygen is consumed at a rate of 10 mL/min per kg, representing the
amount of maternal oxygen needed to supply the placenta and fetus for metabolic functions. Approximately
20% of placental oxygen uptake is used by the placenta; the remainder diffuses to the fetus. Glucose, the
principle metabolic carbon source of the placenta, is converted to lactate or is oxidized to CO2. Placental tissue
requires energy for maintenance of active transfer systems, hormone production, and substrate metabolism.
PLACENTA AS AN ENDOCRINE ORGAN

It is the placenta that maintains the maternal milieu most favorable for pregnancy by elaboration of large
amounts of steroid and peptide hormones, which are required to maintain the fetoplacental unit. Generally,
placentally produced steroid hormones, but not peptide hormones, cross to the fetal side.
Human Chorionic Gonadotropin
Sensitive techniques demonstrate that the secretion of human chorionic gonadotropin (hCG) begins during
implantation, when the cytotrophoblast differentiates into the syncytiotrophoblast. Although the messenger RNA
for hCG can be found in the cytotrophoblast, this cell is thought not to be the origin of this peptide hormone, but
only gains the ability to secrete hCG after it differentiates into a syncytiotrophoblast. The maternal plasma hCG
level rises after implantation, peaks by 10 menstrual weeks of pregnancy, and declines to a nadir in the second
trimester, after which levels remain low (Fig. 11-6).
The only well-established role of hCG is continued stimulation of the ovarian corpus luteum to produce 17hydroxyprogesterone for maintenance of pregnancy. Although placental production of progesterone occurs early
in gestation, the transition to placental autonomy from the ovary occurs
P.155
between 10 and 12 menstrual weeks. Before this transition, loss of the corpus luteum results in loss of the
pregnancy, unless exogenous progesterone is administered. Primary control of trophoblastic hCG production has
not been determined, but hormonal modulation is apparent (21). Proposed roles for hCG include the
immunologic protection of the trophoblast and regulation of placental progesterone production. Falling levels of
hCG before 10 menstrual weeks heralds pregnancy loss and is associated with miscarriage or ectopic gestation.
Higher-than-normal hCG levels are seen with multiple gestations, hydatidiform mole, choriocarcinoma, fetal
triploidy when associated with molar changes of the placenta, and Down syndrome (22).
Figure 11-6 Maternal blood levels of the major hormones produced by the placenta throughout pregnancy.
(From Ashitaka Y, Nishimura R, Takemori M, et al. Production and secretion of hCG and hCG subunits by
trophoblastic tissue. In: Segal S, ed. Chorionic gonadotropins. New York: Plenum Press, 1980:151; Selenkow
HA, Varma K, Younger D, et al. Patterns of serum immunoreactive human placental lactogen and chorionic
gonadotropin in diabetic pregnancy. Diabetes 1971;20:696; and Speroff L, Glass RH, Kase NG. Clinical
gynecologic endocrinology and infertility, 4th ed. Baltimore: Williams & Wilkins, 1989, with permission.)
Human Placental Lactogen

Human placental lactogen (hPL) is a single-chain polypeptide that has approximately 85% similarity to human
growth hormone. The quantity of hPL synthesized by the placental syncytiotrophoblast parallels placental mass,
reaching its peak at term and falling dramatically after delivery of the placenta.
Functionally, hPL can be considered a fetal growth hormone, because it maintains the maternal metabolic milieu
optimal for delivery of nutrients to the fetus. Despite its name, hPL has not been demonstrated conclusively to
exert a lactogenic effect in humans. The physiologic role of hPL appears to be in shifting the pattern of maternal
energy metabolism during pregnancy from carbohydrate to one that depends on fat. The hormone promotes
adipolysis and increases free fatty acid availability for maternal metabolism, saving glucose and amino acids for
transfer to the fetus. Free fatty acids do not cross the placenta as readily as amino acids and glucose.
Human placental lactogen has antiinsulin effects thought to be mediated by the elevated free fatty acids, which
promote peripheral tissue resistance to insulin. The subsequent increased pancreatic production of insulin leads
to down regulation of peripheral insulin receptors.
Maternal blood levels of hPL correlate with placental function. It was once thought that low hPL levels could
predict pregnancies with deteriorating placental function and those with fetal compromise (23). Unfortunately,
hPL levels are not as clinically useful as other methods. Similarly, elevated maternal hPL levels were thought to
be predictive of gestational diabetes or outcome in preexisting diabetes, but large biological variations in
maternal levels preclude its use in prediction or diagnosis. There are no clinical applications for hPL levels at this
time.
Estrogen
Estrogen production by the syncytiotrophoblast requires an elaborate concerted effort by the mother, fetus, and
placenta (Fig. 11-7). Because there is no activity of 17-hydroxylase and 17,20-desmolase in the human
placenta, estro gen precursors must be obtained from the fetal adrenal gland. The placenta produces three
major estro gensestradiol (E2), estriol (E3), and estrone (E1)which are secreted predominantly into the
maternal circulation. Maternal estrogen levels increase with gestational age.
Little is known about the specific functions of estrogen during pregnancy. Estrogens effect many general
changes in the mother to prepare for and maintain pregnancy. The uterine myometrium responds exquisitely
with increased protein synthesis and cellular hypertrophy. Estrogens cause vascular relaxation and increased
blood flow to the uterus. Uterine contractility is increased by estrogens, supporting a role in the onset of
parturition. Placental sulfatase deficiency, an X-linked fetal disorder, is associated with low estrogen levels.
Except for dysfunctional labor, women with these fetuses have normal pregnancies.
A specific role for estrogens in the fetus has not been determined. The fetal liver can metabolize E3 to estetrol
(E4), which binds to fetal estrogen receptors but has no estrogenic activity, protecting fetal tissue from massive
amounts of free estrogen.
Progesterone
The placental syncytiotrophoblast produces progesterone from maternally derived LDL cholesterol (Fig. 11-8).
Fetal
P.156
contribution to progesterone synthesis is minimal. Maternal progesterone levels increase with gestational age.
The major role of progesterone is to maintain the pregnancy. Early production of progesterone is by the ovarian
corpus luteum. After a transition period of shared function between 6 and 12 weeks of gestation, the placenta
becomes the dominant producer of progesterone, and the pregnancy continues even if the corpus luteum is
removed. Low levels of progesterone may be associated with first-trimester pregnancy loss.
Figure 11-7 Placental estrogen synthesis from fetal and maternal precursors. After 20 weeks of gestation, the
fetal compartment supplies most steroid precursors for placental estrogen production. The fetal adrenal uses
low-density lipoprotein cholesterol, produced by the fetal liver or transferred from the maternal compartment,
to synthesize dehydroepiandrosterone sulfate (DS). DS is converted to 16-OH-DS in the fetal liver. DS and
16-OH-DS undergo placental metabolism to estradiol (E2) and estriol (E3), respectively, which are released
predominantly on the maternal side.
Figure 11-8 Placental progesterone synthesis depends only on maternal precursors.
The most important role of progesterone may be that of principal substrate for fetal adrenal gland production of
glucocorticoids and mineralocorticoids. Progesterone may have a role in parturition and in suppressing the
maternal immunologic response to fetal antigens.
AMNIOTIC FLUID
Formation and circulation of the amniotic fluid reflect intimate and dynamic maternal and fetal interactions.
Amniotic fluid is ultimately derived from maternal water. Very early in pregnancy, amniotic fluid is cellular
transudate with the same tonicity, but lower protein content, as maternal plasma. By at least 8 gestational
weeks, when the maternal and fetal blood circulations are well established, most amniotic fluid water is thought
to be derived from maternal plasma water by direct transfer from the maternal circulation to fetal capillaries in
response to osmotic and hydrostatic forces. Once circulating in the fetus, water is filtered and excreted by the
urinary system into the amniotic cavity. By 8 gestational weeks, the urethra is patent, and the fetal kidneys
begin to form urine; by 10 to 11 weeks, a fetal bladder can be seen ultrasonographically. Concurrently, the fetus
begins to swallow. Swallowed amniotic fluid is

P.157
reabsorbed into the fetal circulation to be reexcreted by the kidneys or transferred across the placenta to the
mother. By the end of the first trimester, the amniotic fluid circulation has been established. Before fetal skin
keratinization at 22 weeks, additional water transfer can occur directly through the highly permeable fetal skin.
The content of amniotic fluid changes over gestation. The early electrolyte content is similar to that of
extracellular fluid. As the kidneys become functional, fetal urine electrolytes and excretory products become
major components. As gestation progresses, the fetal kidneys mature and are better able to retain electrolytes
and produce more dilute urine. At 20 weeks, the sodium content of amniotic fluid is 136 mEq/L and the
osmolality is 276 mOsm/L; at 40 weeks, the sodium content is 124 mEq/L and the osmolality is 258 mOsm/L.
Amniotic fluid contains excretory products of the fetus and maternal products that can diffuse directly across the
fetal membranes from the maternal compartment. In addition to electrolytes and proteins, amniotic fluid
contains carbohydrates, amino acids, urea, creatinine, lactate, pyru vate, lipids, enzymes, hormones, and
various other metabolites reflective of the fetal milieu. Their presence and the presence of desquamated fetal
cells allow diagnosis of many fetal abnormalities by biochemical and genetic analysis of amniotic fluid. Amniotic
fluid also contains fetal pulmonary fluid. The usefulness of the lecithin to sphingomyelin ratio in predicting
maturity of the fetal lungs has its basis in contributions from this fluid. In fetal sheep, there is constant
outpouring of pulmonary fluid into the trachea (24).
Although the range of normal amounts of amniotic fluid varies and depends on gestational age, abnormal
volumes often herald fetal structural or growth abnormality. Maternal factors can affect amniotic fluid volume.
Maternal plasma volume correlates with amniotic fluid volume. In hypovolemic mothers, expansion of plasma
volume with albumin increases amniotic fluid volume (25). Maternal use of diuretics may influence amniotic fluid
volume indirectly by decreasing maternal intravascular volume and directly by increasing fetal micturition after
transplacental passage.
FETAL MEMBRANES
The fetal amnion and chorion, although simple in anatomic design, are intricately involved in fetomaternal
interactions. The thin, avascular layer of epithelial cells that makes up the amnion arises from fetal ectodermal
cells, and the chorion, several layers thick, arises from extraembryonic somatic mesoderm and a trophoblast
layer. The trophoblast layer in the area of the chorion, which is destined to be the fetal surface of the term
placenta, undergoes rapid proliferation and branching into villi. By 8 menstrual weeks, the trophoblast layer of
the remaining chorion becomes compressed, attenuated, and microscopic. These microscopic cells are intimately
intermingled with the out ermost maternal layer, the decidua or gestational endometrium, to allow paracrine
interaction of these cells.
Paracrine interactions of the fetal chorionic cells with maternal decidual cells may be involved in the control of
maternal production of prolactin and amniotic fluid volume regulation. Transport studies with tritiated water
suggest that the net transport of water by the chorioamnion with adherent decidua is greatest in the
fetomaternal direction, suggesting a net outflow of water from the amniotic fluid compartment to maternal
circulation (26). Lower osmolality in the amniotic fluid compartment favors movement of water from the
amniotic cavity to the maternal compartment.
The cells of the amnion, rich in esterified arachidonic acid, are active in prostaglandin metabolism and are at
least indirectly involved in cervical ripening and the onset or maintenance of labor. Initiation of human labor
may involve autocrine and paracrine mechanisms within the fetal membranes and possibly maternal decidua,
resulting in amnion cell production of prostaglandin E2 (PGE2), a potent cervical-ripening and uterotonic agent
(27). Although amnion cell production of PGE2 has been associated with initiation and maintenance of labor,
control of its production is less well understood. Various inflammatory cytokines, in particular, tumor necrosis
factor interleukin-1, interleukin-1, interleukin-6, and interleukin-8, play a role affecting the common
pathway of amnion cell production of PGE2 and subsequent labor (28,29,30,31).
UMBILICAL CORD
The umbilical cord contains one fetal vein and two fetal arteries, which are supported and protected by Wharton
jelly, a gel-like connective tissue composed of a ground substance of open-chain polysaccharides in a network of
collagen and microfibrils (32). Externally, the cord is covered by amnionic epithelium, but there is no chorionic
epithelium. The length of the umbilical cord ranges from 30 to 100 cm (mean 55 cm) with reported extremes
from 0 to 155 cm. Fetuses with no umbilical cords have a severe, fatal abdominal wall defect because of failure
of formation of the body stalk. The association of a short cord with a low IQ raises the question of whether the
length of the cord is determined by fetal movement mediated by antenatal neurologic function. Fetuses with
long cords are more likely to have cord entanglement. Short cords are more likely to stretch and avulse during
descent of the fetus, resulting in signs of fetal distress during expulsion or hemorrhage at birth.
The normal umbilical cord increases in circumference with gestation until term, when the average circumference
is 3.8 cm (33). Although excessive Wharton jelly (i.e., thick cords) has not been associated with fetal
abnormalities, lack of this connective tissue has. Thin cords frequently are seen with growth-retarded fetuses
and may be associated with cord strictures, umbilical vessel rupture, or thrombi (34). Thin cords are more likely
to allow symptomatic
P.158
external compression, stretch, or occlusion of umbilical vessels. Resolution of umbilical venous bleeding after
percutaneous umbilical blood sampling is facilitated by Wharton jelly surrounding the vessel.
PLACENTAL PHYSIOLOGY IN DISEASE STATES

Preeclampsia and Hypertensive Disorders of Pregnancy
Preeclampsia, a hypertensive disease unique to pregnancy, has its genesis in the placenta. It is more common in
women with multiple gestations (i.e., multiple placentas) and molar pregnancies (i.e., excessive trophoblastic
tissue), and it abates after delivery of the placenta (35). Altered placentation may be responsible for maternal
vasospasm, the basic physiologic abnormality in preeclampsia (36). Vascular constriction in most maternal
organs results in hypoperfusion of the uterus and placenta, making the fetus at risk for intrauterine growth
restriction, placental abruption, and fetal death.
Placentas of preeclamptic women have a significantly lower total volume and volume of parenchymal and villous
surface area than placentas of normal controls (37). There is a greater proportion of centrally located, infarcted
areas because of poor intervillous maternal circulation. Histologically, cellular degenerative changes are found
(38). An abundance of syncytial budding is common in histologic sections and teased preparations of villi.
Budding, or knotting of the placental syncytiotrophoblast, consists of bunching of the cytoplasm of these cells
with aggregation of nuclei, a histologic finding suggestive of perfusional compromise.
The maternal vascular response to placentation is inadequate in women who develop preeclampsia (39).
Normally during implantation, cytotrophoblasts, or early placental cells, invade the lumen of the uterine spiral
arterioles, anchor to the vessel wall, and cause dilatation of the arterioles to allow adequate perfusion of the
placenta and, ultimately, the fetus. There is evidence of a vascular transformation of the cytotrophoblasts that
invade the spiral arterioles manifested by at least modulation of adhesion molecules (40). In preeclampsia there
is a loss of the normal modulation of adhesion molecules and an incomplete vascular invasion by the
cytotrophoblasts (41). Some spiral arteries show absence of physiologic changes throughout the entire length,
suggesting a complete lack of invasion by trophoblasts (Fig. 11-9).
Figure 11-9 Maternal blood supply to the placenta in normal pregnancy (left) and preeclampsia (right). Notice
the lack of normal physiologic dilation of radial arteries and of some decidual segments of spiral arteries in
preeclampsia. (From Khong TY, De Wolf F, Robertson, WB, et al. Inadequate maternal vascular response to
placentation in pregnancies complicated by preeclampsia and by small for gestational age infants. Br J Obstet
Gynaecol 1986;93:1049, with permission.)
Certain spiral arterioles at the implantation site undergo acute atherosis. Initially, fibrinoid degeneration and
mural thrombosis of decidual vessels occurs. The vessel wall then becomes replaced by fibrin, and the intima is
replaced by cholesterol-laden macrophages. Eventually, fibrinoid necrosis and total obstruction of the lumen lead
to loss of maternal blood flow, which is likely responsible for placental infarcts (42).
There are no microscopic or macroscopic placental changes that are pathognomonic for preeclampsia. The
pathologic changes in placentas of preeclamptic women suggest the disease is indistinguishable from that of
women with lupus anticoagulant syndrome, signifying a shared physiology. In the lupus anticoagulant syndrome,
autoimmune antiphospholipid antibodies, detectable in the maternal bloodstream, are associated with maternal
arterial and venous thrombosis, recurrent miscarriage, early onset preeclampsia, placental and fetal growth
restriction, and fetal death. Placentas from these women, whether or not they show signs of overt preeclampsia,
have infarctions, fibrosis, a decrease in vasculosyncytial membranes, and an increase in syncytial knots (43).
Preeclampsia is a disease resulting from abnormal maternoplacental interaction, and its cause is unknown. More
recently, preeclampsia has been considered a syndrome with various etiologies leading to a final common
P.159
constellation of signs and symptoms, hypertension, proteinuria, and edema. One of the more attractive
pathogenetic mechanisms of preeclampsia is that of an immune-mediated disease, supported by similarities with
the lupus anticoagulant syndrome. However, in preeclampsia, neither the exact antigenic stimulus nor the
antibody response has been defined, although the former probably is derived from trophoblasts.
Preeclampsia may be caused by placental production of a circulating humoral factor, perhaps immunologic,
which affects prostaglandin homeostasis. Prostaglandins are involved in normal vasodilation of pregnancy.
Prostacyclin (PGI2), because of its potent effect in relaxing vascular smooth muscle and lowering systemic
arterial pressure, is thought to be most involved. Prostacyclin is produced in vascular endothelial cells and is
thought to exert its effect on vascular smooth muscle in a paracrine fashion. It is an inhibitor of platelet
aggregation and of uterine contractility. The combined effects of this prostaglandin prevent maternal
hypertension, prevent platelet aggregation, and promote uteroplacental blood flow. Thromboxane, produced
predominantly by platelets, is a powerful vasoconstrictor, stimulator of platelet aggregation, and stimulator of
uterine contractility, favoring maternal hypertension, decreased uteroplacental blood flow, and intrauterine
growth restriction.
Abnormal prostaglandin homeostasis has been found in preeclampsia (44). Excessive placental production of
thromboxane and insufficient production of PGI2 result in an abnormally high ratio of thromboxane to PGI2.
Prosta-cyclin production is decreased in umbilical arteries, placental veins, and uterine vessels. Thromboxane
production is increased in placental tissue and in circulating platelets from preeclamptic women with infants who
are small for gestational age (45).
Compared with normal placentas, placental findings in women with chronic or essential hypertension vary from
lower total volume, lower parenchymal tissue, and infarcts to normal volumes and large, villous surface areas.
Findings differ with various degrees of severity of disease and lack of differentiation between chronic
hypertension and chronic hypertension with superimposed preeclampsia.
Diabetes
Just as the infant of a diabetic mother can be macrosomic, growth restricted, or normally grown, the diabetic
placenta can have various findings. Some investigators associate these findings with severity of maternal
diabetes, especially the duration and complications of the disease, and others associate them with the degree of
glycemic control.
Placentas of diabetic mothers without significant vascular disease (i.e., White's classes A-D) differ from normal
by having more parenchymal and villous tissue, a higher cellular content, and a larger surface area of exchange
between mother and fetus in terms of peripheral villous and capillary surface areas and intervillous space
volume (46,47,48). These larger placentas are able to adequately support growth of large fetuses. Placentas of
diabetic mothers with appropriate-for-gestational-age newborns are morphologically closer to control,
nondiabetic placentas. It is the placentas of macrosomic infants that are heavier, predominantly because of a
significant accumulation of nonparenchymal and parenchymal tissue. These placentas have retarded maturation
of surface areas of terminal villi. Grossly, they appear large, thick, and plethoric. Microscopically, focal
immaturity (i.e., dysmaturity) and villous edema are found.
The excessive growth and dysmaturity of these placentas suggest an accelerated growth process or a loss of the
normal growth process that occurs in placentas of healthy women. Because the placenta is essentially a fetal
organ, which is different from other fetal organs only in that it is subject to more direct maternal modulation, it
is not surprising that macrosomic fetuses have large placentas. The major mechanism for organ enlargement, or
macrosomia, in the fetus of a diabetic mother involves anabolic metabolism of excess glucose and its deposition
as glycogen and fat. The placenta of a macrosomic infant of a diabetic mother has neither excessive fat nor
glycogen, suggesting a different mechanism by which the fetus (or mother) increases placental size and surface
areas of terminal villi to maintain fetal nutrition (the size of the placenta and quality and topography of the
transfer surface regulate nutrient availability to the fetus).
Placental cells respond to maternally or fetally produced hormones directly by alterations in growth and
indirectly by elaboration of certain substances that control their own growth. Insulin and its associated family of
growth-promoting peptide hormones, such as insulin-like growth factor I (IGF-I) and insulin-like growth factor II
(IGF-II), have been implicated in excessive placental growth of the diabetic placenta.
Insulin receptors have been localized to the apical brush border of the syncytiotrophoblast (bathed in maternal
blood) (49). Maternal insulin binds to these receptors, is internalized, and eventually degraded by this cell.
Investigators using the in vitro perfused human placental cotyledon model have shown that placental-facilitated
uptake and metabolism of glucose do not appear to be regulated by insulin (50). The lack of insulin regulation of
glucose transport in the intact placenta correlates with glucose transporter genes in placenta, GLUT1 and GLUT3,
both thought to code for insulin-unresponsive glucose transporter proteins. Why then is there active insulin
uptake by the placental syncytiotrophoblast? Insulin, more likely fetal but possibly maternal, may have growthpromoting activity in the placenta and thereby affect placental size.
The placental trophoblast microvillous brush border membrane contains specific heterotetrameric IGF-I
receptors that have been found in trophoblasts as early as 6 weeks of gestation (51). This somatomedin has
been measured in placental explant cultures and placental fibroblast culture fluid and is thought to be involved in
control of placental growth (52). Maternal serum IGF-I levels and the ratio of cord serum IGF-I to its binding
protein correlate with birth-weight (53). Cord serum IGF-II levels are 50%
P.160
higher in infants of diabetic mothers than in those of nondiabetic mothers. Because fetal macrosomia in diabetic
pregnancies cannot always be prevented, despite excellent maternal blood glucose control, attention has been
turned to excessive placental transfer of fetal fuels other than glucose. Some investigators suggest a greater
diffusional transfer of free fatty acids in diabetic pregnancies, probably related to greater maternal availability (e.
g., maternal hyperlipidemia in diabetic pregnancies) and greater placental transfer surface (54). Excessive
placental transfer of insulin-secretagogue amino acids (e.g., arginine) also may be responsible for fetal
hyperinsulinemia and macrosomia.
Placentas of diabetic women with vascular complications (e.g., nephropathy, retinopathy, heart disease, White's
class F, R, or H) frequently have infarcts and are associated with growth-retarded fetuses. The weights of the
placentas of these diabetic women are lower than normal gestational age-matched placentas. There have been
no specific placental findings to explain the higher stillborn rate in macrosomic fetuses of diabetic mothers.
Erythroblastosis
Erythroblastosis fetalis, or hemolytic disease of the newborn, is a condition in which specific IgG antibodies
formed by the mother against erythrocyte antigens of the fetus cross the placenta by receptor-mediated
endocytosis and coat fetal erythrocytes, causing splenic sequestration, intravascular hemolysis, anemia, and
unconjugated hyperbilirubinemia. Unconjugated bilirubin is transported easily to the maternal side, conjugated,
and excreted by the mother. Anemia stimulates fetal hematopoiesis, especially in the liver and spleen, resulting
in release of immature erythrocyte precursors into fetal blood. Severe fetal anemia causes fetal and placental
hydrops, and often hypoproteinemia and thrombocytopenia. The placenta of newborns with erythroblastosis
fetalis is pale and enlarged, displaying villous immaturity, edema, and an increase in Hofbauer cells (i.e.,
macrophages). Erythrocyte precursors are found in the vascular spaces. The severity of placental changes
parallels the severity of fetal disease. There is ultrasonographic evidence of reversal of placental thickening and
edema as fetal hydrops improves with treatment.
Placental changes are secondary to the disease process and do not contribute to its formation. Hydropic
placentas from fetuses with erythroblastosis fetalis are indistinguishable from those with other causes. In
placentas of newborns with erythroblastosis fetalis, compensatory placental hematopoiesis, specifically in the
villous stroma, is suggested because numerous erythrocyte precursors are found packed in fetal villous sinusoids
mimicking de novo erythrocyte synthesis in the placenta. No specific erythrocyte synthesis occurs in the
placenta.
Hydropic placentas produce elevated titers of hCG. Serum levels of this hormone are significantly higher than
normal in women with hydropic fetuses and placentas (55). It is unclear whether this is because of
overproduction of the hormone or normal production by larger placental cell mass. Other placental hormones,
including hPL, are found in elevated quantities in serum of women with hydropic placentas.
Preeclampsia occurs frequently in mothers with fetal and placental hydrops. Reversal of preeclamptic signs and
symptoms has been observed after fetal and placental hydrops resolved spontaneously or was reversed by fetal
transfusion (56). Because hydropic placentas release greater amounts of placental hormones into the maternal
circulation, this finding gives credence to a humoral placental product theory as the cause of preeclampsia.
Twin-to-Twin Transfusion Syndrome

Abnormal placental physiology because of congenital placental vascular malformations in monozygotic twins is
the basis of twin-to-twin transfusion syndrome.
Although dizygotic twinning always involves separate placentas and membranes (i.e., diamnionic dichorionic),
monozygotic twinning results in separation ranging from complete separation of the placentas and membranes (i.
e., diamnionic dichorionic) to shared placentas and chorion (i.e., diamnionic monochorionic), to shared
placentas, chorion, and amniotic cavity (i.e., monoamnionic monochorionic), to shared fetal tissues (i.e.,
monoamnionic conjoined).
Twin-to-twin transfusion syndrome is a disease of monochorionic twins. Extraembryonic somatic mesoderm
lining the extraembryonic cavity of the developing embryo gives rise to the chorion and to the mesenchymal
core of the placental villus and associated fetal vessels. Monozygotic twins who share chorions (i.e.,
monochorionic) also share fetal circulations by vascular connections. Fetal vascular anastomoses almost never
are found on placentas of dichorionic twins, but they are usually found on those of monochorionic twins (57).
In monochorionic twin placentas, the normal 1:1 ratio between the artery supplying and the vein draining a
cotyledon is lost. Haphazard vascular patterns exist (Fig. 11-10).
P.161
Certain cotyledons are supplied from a single or anastomotic artery from both twins and are drained by one or
two veins. Others are supplied by an artery of one fetus and drained by a vein of the other. Drainage by the vein
of the opposite fetus may not be obvious when capillaries join the venous system in a neighboring cotyledon
before surfacing as a contralateral fetal vessel. An intertwin arteriovenous shunt should be suspected when an
artery diving into a cotyledon is not accompanied by an adjacent emerging vein.
Figure 11-10 Diamnionic, monochorionic twin placenta, with the intervening amnion rolled in the center.
Notice the haphazard vascularization (large arrow). The velamentous cord insertion (small arrow) of the smaller
twin of this discordant pair is not an unusual finding in monochorionic twins.
The most common fetal vascular anastomosis on the surface of the placenta is artery to artery, occurring in two
thirds of monochorionic placentas (58). Vein-to-vein anastomoses are the least common, occurring in only 1 of
20 monochorionic placentas. Arteriovenous anastomosis, or arterial supply with contralateral fetal venous
drainage of a cotyledon, which occurs in approximately two thirds of monochorionic placentas, is thought to be
the most common placental lesion leading to inequality of blood flow and twin transfusion syndrome. Placental
injection studies using contrast have been performed to ascertain vascular anastomosis, but injection studies
only demonstrate the presence of anastomosis and do not establish overall inequality of flow.
In twin-to-twin transfusion syndrome, blood from the donor twin flows to the recipient twin through intertwin
placental vascular connections. Remaining anastomoses are insufficient to allow return of the lost blood volume,
and an imbalance of blood flow exists. The donor twin becomes hypovolemic, anemic, malnourished, and growth
restricted; and responds with oliguria (i.e., oligohydramnios) and, in severe cases, anuria (i.e., ahydramnios).
The recipient twin becomes hypervolemic and plethoric, develops cardiomegaly and polyuria (i.e.,
polyhydramnios), and, if severe, develops cardiac failure and hydrops fetalis. Grossly, the placental portion of
the donor twin is anemic, pale, and usually smaller than that of the receiving twin. If hydrops has not yet
occurred, the recipient twin's placental portion is red, thick, and congested. After hydrops ensues, the placenta
becomes pale from villous edema. Microscopically, anemia and villous immaturity are found in the donor's
placental portion and polycythemia and congestion in that of the recipient.
Despite the high frequency of cross-placental vascular anastomoses in monochorionic placentas, the incidence of
clinically evident twin-to-twin transfusion syndrome in monochorionic twin pairs is only 5% to 10%. In some
twin pairs, intertwin vascular connections may produce a chronic twin transfusion syndrome in which the larger
twin responds with cardiac hyperplasia and hypertrophy (59). Cardiac hyperplasia may be compensatory, and
certain sets of monochorionic twins may withstand the vascular inequalities throughout gestation (i.e.,
subclinical twin-to-twin transfusion syn drome). In others, the pumping capabilities of the enlarged heart are
exceeded, and cardiac failure occurs. It is tempting to speculate that as the heart enlarges and fails, delicate
pressure-flow characteristics in the placental vasculature are disrupted, exacerbating shunting to the recipient
twin.
Twin-to-twin transfusion syndrome is a condition that begins early in embryonic vascular development.
Benirschke (60) described the youngest example, a pair of aborted twin embryos measuring 7 and 8 cm (10
weeks of gestation) in which the heart of the donor twin was half the size of that of the recipient. Intertwin
vascular distribution established early in the embryonic period may control intertwin placental mass distribution.
Inequality in placental mass distribution may have an early, direct effect on fetal growth, causing significant
growth restriction in one fetus. Antenatal ultrasonographic evaluation of placentation in the first trimester, or of
the number of layers in intervening membranes in the latter trimesters, has allowed diagnosis of chorionicity
with high sensitivity and positive predictive value (61). The obstetrician can confirm chorionicity and assess
placental vascular patterns immediately after delivery of the twin placenta and should impart this information to
the pediatrician.
Some antenatal treatment approaches for twin transfusion syndrome attempt to reverse the abnormal placental
physiology. Selective termination of one fetus, usually the donor, prevents further transfusion to the recipient.
This treatment introduces risks to the living twin from passage of emboli from the dead twin through
anastomotic channels (i.e., twin embolization syndrome). Serial amniocentesis to decompress the
polyhydramniotic sac of the recipient twin has reversed twin-to-twin transfusion syndrome in a few cases (62).
Loss of amniotic fluid pressure on a large placental vascular anastomosis allows changes in fetal blood flow.
After amniotic fluid decompression of severe polyhydramnios, ultrasonographically observed placentas appear
thickened and less stretched. Fetoscopic laser occlusion of placental vessels, reported by De Lia and colleagues
(63), may prove to be the most logical therapy, because treatment is directed at the cause of the problem. At
the time of delivery, the placenta is two discs, separated by an area of infarcted cotyledons previously supplied
by the coagulated vessels.
Acute transplacental fetus-to-fetus bleeding can occur in monochorionic twins and is distinct from twin
transfusion syndrome. Acute fetus-to-fetus bleeding occurs in placentas with medium- or large-caliber vascular
anastomosis when loss of established pressure-flow relations occurs. For example, the death of one twin of a
monochorionic pair may allow large shifts of blood from the living twin to the deceased twin. Shifts may be
sufficient to cause anemia and hydrops fetalis in the surviving twin. If the deceased twin was the growthretarded donor of a twin-to-twin transfusion pair, paradoxical plethora of the donor twin may occur.
Acute transplacental fetal bleeding may occur at the time of labor. For example, during the uterine contractions,
umbilical cord compression may occur to a sufficient degree that it diminishes umbilical venous return but not
arterial perfusion. The resulting higher resistance in the placental venous system of the cord-compressed twin
favors return of blood to the co-twin. Acute intraplacental fetal bleeding may occur after delivery of the first
fetus. Loss of established placental pressure-flow relations after
P.162
clamping of the umbilical cord favors intraplacental pooling of the second twin's blood through anastomoses,
resulting in hypovolemia. Acute transplacental fetal bleeding can create disparate newborn hematocrits that do
not reflect hematocrit levels during fetal life.
THE IN UTERO ENVIRONMENT AND FETAL DETERMINANTS OF

ADULT DISEASE
The in utero environment has an overwhelming impact upon birth-weight and newborn disorders. Constriction of
the uterine cavity because of leiomyomata (fibroid tumors), uterine anomalies, or lack of amniotic fluid may lead
to deformation abnormalities such as club feet (64). Persistent, severe oligohydramnios, particularly in the
second and third trimesters may lead to Potter sequence (65). Certain metabolic abnormalities of the mother
causing an abnormal metabolic uterine milieu, such as uncontrolled diabetes (66) or uncontrolled
phenylketonuria (67), may lead to congenital malformations, mental retardation, and growth abnormalities.
Additionally, each fetus has a certain genetic predisposition to various disorders the development of which can
be influenced by the in utero environment. For example, in utero positioning or abnormalities of labor may be
associated with cranial synostosis in predisposed individuals (68).
Recently, the concept of the in utero environment influencing development of adult disease has become popular.
This fetal origins of adult disease hypothesis says that human development is controlled not only by a
combination of predetermined genes, but the effect on the fetus of the many variables of pregnancy.
The epidemiologic associations to support in utero programming and the fetal origins hypothesis were initially
published in 1989 (69). The majority of the abnormal fetal origin hypotheses relate to poor fetal growth.
Generally, investigators in these initial studies noted that outliers of normal birth-weight, either very large or
more often very small newborns, had an increased risk for coronary artery disease, hypertension, and type 2
diabetes. This concept in humans has credibility from animal studies (70,71). The fetal origins hypothesis
suggests, specifically, that lack of availability or use of fetal fuels in utero can program an individual's risk for
chronic disease later in life.
Intrauterine fetal growth restriction is not only heterogeneous in etiology, but often multifactorial. The extensive
placental and maternal physiologies discussed earlier in this chapter apply. Overall, the availability, delivery, and
utilization of nutrients influence the growth of the fetus. Some of these influences can be modified such as
promotion of adequate nutrition, cessation of smoking, and optimization of cardiac output to the uterus and
placenta. Many other etiologies of growth restriction in newborns are evasive and lack both therapeutic and
preventative prenatal measures.
BIRTH-WEIGHT AND CORONARY HEART DISEASE

Epidemiologic investigators have found an association of either low- or high-birth-weight with a higher risk of
coronary heart disease later in life (72,73). The majority of the supporting data come from British
epidemiologists evaluating birth records in three regions of Britain, specifically Hertfordshire, Sheffield, and
Preston.
In a study of men born in Sheffield from1907 to 1924 (74), death because of cardiovascular disease evaluated
by standardized mortality ratios, fell in the birth-weight categories greater than 2,495 g. The standardized
mortality ratio for cardiovascular disease fell from 119 in men who weighed 2,495 g (5.5 pounds) or less at birth
to 74 in men who weighed more than 3,856 g (8.5 pounds). A significant fall was seen also in premature
cardiovascular deaths and was associated as well with smaller head circumference and lower ponderal index.
Notably in the study, men of low-birth-weight with large placentas (highest placenta-to-birth-weight ratio) had
the highest rate of mortality from cardiovascular disease. Additionally, the risk appeared to be associated more
with those who were low-birth-weight newborns who experienced catch-up growth later in life (75).
The association of low-birth-weight and higher incidence of cardiovascular disease has also been seen in other
areas of the world. In an Uppsala study of nearly 15,000 Swedish men born between 1915 and 1929, a 1,000-g
increase in birth-weight correlated with a 0.77 reduction in the rate of coronary artery disease (76).
The association of low-birth-weight and mortality from coronary artery disease appears to be stronger for men
than women, but still significant in the latter. In Nurse's Health Study, performed in the United States, 70,297
women who completed questionnaires from 1976 to 1992 were assessed. An inverse relationship between birthweight and cardiovascular disease and stroke was noted in these women (73).
Risk factors for coronary artery disease, specifically abnormalities of cholesterol metabolism and the coagulation
system, were also linked to low-birth-weight, especially head-sparing growth restriction. In an English cohort
study from Sheffield, a reduction in abdominal circumference and body length at birth predicted higher serum
low-density lipoprotein cholesterol and plasma fibrinogen levels in adult life. It was postulated that the small
abdominal circumference reflected impaired liver growth and reprogramming of liver metabolism (77). In a
somewhat similar British study of newborn size and corresponding adult lipid profile, men and women who had
smaller abdominal circumferences at birth were found to have higher serum levels of total and low density
lipoprotein cholesterol and apolipoprotein B. The association was independent of social class, adult body weight,
cigarette use, or alcohol use (78).
Higher birth-weights have also been associated with a higher risk of adult onset disease, but the association has
P.163
not been as strong. Men born with large abdominal circumferences (presumed large liver as well) were also
noted to have an increased risk for coronary heart disease (79).
BIRTH-WEIGHT AND HYPERTENSION

Birth-weight was associated with adult hypertension at least as early as 1985 by Wadsworth and investigators
(80), who studied more than 5,000 36-year-old men and noted that, in addition to adult weight, newborn weight
was also statistically related to blood pressure elevation. Law and colleagues (81) studied a national sample of
British children and men and women in various age categories. At all age categories beyond the newborn period,
men and women who had lower birth-weights developed a higher blood pressure in adulthood. Investigators in
the Nurses' Health Study evaluated women at 30 to 55 years of age and found that their birth-weight (by
recollection) was inversely proportional to blood pressure in adult life (82).
Barker and associates (83) studied adult blood pressure of 327 men and women 46 to 54 years of age in
relation to the placental weight and the newborn parameters of length, ponderal index, and head circumference.
Even in this small cohort, there was a strong trend of higher blood pressure in adult life with lower birth-weight.
Those newborns with high placental weights were also at higher risk for development of hypertension later in life.
The association of low-birth-weight with a higher risk of hypertension in adulthood has not been found in all
studies. Birth-weight has been found in at least one study (84) to have a nearly negligible influence upon adult
hypertension when compared to the effect of adult body mass index. In that study, birth-weight accounted for
only 5% of the risk of adult hypertension when compared to adult body mass index, which accounted for 12% of
the risk.
BIRTH-WEIGHT AND DIABETES

Hales and coworkers (85) evaluated fasting plasma glucose, insulin, proinsulin, and 32/33 split proinsulin
concentrations, as well as glucose levels after a 75 g glucose challenge, in 468 men in a British cohort born and
raised in Hertfordshire. Men of lower birth-weight had the highest rate of glucose intolerance and diabetes. The
association was independent of adult body mass at the time of the study. A higher level of 32/33 split proinsulin
was also related to low-birth-weight and interpreted as representation of pancreatic -cell dysfunction. In this
cohort, hypertension in adulthood was also associated with impaired glucose tolerance in adulthood. Low-birthweight has also been associated with a higher risk of syndrome X (type 2 diabetes, hypertension, and
hyperlipidemia) in adult life (86).
In a British study of forty 21-year-old men, glucose intolerance after a 75 g glucose challenge was associated
with lower birth-weight independent of gestational age at birth and adult body mass, height, and social class
(87). In a different study, ponderal index at birth was indirectly correlated with insulin resistance in late
adulthood as measured by the rate of drop of glucose concentration after injection of insulin (88). Thinner
newborns had greater glucose intolerance in adulthood. The authors remarked that thinness at birth and
thinness in adult life had opposing effects. Insulin resistance and glucose intolerance increased with a low-birthweight, but decreased with high adult weight.
Huxley and colleagues (89), as well as others (90), question the fetal origins of adult disease hypothesis. They
conducted a meta-analysis of 55 studies and concluded that the association of at least low fetal weight and an
increased risk of high blood pressure in adulthood is at best weak. They believe that the strength of this
association was primarily because of random error, selective emphasis of particular results, and inappropriate
adjustments for adult weight and confounding factors.
Current epidemiologic data regarding fetal origins, fetal programming, and intrauterine effects on development
of adult disease need further evaluation to confirm both association and causation. Both genetic and
environmental factors affect fetal growth. For example, women with hypertension (and at risk for coronary
artery disease later in life) are more likely to have growth-restricted newborns because of their hypertension
during the pregnancy. These growth-restricted newborns are genetically at higher risk of developing
hypertension and thus coronary heart disease irrespective of intrauterine conditions. Confounding variables need
further evaluation. It remains to be seen if a newborn's small size alone is a risk factor for adult-onset disease.
Nonetheless, it is exciting to think that efforts to promote a healthy pregnancy and optimize normal fetal growth
may decrease development of certain common and widespread adult disorders.
REFERENCES
1. Ramsey EM. The placenta: human and animal. New York: Praeger, 1982:9.
2. Pridjian G, Moawad AH, Whitington PF. Handling of beta-hydroxybutyrate in the human placenta. In:
Scientific program and abstracts. St Louis, MO: Society for Gynecologic Investigation, 1990:230(abst).
3. Lind T, Aspillaga M. Metabolic changes during normal and diabetic pregnancies. In: Reece EA, Coustan DR,
eds. Diabetes mellitus in pregnancy: principles and practice. New York: Churchill Livingstone, 1988:75.
4. Neufeld ND, Corbo L. Increased fetal insulin receptors and changes in membrane fluidity and lipid
composition. Am J Physiol 1982;243:E246.
5. Dancis J. Aspects of bilirubin metabolism before and after birth. Pediatrics 1959;24:980.
6. Kayano T, Fukumoto H, Eddy RL, et al. Evidence for a family of human glucose transporter-like proteins. J
Biol Chem 1988;263: 15245.
7. Tadakoro C, Yoshimoto Y, Sakata M, et al. Localization of human placental glucose transporter 1 during
pregnancy. An immunohistochemical study. Histol Histopathol 1996;11:673.
8. Johnson LW, Smith CH. Monosaccharide transport across microvillous membrane of human placenta. Am J
Physiol 1980; 238:C160.
9. Yudelivech DL, Sweiry JH. Transport of amino acids in the placenta. Biochim Biophys Acta 1985;822:169.
P.164
10. Miller RK, Berndt WO. Characterization of neutral amino acid accumulation by human term placental
slices. Am J Physiol 1974; 227:1236.
11. Schneider H, Mohlen KH, Dancis J. Transfer of amino acids across the in vitro perfused human placenta.
Pediatr Res 1979;13:236.
12. Enders RH, Judd RM, Donohue TM, et al. Placental amino acid uptake. III. Transport systems for neutral
amino acids. Am J Physiol 1976;230:706.
13. Hibbard JU, Pridjian G, Whitington PF, et al. Taurine transport in the in vitro perfused human placenta.
14. Fortunato SJ, Bawdon RE, Swan KF, et al. Transfer of azidothymidine (AZT) across the in vitro perfused
human placenta. In: Scientific program and abstracts. San Diego, CA: Society for Gynecologic Investigation,
1989:82(abst).
15. Connor EM, Sperlin RS, Gelber R, et al. Reduction of maternal- infant transmission of human
immunodeficiency virus type 1 with zidovudine treatment. N Engl J Med 1994;331:1173.
16. Ockleford CD, Whyte A. Differentiated regions of human placental cell surface associated with the
exchange of materials between maternal and fetal blood. The structure, distribution, ultrastructural
cytochemistry and biochemical composition of coated vesicles. J Cell Sci 1977;25:293.
17. McNabb T, Koh TY, Dorrington KJ, et al. Structure and function of immunoglobulin domains V. Binding of
immunoglobulin G and fragments to placental membrane preparations. J Immunol 1976; 117:182.
18. Niezgodka M, Mikulska J, Ugorski M, et al. Human placental membrane receptor for IgG-1. Studies on the
properties and solubilization of the receptor. Mol Immunol 1981;18:163.
19. Bussel JB, Zabusky MR, Berkowitz RL, et al. Fetal alloimmune thrombocytopenia. N Engl J Med
1997;337:22.
20. Okuyama T, Tawada MD, Furuya H, et al. The role of transferrin and ferritin in the fetal-maternal-placental
unit. Am J Obstet Gynecol 1985;152:344.
21. Ringler GE, Kallen CB, Strauss JF. Regulation of human trophoblast function by glucocorticoids:
dexamethasone promotes increased secretion of chorionic gonadotropin. Endocrinology 1989;124:1625.
22. Eldar-Geva T, Hochberg A, deGroot N, et al. High maternal serum chorionic gonadotropin level in Downs'
syndrome pregnancies is caused by elevation of both subunits messenger ribonucleic acid level in
trophoblasts. J Clin Endocrinol Metab 1995;80:3528.
23. Cohen M, Haour F, Dumont M, et al. Prognostic value of human chorionic somatomammotropin plasma
levels in diabetic patients. Am J Obstet Gynecol 1973;115:202.
24. Adams FH, Fujiwara T. Surfactant in fetal lab tracheal fluid. J Pediatr 1963;63:537.
25. Goodlin RC, Anderson JC, Gallagher TF. Relationship between amniotic fluid volume and maternal plasma
volume expansion. Am J Obstet Gynecol 1983;146:505.
26. McCoshen JA. Associations between prolactin, prostaglandin E2 and fetal membrane function in human
gestation. In: Mitchell BF, ed. The physiology and biochemistry of human fetal membranes. Ithaca, NY:
Perinatology Press, 1988:117.
27. Okazaki T, Casey ML, Okita JR, et al. Initiation of human parturition, XII. Biosynthesis and metabolism of
prostaglandins in human fetal membranes and uterine decidua. Am J Obstet Gynecol 1981;139:373.
28. Romero R, Brody DT, Oyarzun E, et al. Infection and labor III. Interleukin-1: a signal for the onset of
parturition. Am J Obstet Gynecol 1989;160:1117.
29. Keelan JA, Sato T, Mitchell MD. Interleukin (IL)-6 and IL-8 by human amnion: regulation by cytokines,
growth factors, glucocorticoids, phorbol esters, and bacterial lipopolysaccharide. Biol Reprod 1997;57:1438.
30. Romero R, Avila C, Santhanam U, et al. Amniotic fluid interleukin 6 in preterm labor: association with
infection. J Clin Invest 1990;85:1392.
31. Goldberg RL, Hauth JC, Andrews WW. Intrauterine infection and preterm delivery. N Engl J Med
2000;342:1500.
32. Benirschke K, Kaufmann P. Pathology of the human placenta, 2nd ed. New York: Springer-Verlag,
1990:182.
33. Silver RK, Dooley SL, Tamura RK, et al. Umbilical cord size and amniotic fluid volume in prolonged
pregnancy. Am J Obstet Gynecol 1987;157:716.
34. Robertson RD, Rubinstein LM, Wolfson WL, et al. Constriction of the umbilical cord as a cause of fetal
demise following midtrimester amniocentesis. J Reprod Med 1981;26:325.
35. Pridjian G, Puschett JB. Preeclampsia part I: clinical and pathophysiologic considerations. Obstet Gynecol
Surv 2002;57:598.
36. Pridjian G, Puschett JB. Preeclampsia part II: experimental and genetic considerations. Obstet Gynecol
Surv 2002;57:619.
37. Boyd PA, Scott A. Quantitative structural studies on human pla centas associated with preeclampsia,
essential hypertension and intrauterine growth retardation. Br J Obstet Gynaecol 1985; 92(7):714721.
38. Cibils LA. The placenta and newborn infant in hypertensive conditions. Am J Obstet Gynecol 1974;118:256.
39. Khong TY, De Wolf F, Robertson, WB, et al. Inadequate maternal vascular response to placentation in
pregnancies complicated by preeclampsia and by small for gestational age infants. Br J Obstet Gynaecol
1986;93:1049.
40. Damsky CH, Fitzgerald ML, Fisher SJ. Distribution patterns of extracellular matrix components and
adhesion receptors are intricately modulated during first trimester cytotrophoblast differentiation along the
invasive pathway, in vivo. J Clin Invest 1992;89: 210.
41. Zhou Y, Damsky CH, Fisher SJ. Preeclampsia is associated with failure of human cytotrophoblasts to mimic
a vascular adhesion phenotype. Once cause of defective endovascular invasion in this syndrome? J Clin Invest
1997;99:2152.
42. Zeek PM, Assali NS. Vascular changes in the decidua associated with eclamptogenic toxemia of pregnancy.
Am J Clin Pathol 1950;20:1099.
43. Out HJ, Kooijman CD, Bruinse HW, et al. Histopathological findings in placentae from patients with
intrauterine fetal death and antiphospholipid antibodies. Eur J Obstet Gynecol Reprod Biol 1991;41:179.
44. Walsh SW. Preeclampsia: an imbalance in placental prostacyclin and thromboxane production. Am J
45. Wallenburg HC, Rotmans N. Enhanced reactivity of the platelet thromboxane pathway in normotensive
and hypertensive pregnancies with insufficient fetal growth. Am J Obstet Gynecol 1982;144:523.
46. Teasdale F. Histomorphometry of the placenta of the diabetic woman. Class A diabetes mellitus. Placenta
1981;2:241.
47. Teasdale F. Histomorphometry of the human placenta in class B diabetes mellitus. Placenta 1983;4:1.
48. Teasdale F. Histomorphometry of the human placenta in class C diabetes mellitus. Placenta 1985;6:69.
49. Deal CL, Guyda HJ. Insulin receptors of human term placental cells and choriocarcinoma (JEG-3) cells:
characteristics and regulation. Endocrinology 1983;112:1512.
50. Challier JC, Hauguel S, Desmaizieres V. Effect of insulin on glucose uptake and metabolism in the human
placenta. J Clin Endocrinol Metab 1986;62:803.
51. Grizzard JD, D'Ercole AJ, Wilkins JR, et al. Affinity-labeled somatomedin-C receptors and binding proteins
from the human fetus. J Clin Endocrinol Metab 1984;58:535.
52. Fant M, Monro H, Moses AC. An autocrine/paracrine role for insulin-like growth factors in the regulation of
human placental growth. J Clin Endocrinol Metab 1986;63:499.
53. Hall K, Hansson U, Lundin G, et al. Serum levels of somatomedins and somatomedin-binding protein in
pregnant women with type I or gestational diabetes and their infants. J Clin Endocrinol Metab 1986;63:1300.
54. Thomas CR. Placental transfer of non-esterified fatty acids in normal and diabetic pregnancy. Biol Neonate
1987;51:94.
55. Hatjis CG. Nonimmunologic fetal hydrops associated with hyperreactio luteinalis. Obstet Gynecol 1985;65
[Suppl]:11S-13S.
56. Pryde PG, Nugent CE, Pridjian G, et al. Spontaneous resolution of nonimmune hydrops fetalis secondary to
parvovirus B19 infection. Obstet Gynecol 1992;79:869.
57. Robertson EG, Neer KJ. Placental injection studies in twin gestation. Am J Obstet Gynecol 1983;147:170.
58. Benirschke K, Kaufmann P. Pathology of the human placenta, 2nd ed. New York: Springer-Verlag,
1990:658.
59. Pridjian G, Nugent CE, Barr M. Twin gestation: influence of placentation on fetal growth. Am J Obstet
Gynecol 1991;165:1394.
60. Benirschke K. Prenatal cardiovascular adaptation, comparative pathophysiology of circulatory

disturbances. In: Bloor CM, ed. Advances in experimental medicine and biology. New York: Plenum Press,
1972:3.
61. D'Alton ME, Dudley DK. The ultrasonographic prediction of chorionicity in twin gestation. Am J Obstet
Gynecol 1989;160:557.
P.165
62. Elliott JP, Urig MA, Clewell WH. Aggressive therapeutic amniocentesis for treatment of twin-twin
transfusion syndrome. Obstet Gynecol 1991;77:537.
63. De Lia JE, Cruikshank DP, Keye WR. Fetoscopic neodymium:YAG laser occlusion of placental vessels in
severe twin-twin transfusion syndrome. Obstet Gynecol 1990;75:1046.
64. Christianson C, Huff D, McPherson E. Limb deformations in oligohydramnios sequence: effects of

gestational age and duration of oligohydramnios Am J Medical Genetics. 1999;86:430.
65. Curry CJ, Jensen K, Holland J, et al. The Potter sequence: a clinical analysis of 80 cases. Am J Med Genet
1984;19:679.
66. Simpson JL, Elias S, Martin AO, et al. Diabetes in pregnancy, Northwestern University series (19771981).
I. Prospective study of anomalies in offspring of mothers with diabetes mellitus. Am J Obstet Gynecol
1983;146:263.
67. Rouse B, Azen C, Koch R, et al. Maternal Phenylketonuria Collaborative Study (MPKUCS) offspring: facial
anomalies, malformations, and early neurological sequelae. Am J Med Genet 1997;69:89.
68. Shahinian HK, Jackle R, Suh Rh, et al. Obstetrical factors governing the etiopathogenesis of lambdoid
synostosis. Am J Perinatol 1998;15:281.
69. Barker DH, Winter PD, Osmond C, et al. Weight in infancy and death from ischaemic heart disease. Lancet
1989;2:577.
70. Koukkou E, Ghosh P, Lowy C, et al. Offspring of normal and diabetic rats fed saturated fat in pregnancy
demonstrate vascular dysfunction. Circulation 1998;98:2899.
71. Vickers MH, Ikenasio BA, Breier BH. Adult growth hormone treatment reduces hypertension and obesity
induced by an adverse prenatal environment. J of Endocrinology 2002;175:615.
72. Barker DJP, Osmond C. Infant mortality, childhood nutrition, and ischaemic heart disease in England and
Wales. Lancet 1986;1:1077.
73. Rich-Edwards J, Stampfer M, Manson J, et al. Birth weight and risk of cardiovascular disease in a cohort of
women followed up since 1976. BMS 1997;315:396400.
74. Barker DJ. Osmond C, Simmonds SJ, et al. The relation of small head circumference and thinness at birth
to death from cardiovascular disease in adult life. BMJ 1993;306:422.
75. Barker DJ, Eriksson JG, Forsen T, et al. Fetal origins of adult disease: strength of effects and biological
basis. Int J Epidemiol 2002; 31:1235.
76. Leon DA, Lithell HO, Vg D, et al. Reduced fetal growth rate and increased risk of death from ischaemic
heart disease: cohort study of 15, 000 Swedish men and women born 191529. BMJ 1998; 317:241.
77. Roseboom TJ, van der Meulen JH, Ravelli AC, et al. Plasma fibrinogen and factor VII concentrations in
adults after prenatal exposure to famine. Br J Haematol 2000;111:112117.
78. Barker DJ, Martyn CN, Hales, et al. Growth in utero and serum cholesterol concentrations in adult life. BMJ
1993;307:1524.
79. Barker DJP, Martyn CN, Osmond C, et al. Abnormal liver growth in utero and death from coronary heart
disease. BMJ 1995; 310:703.
80. Wadsworth ME, Cripps HA, Midwinter RE, et al. Blood pressure in a national birth cohort at the age of 36
related to social and familial factors, smoking and body mass. BMJ 1985;291:1534.
81. Law CM, de Swiet M, Osmond C, et al. Initiation of hypertension in utero and its amplification throughout
life. BMJ 1993; 306:24.
82. Curham GC, Chertow GM, Willett WC, et al. Birth weight and adult hypertension and obesity in women.
Circulation 1996;94: 1310.
83. Barker DJ, Godfrey KM, Osmond, et al. The relation of fetal length, ponderal index and head circumference
to blood pressure and the risk of hypertension in adult life. Paediatr Perinat Epidemiol 1992;6:35.
84. Holland FJ, Stark O, Ades AE, et al. Birth weight and body mass index in childhood, adolescence, and
adulthood as predictors of blood pressure at age 36. J Epidemiol Commun H 1993;47(6): 432435.
85. Hales CN, Barker DJ, Clark PM, et al. Fetal and infant growth and impaired glucose tolerance at age 64.
BMJ 1991;303:1474.
86. Barker DJ, Hales CN, Fall CH, et al. Type 2 (non-insulin-dependent) diabetes mellitus, hypertension and
hyperlipidemia (syndrome X): relation to reduced fetal growth. 1993;36:62.
87. Robinson S, Walton RJ, Clark PM, et al. The relation of fetal growth to plasma glucose in young men.
Diabetologia 1992; 35:444.
88. Phillips DI, Barker DJ, Hales CN, et al. Thinness at birth and insulin resistance in adult life. Diabetologia
1994;37:150.
89. Huxley R, Neil A, Collins R. Unraveling the fetal origins hypothesis: is there really an inverse association
between birthweight and subsequent blood pressure? Lancet 2002;360:659665.
90. Williams S, Poulton R. Birth size, growth, and blood pressure between the ages of 7 and 26 years: failure
to support the fetal origins hypothesis. Am J Epidemiol 2002;155:849.
Chapter 12
Fetal Imaging: Ultrasound and Magnetic Resonance
Imaging
Dorothy I. Bulas
Advances in high-resolution sonographic imaging and rapid sequencing magnetic resonance imaging (MRI) have
provided exquisite detail regarding the fetus and intrauterine environment. The ability to assess the health of
the fetus and identify anomalies has changed the practice of both obstetrics and neonatology. By identifying the
fetus with anomalies or at risk for intrauterine compromise, management can be guided by appropriate
specialists with resultant improved outcome.
ASSESSMENT OF FETAL AGE, GROWTH, AND MATURITY

Accurate assessment of fetal age is crucial for perinatal management. Overestimating the age of a fetus can
result in the delivery of a premature infant, whereas underestimating the age of a growth-retarded fetus can
result in a fetal death.
Fetal age can be determined based on obstetric history and clinical data, but estimations can be inaccurate in up
to 40% of patients (1). The use of ultrasound provides a method of significantly improving the accuracy of
gestational age determination. Measurements used to assess age have been chosen on the basis of their
relationship with gestational age and ease in obtaining these measurements and reproducibility. The fetal
biparietal diameter (BPD) measurement was the first measurement used because the cranium was the most
easily visualized body part (2). There are many tables and nomograms that describe the normal growth of
various fetal organs. Most predictions of gestational age are based on the 50th percentile measurement with a
wide range of normal values. At the various gestational ages, each fetal dimension differs in how easily and
reliably it can be measured. Thus, a combination of measurements is superior to a single measurement, with
computer software that permits instantaneous calculation of fetal age and weight estimates and plotting growth
curves. Functional studies of the fetus, such as breathing patterns, can also aid in the determination of
gestational age.
DETERMINING FETAL AGE

First-Trimester Assessment
In the first trimester, fetal imaging using a transabdominal approach is often difficult because the uterus is low
within the maternal pelvis and may be obscured by overlying gas. Transvaginal ultrasound transducers greatly
improve fetal visualization in early pregnancy. The probe is placed into the maternal vagina in close proximity to
the uterus and fetus. High-resolution fetal images can thus be obtained for accurate assessment of fetal
morphology. By use of transvaginal scanning, a gestational sac can be seen by 5 weeks menstrual age, and fetal
heart activity by 6 weeks menstrual age. With abdominal scanning, these findings are first identified reliably a
week later, at 6 and 7 weeks, respectively.
Crown-rump length is the most accurate dimension for assessing gestational age in the first trimester, with a
95% confidence interval of 6 2 to 3.5 days (3). Measurement of the gestational sac diameter can also be used to
estimate age, although it is not as accurate as a result of variability in the shape of the sac. Fetal biparietal
diameter and abdominal circumference are also relatively good predictors of menstrual age in the first trimester
(4).
TABLE 12-1 RELATIONSHIPS AMONG SELECTED FETAL INDICES AS MEASURED BY ULTRASOUND AND
GESTATIONAL AGE AS DETERMINED BY MENSTRUAL WEEKS
Menstrual Age
Biparietal
Head Circumference
Abdominal
Femur
(wk)
Diameter (cm)
(cm)
Circumference (cm)
Length (cm)
12
13
14
15
16
2
2.3
2.7
3
3.3
7.1
8.4
9.8
11.1
12.4
5.6
6.9
8.1
9.3
10.5
0.8
1.1
1.5
1.8
2.1
17
18
19
20
21
22
23
24
25
3.7
4
4.3
4.6
5
5.3
5.6
5.8
6.1
13.7
15
16.3
17.5
18.7
19.9
21
22.1
23.2
11.7
12.9
14.1
15.2
16.4
17.5
18.6
19.7
20.8
2.4
2.7
3
3.3
3.6
3.9
4.2
4.4
4.7
26
27
28
29
30
31
32
6.4
6.7
7
7.2
7.5
7.7
7.9
24.2
25.2
26.2
27.1
28
28.9
29.7
21.9
22.9
24
25
26
27
28
4.9
5.2
5.4
5.6
5.8
6.1
6.3
33
34
35
36
37
8.2
8.4
8.6
8.8
9
30.4
31.2
31.8
32.5
33.1
29
30
30.9
31.8
32.7
6.5
6.6
6.8
7
7.2
38
9.1
33.6
33.6
39
9.3
34.1
34.5
40
9.5
34.5
35.4
From Hadlock FP, Deter RL, Harrist RB. Computer assisted analysis of fetal age using multiple
7.3
7.5
7.6
fetal
growth parameters. J Clin Ultrasound 1983;11:313, with permission.

P.167
Second- and Third-Trimester Assessment

After 12 weeks of gestation, the crown-rump length is difficult to obtain accurately. In the second and third
trimesters, the main parameters of fetal age are the BPD, head circumference, abdominal circumference, and
femoral length (Table 12-1).
The fetal BPD is quick, reliable, and easy to obtain. It measures the transverse diameter of the upper midbrain
at the level of the thalamic nuclei and cavum septum pellucidum from the external surface of one proximal
parietal bone to the inner surface of the contralateral parietal bone (Fig. 12-1) (5). The earlier the gestational
age the greater the predictive accuracy of this measurement.
Inaccuracies in measurements of the BPD may result from improper selection of the measurement plane (6). In
late pregnancy, the fetal head may be too low in the pelvis to measure. Cranial molding, either from fetal
position or oligohydramnios, may result in BPD measurements that do not accurately reflect true gestational age
(Fig. 12-2).
P.168
The cephalic index, which measures the relationship between the short and long axes of the fetal head, should
be measured if the head shape appears abnormal. If this index is below 70 or above 86, the biparietal diameter
should not be used in the age and weight estimates (7). The accuracy of BPD measurements decreases with
advancing gestational age. In the second trimester, a BPD measurement carries a predictive accuracy of 7
days, whereas in the late third trimester it carries a predictive accuracy of 3 weeks (7). Serial measurements
of BPD can improve the predictive accuracy (8). The fetal head circumference is obtained from the same axial
image used for the BPD. This measurement is more accurate than the BPD in the third trimester, as it is less
affected by shape (7).
Figure 12-1 Fetal biparietal diameter (BPD). Axial scan of the head at the standardized level for BPD
measurement shows the thalami.
Figure 12-2 Transverse image of the brain of a 23-week fetus with severe oligohydramnios demonstrates a
narrow and elongated skull. Cephalic index was abnormally low.
The measurement of the fetal abdominal circumference is made from an axial image of the abdomen at the level
of the portal vein. The circumference of the head is typically larger than the abdominal circumference up to 34
weeks of gestation, and the abdomen becomes larger thereafter. In the presence of intrauterine growth
retardation (IUGR), head size tends to be preserved compared to abdominal size altering the head
circumference to abdominal circumference ratio.
Femur length is defined as the distance between the greater trochanter and the distal end of the femur (Fig. 123) (9). Gestational age prediction from femur length is subject to error either from measurement difficulties
caused by difficulty in visualizing the ends of the bone or from biological variation. The predictive accuracy of
femur length ranges from 1 week at 12 weeks of gestation to 3 weeks at 36 weeks of gestation (10).
Estimation of fetal age is best accomplished using a composite assessment of multiple fetal measurements. A
combination of head size (BPD or head circumference), femur length, and abdominal circumference is most
commonly used. Estimates have an error of 1.46 weeks between 24 and 30 weeks of gestation and increase to
2.3 weeks at 36 weeks of gestation (11). Individual measurements should not be used for assessing gestational
age when they are affected by a pathological process; for example, head measurements in a hydrocephalic fetus
or long bone measurements in the fetus with a bone dysplasia. After 22 weeks, age-independent fetal body
ratios are useful in identifying asymmetric measurements. If the cephalic index is normal, the ratio of femur
length to BPD can be measured, with normal results ranging between 71 and 87. The ratio of femur length to
abdominal circumference should range between 20 and 24 (12).
Figure 12-3 Longitudinal scan of the fetal thigh demonstrates a femur length of 3.5 cm consistent with a 21week gestation. Measurement is from the greater trochanter to the distal end of the femur.
FETAL WEIGHT ESTIMATES

It would be helpful in the assessment of fetal age, growth, and maturity if an accurate, reproducible method of
determining fetal weight were available. Ultrasonographic estimates of fetal weight have been derived from
single measurements such as BPD or abdominal circumference (13) and by a combination of measurements
such as abdominal circumference and BPD (14). All ultrasonographic methods for fetal weight estimations,
however, contain inherent error. Measurement of abdominal circumference yields an average error of 18% (13).
The BPD and abdominal circumference yield a similar error (20%) (15). The absolute estimated error decreases
as birth weight decreases.
The use of multiple parameters, especially head, abdomen, and femur measurements, provides the most
accurate measurement of fetal weight. However, it is still not accurate in extremely low or high birth weight
fetuses. Because fetal weight estimates remain imprecise, estimation of fetal age based solely on estimates of
fetal weight should not be used. Incremental fetal weight change can be determined with serial ultrasounds and
is one way of detecting abnormal fetal growth (10). In the future, three-dimensional fetal MRI reconstruction
and volume estimates may improve the accuracy of fetal weight determination (16).
P.169
ASSESSMENT OF FETAL GROWTH

Fetal growth abnormalities are often associated with increased risk of perinatal morbidity and mortality. The rate
of growth in the normal fetus varies with gestational age, with an increasing exponential curve in the first
trimester, a linear curve in the second and most of the third trimester, and a decreasing exponential curve after
36 weeks of gestation (17). Serial ultrasonographic measurements can be plotted to evaluate fetal growth and
may include weight estimates derived from a combination of these variables.
Intrauterine Growth Retardation

The infant whose birth weight is below the tenth percentile for gestational age is considered small for gestational
age (SGA). The population of SGA infants whose weight falls below the tenth percentile is heterogeneous.
Normal but small fetuses represent up to 60% of the SGA group, fetuses with congenital anomalies account for
up to 15%, and fetuses with true IUGR account for 25% (18). Management varies among these groups. The
normal but small fetus is not at increased risk and does not require intervention. The small fetus with true
growth restriction may be at high morbidity or mortality risk and may require immediate intervention. The small
infant with congenital abnormalities may or may not require intervention. It can be difficult determining which
group an SGA fetus falls into, particularly when the true gestational age is unknown.
Ultrasound's ability to identify the truly IUGR fetus is variable and is most successful when the disease state is
severe. A slowing of the rate of head and femur growth is apparent with severe IUGR (19,20) but is more
difficult to identify when the retardation is mild or early. Growth-adjusted sonographic age determination based
on BPD or abdominal circumference (6,20) uses serial ultrasonographic measurements. This method assumes
that fetal growth should remain within a narrow percentile band and that deviation from this may be an early
sign of growth disorders (21). Because of the inherent errors, serial estimation of fetal weight identifies only the
most severe growth disturbances. Asymmetric growth suggests IUGR. Abnormal ratios of head circumference to
abdominal circumference, as a result of loss of liver mass with normal head growth, can be seen in severe IUGR
but may not be present in milder cases (22).
Secondary signs aid in assessing the presence and severity of IUGR. Oligohydramnios often develops during
episodes of uteroplacental insufficiency, most likely because of decreased fetal urine production and pulmonary
fluid during episodes of fetal hypoxemia (23). Amniotic fluid volume may be assessed subjectively or by the
semiquantitative method of measuring the vertical diameter of the largest visible pocket of amniotic fluid. The
amniotic fluid index (AFI) is calculated by adding the vertical depths of the largest pocket in each of four equal
uterine quadrants (24). Oligohydramnios has been defined by the absence of identifiable amniotic fluid pockets
or when the maximum vertical pocket measures less than 2 cm in two perpendicular planes or when the
amniotic fluid index measures less than 5 cm. Oligohydramnios as defined by any one of the methods is
predictive of increased peripartum morbidity and mortality (24,25,26). Milder cases of IUGR, however, may not
be associated with oligohydramnios (27).
Assessing fetal condition using the biophysical profile score (BPS) or Doppler velocity ratios of the umbilical
artery and/or fetal middle cerebral artery are additional useful adjuncts in the diagnosis of IUGR (see Doppler
Assessment of Blood Velocities in Umbilical and Fetal Vessels) (28). If a fetus who is being followed for IUGR has
no signs of asphyxial compromise, continued observation rather than intervention may be reasonable.
Macrosomia
The infant whose birth weight is above the 90th percentile for gestational age is labeled large for dates (LGA) or
macrosomic. This group of large infants is heterogeneous, composed of normal but large infants and infants with
abnormally increased growth. Macrosomia is usually associated with maternal glucose abnormalities.
Macrosomia may lead to obstetric complications such as shoulder dystocia, although maternal diabetes may
cause neonatal complications such as hypoglycemia, polycythemia, and cardiac abnormalities (29). Detection of
macrosomia sonographically includes serial assessment of growth and the relationship of the variables. The ratio
of abdominal circumference to head circumference is particularly useful in identifying an enlarging abdomen
indicative of excessive weight gain. Other measurements have been proposed to aid in the diagnosis of
macrosomia, including soft tissue thickness of the humerus or femur and cheek-to-cheek diameter (30).
Fetal Maturity
The sonographic assessment of fetal maturity is useful in the optimal timing of perinatal management. The
perinatologist must maintain a balance between the risk of fetal death and the risk of neonatal death based on
an estimate of whether a fetus has reached an age and weight at which lung maturity is possible. Fetal age and
weight determinations, evaluation of placental architecture, and the biophysical profile have been used to assess
fetal maturity. Studies on elective deliveries based on BPD, or weight estimates based on BPD and abdominal
circumference, have shown good results (31,32 and 33). The grading of placental maturity sonographically,
however, has not been shown to be as accurate in assessing lung maturity as examination of the amniotic fluid
(34).
FETAL ANATOMY
Congenital anomalies are present in 2% to 5% of newborns and account for 20% to 30% of perinatal deaths
(35). High-resolution ultrasonography has produced significant
P.170
changes in the diagnosis and management of these anomalies, including the potential for therapeutic
intervention. Factors to be considered in the counseling of these cases include gestational age, effect on
maternal outcome, and neonatal prognosis with or without therapy. Amniocentesis should be considered in cases
at risk for a chromosomal anomaly.
Figure 12-4 Transvaginal scan of the fetal brain in the coronal plane clearly demonstrates mild dilatation of
both occipital horns.
Ultrasound is the initial imaging modality of choice for the assessment of the fetus as it is safe, inexpensive, and
easily performed. A comprehensive sonographic review includes a survey of the fetus, followed by evaluation of
amniotic fluid volume, cord structure, and the placenta. A functional review of the fetus including hand clenching
and swallowing is informative. Because some anomalies may not become evident until later in gestation (such as
progressive hydrocephalus, congenital diaphragmatic hernia), follow-up evaluation of the fetus at risk is
important. Transvaginal sonography and three-dimensional sonographic imaging have provided a means of
improved visualization of fetal anomalies, particularly of the face and brain (Fig. 12-4).
Fetal Magnetic Resonance Imaging

Factors such as fetal lie, maternal obesity, and oligohydramnios may interfere with the sonographic evaluation of
fetus. MRI is an alternative modality to evaluate the fetus as it uses no ionizing radiation, has excellent tissue
contrast, large field of view and multiplanar capabilities (Fig. 12-5) (36,37). The development of faster scanning
techniques now allow studies to be performed without sedation in the second trimester with superior contrast,
and without significant motion artifact.
Although there currently is no evidence that the use of clinical MRI produces deleterious effects on human
embryos, the safety of MRI during pregnancy has not been definitively demonstrated (37,38). While there has
been concern that prolonged exposure to electromagnetic radiation may be linked to deleterious effects on
embryogenesis, animal studies on embryos have so far demonstrated no growth abnormalities (40,41). Limited
follow-up studies on children exposed to fetal MRI also have not demonstrated developmental abnormalities
(40,42,43). Due to the possibility of unknown effects, MRI currently is used as a complementary study only
when ultrasound results are inadequate for diagnosis or when MR imaging is expected to provide important
additional information.
Figure 12-5 Large field of view T2 weighted sagittal image demonstrates the brain, chest, abdomen, and leg of
a 20 week fetus.
With advances in the diagnosis and treatment of fetal abnormalities there is an increasing need for precise
depiction of abnormalities. Fetal MRI not only confirms the presence of lesions noted by ultrasound, but also can
demonstrate additional subtle anomalies. This information helps appropriate specialists, including the
neonatologist, counsel families when difficult decisions must be made regarding early and potentially invasive
intervention and allows for optimal delivery planning.
Fetal Central Nervous System Abnormalities

Fetal central nervous system (CNS) abnormalities may be detected in the first trimester in severe cases such as
anencephaly. The most favorable time to evaluate the fetal neural axis, however, is 18 to 24 weeks of gestation,
when resolution is sufficient to identify most structures including the spine. Abnormal measurements of BPD and
head circumference may be the first suggestion of a CNS abnormality. Structural evaluation includes the shape
and size of the head, ventricles, and posterior fossa and an evaluation of the spine. If a CNS anomaly is
identified, careful evaluation for other anomalies, especially of the face, heart, and kidneys, follows.
The ventricular diameter normally measures less than 10 mm throughout the second and third trimesters
P.171
(44,45). Other criteria of ventriculomegaly include separation of choroid plexus from the medial ventricular wall.
If the diagnosis of ventriculomegaly is established, a search for its etiology should be made. A differential
diagnosis ranges from cerebral atrophy (hydranencephaly, porencephaly) (Fig. 12-6), dysgenesis
(holoprosencephaly, agenesis of the corpus callosum), to true hydrocephalus.
Figure 12-6 A: Axial sonogram of the brain at 29 weeks of gestation demonstrates unilateral dilatation of the
left lateral ventricle (cursers). B: Follow-up CT scan at birth confirms the finding of left hemispheric
porencephaly (arrow).
There are significant limitations in the sonographic evaluation of the fetal brain as a result of refraction artifact
from the skull and low sensitivity to cortical malformations. MRI better defines anomalies of the posterior fossa,
corpus callosum, and developing gray and white matter. Three planes of the brain can be obtained with excellent
delineation of the extraaxial spaces (Fig. 12-7). Sulcal and gyral development can be assessed allowing for
estimation of brain maturity (Fig. 12-8) (46,47 and 48).
Figure 12-7 Coronal T2 weighted image through the brain of a 25 week fetus with aqueductal stenosis
demonstrates moderate lateral and third ventricular dilatation.
Hydrocephalus
Hydrocephalus can be caused by an increased rate of cerebrospinal fluid (CSF) formation as in choroid plexus
papilloma, decreased resorption of CSF after hemorrhage, or obstruction of CSF flow as a result of tumor or
aqueductal stenosis (49). In many cases, the etiology of hydrocephalus remains unknown. Up to 85% of cases
are associated with other neurological, cardiac, renal, gastrointestinal (GI), or skeletal anomalies. Up to 10% are
associated with chromosomal abnormalities (50). Fetuses with other anatomic or chromosomal abnormalities
have a grim prognosis for long-term development. When ventriculomegaly is mild, and no associated
abnormalities are identified, outcome is less severe. Up to 20% of these cases, however, may demonstrate
neurological abnormalities later in life (51,52 and 53).
Figure 12-8 Oblique T2 weighted image through the brain of a 32 week fetus with trisomy 21 demonstrates
moderate ventriculomegaly with a more mature sulcation pattern.
Figure 12-9 Coronal T2 weighted image through the brain of 19 week gestation demonstrates agenesis of the
corpus callosum and high riding third ventricle.
P.172
MRI is useful in confirming the presence of ventriculomegaly, and can assess symmetry of ventricular size and
contour. Irregularity of the ventricular wall and adjacent parenchymal abnormalities may indicate post
inflammatory and destructive changes. Atrophy, hemorrhages, or masses may be identified (48). Identifying the
etiology for ventriculomegaly, however, may still be difficult as subtle gyral abnormalities and parenchymal
changes may be impossible to detect by MRI in the second trimester.
Multiplanar views are particularly useful in the evaluation of the corpus callosum, which develops between 8 to
17 weeks gestation (Fig. 12-9). When severe, holoprosencephaly can be identified by ultrasound; less severe
semilobar and lobar types are more difficult to diagnose even by MRI.
The posterior fossa and brain stem are well delineated by MRI and can help differentiate Dandy Walker
malformation for posterior fossa arachnoid cysts or mega cisterna magna (Fig 12-10).
Figure 12-10 Sagittal T2 weighted image of a 22 week gestation demonstrates a Dandy Walker cyst with
absent inferior vermis.
Vascular CNS anomalies including vein of Galen aneurysms are well demonstrated by MRI. Hemorrhage and or
encephalomalacia suggests a poorer prognosis and is better demonstrated by MRI.
Neural Tube Defects

The most common condition leading to ventriculomegaly is spina bifida. Evaluation of the fetal spine and
cerebellum allows prenatal identification of over 95% of fetuses with meningomyeloceles. Screening programs
using maternal serum -fetoprotein identify pregnancies at risk for neural tube or ventral wall defects.
Sonographic findings of meningomyelocele include splaying of pedicles, presence of a cystic collection posterior
to the fetal spine with absent skin covering, and scoliosis. Because of the difficulty in identifying small spinal
defects prenatally, the association of cranial abnormalities is useful in evaluating cases at risk. The Arnold-Chiari
malformation, which often accompanies spina bifida, results in the herniation of the cerebellum through the
foramen magnum into the upper cervical spinal canal. Cranial sonographic findings of this malformation include
frontal bone scalloping (lemon sign), abnormal curvature of the cerebellum (banana sign), and obliteration of
the cisterna magna (54,55). MRI is useful in the confirmation of Chiari malformations, with amount of tonsillar
herniation measured (46).
Cranial defects involving brain (encephalocele) or meninges (meningocele) may present with ventriculomegaly
(Fig 12-11). Typically on ultrasonography there is a posterior paracranial mass that is cystic or solid. It is
important to identify a true skull defect because the diagnosis can be confused with scalp edema, cystic
hygroma, or scalp hemangiomas, which carry a better prognosis (56). MRI is particularly useful in identifying
intracranial anomalies associated with encephaloceles. Because encephalocele is a feature of other syndromes
including Meckel-Gruber, additional anomalies should be excluded (57).
Figure 12-11 Sagitta T2 weighted image of a 24 week fetus demonstrates an occipital meningocele with
microcephaly and sulcation anomalies.
Figure 12-12 Transverse image through the chest of a 25-week fetus demonstrates bilateral large pleural
effusions (white arrowheads). The heart (black arrow) is slightly deviated to the right.
P.173
Fetal Chest Abnormalities

Many major intrathoracic structures can be identified sonographically by the second trimester. A small chest
circumference suggests pulmonary hypoplasia in the presence of oligohydramnios or skeletal dysplasia (58).
Pleural effusions appear as fluid collections surrounding collapsed echogenic lung (Fig. 12-12). Effusions may be
primary in chylothorax or secondary in nonimmune hydrops. In the presence of effusions, prognosis is variable,
with 15% mortality if an effusion is isolated but up to 95% mortality if it is associated with severe hydrops
(59,60).
The normal fetal mediastinum lies in the center of the chest, anterior to the spine, with the cardiac
interventricular septum forming a 45-degree angle with the midline. Deviation of this axis or shift of the heart
suggests a mass effect or cardiac pathology.
Figure 12-13 Cystic adenomatoid malformation. Sagittal image of the left fetal chest at 26 weeks of gestation
demonstrates multiple cysts (arrows) within the left hemithorax dispersed between unusually echogenic lung
parenchyma.
Fetal lung parenchyma should be homogeneously echogenic. If it is heterogenous and/or there is shift of midline
structures, a chest mass is likely. Sonographic findings will vary depending on the type of chest mass. Cystic
adenomatoid malformations (CCAM) type 1 are the most common and appear as a single or multiple
macrocysts. Type 2 CCAM contain small cysts less than 1 cm in size and are often associated with other
anomalies. Type 3 CCAM contain multiple microcysts that appear sonographically as a homogeneously echogenic
mass (Fig. 12-13) (61,62). Sequestrations are intralobar or extralobar masses of pulmonary tissue that lack a
tracheobronchial communication and have a vascular supply from the aorta. They present as a solid or mixed
solid and cystic mass in the inferior portion of the chest (63). Up to 25% have a CCAM component that may be
cystic. With color Doppler, a vessel can often be seen coursing from the aorta to feed the mass, confirming the
diagnosis. Bronchogenic cysts typically present as a simple mediastinal or lower lobe cyst without a feeding
vessel (64).
MRI is particularly valuable in the evaluation of fetal thorax. Fetal lungs are homogeneously intermediate in
signal and clearly delineated from mediastinal structures, liver, and bowel (64,65). Cystic lung masses are
typically higher in signal as a result of the high fluid content and can be separated from normal surrounding lung
parenchyma (Fig. 12-14a and b) (67,68).
Prognosis depends on the size of the intrathoracic mass because marked lung compression during fetal
development leads to hypoplastic lungs. Polyhydramnios, ascites, and hydrops are likely secondary to
compression of the esophagus and vena cava and correlate with poor outcome (61). Type 3 CCAM tend to have
the worst prognosis. The prenatal placement of shunts in cysts or pleura has resulted in some success in lung
reexpansion (68). Spontaneous resolution of pleural effusions and pulmonary masses in utero has been reported
(Fig. 12-13) (60,61,62). Prenatal therapy is reserved for those cases at highest risk for poor outcome (e.g.,
hydrops) (61,69,70,71). A fetus with a lung mass has an excellent prognosis if there is no hydrops and only
minimal pulmonary hypoplasia (71).
Fetal neck masses such as cystic teratomas and lymphangiomas can occlude the airway resulting in hypoxic
brain injury or death on delivery. The prenatal detection of a neck mass allows for the planning of a more
controlled delivery such as Exutero Intrapartum Treatment (EXIT) procedure to establish an airway (72,73).
Fetal MRI is particularly useful in providing additional detailed anatomic information in three planes of the larynx
and trachea which are not visualized sonographically (72,73).
Congenital Diaphragmatic Hernia

When a fetal thoracic mass is identified, the differential diagnosis includes congenital diaphragmatic hernia. This
defect results from incomplete fusion of the pleuroperitoneal membranes during the seventh week of gestation.
Passage of bowel and liver may occur into the thorax at any
P.174
time during pregnancy. Sonographic findings include contralateral mediastinal shift and fluid-filled loops of bowel
within the chest that mimic pulmonary cystic masses (Fig. 12-15). Findings specific to congenital
diaphragmatic hernia (CDH) include absence of an intraabdominal stomach bubble or loops of bowel within the
chest. Doppler flow studies demonstrating mesenteric vessels extending into the hemithorax confirm the
diagnosis. Bowing of the umbilical portion of the portal vein suggests liver herniation. Associated anomalies
often present include cardiac, genitourinary, CNS, and GI. Other conditions associated with congenital
diaphragmatic hernia are Fryns' syndrome with nuchal thickening and limb anomalies and Pallister Killian
syndrome (74).
Figure 12-14A. Coronal T2 weighted image of a fetal chest at 22 weeks gestation demonstrates multiple high
signal cysts occupying the left hemithorax. B. Axial CT of the chest following delivery demonstrates multiple air
filled cysts in the left upper lobe compatible with a cystic adenomatoid malformation.
Figure 12-15 Congenital diaphragmatic hernia. Transverse sonographic image through the chest of a 23-week
fetus demonstrates the heart (white arrow) to be shifted to the right by a herniated fluid-filled stomach (black
arrow).
Further evaluation includes fetal chromosomal analysis and fetal echocardiograms. Fetal MRI is particularly
useful in differentiating a CCAM from a CDH. MRI can assess how much residual lung parenchyma remains
expanded, and determine the amount of liver and bowel that has herniated into the thorax (Fig. 12-16) (36).
Serial sonograms every 2 to 4 weeks are important as up to 75% of cases develop polyhydramnios, and all are
at risk for IUGR (75). The mortality rate for infants with diaphragmatic hernia is variable. Cases with multiple
anomalies are likely to be lethal. Less clear is whether sonographic findings such as marked mediastinal shift,
IUGR, hydrops, and polyhydramnios predict uniformly poor outcome (75,76). Ratios of right lung area to head
circumference and the presence of liver in the chest appear most useful in identifying which cases carry the
poorest prognosis (77,78).

Congenital heart disease is one of the most commonly recognized birth defects with a reported frequency of
0.8%. Any woman with a risk factor for a cardiac anomaly should undergo a detailed ultrasound examination of
the fetal heart. Risk factors include nonimmune hydrops, suspected abnormality by screening sonogram,
teratogen exposure, parental or sibling heart defect, aneuploidy, extracardiac anomalies, maternal diabetes, and
fetal arrhythmia (80).
Detailed cardiac evaluation is difficult before 18 weeks of gestation. Examination includes visualization of the
P.175
four chambers and outflow tracts. A dedicated fetal echocardiogram requires additional cross-sectional views to
demonstrate fetal heart integrity. M-mode echocardiography measures chamber size, wall thickness, and wall
motion and facilitates cardiac rhythm assessment. Pulsed Doppler and color flow Doppler are useful to define
blood flow. Sensitivity in detecting anomalies is dependent on fetal position, maternal size, and amniotic fluid
volume and equipment and expertise. Ventricular septal defects, anomalous pulmonary venous return, and
aortic or pulmonic stenosis are especially difficult to diagnose. Ott, in 1995, reported a 14% sensitivity for
detection of congenital heart disease in a low-risk population compared with a 63% sensitivity in the high-risk
group (81). Heart disease diagnosed prenatally is often severe and associated with a poor long-term prognosis
with chromosomal or extracardiac structural defects often present (82).
Figure 12-16 Congenital diaphragmatic hernia. Coronal T2 weighted image of a chest at 32 weeks gestation
demonstrates herniation of stomach, bowel, and spleen into the left hemithorax. There is minimal deviation of
mediastinal structures with the right lung and left upper lobes expanded.
Fetal Gastrointestinal Tract

The ultrasound appearance of fetal bowel varies, with significant overlap between normal and abnormal patterns
(83). The stomach is first visualized at 13 to 16 weeks of gestation. If it is not identified by 16 weeks, especially
in the presence of polyhydramnios, an abnormality such as esophageal atresia or congenital diaphragmatic
hernia should be suspected (84). The esophagus is typically collapsed and not visualized by ultrasound scans
(US), but the absence of a fluid-filled stomach and presence of polyhydramnios and a dilated proximal pouch are
findings suggestive of esophageal atresia. In cases of tracheoesophageal fistula, 70% have a fluid-filled
stomach, with polyhydramnios developing in only 60% of cases. Thus, sonographic detection of
tracheoesophageal fistula (TEF) in unlikely (85). TEF may be detected by MRI if a proximal dilated esophageal
pouch is noted and the distal esophagus is not visualized (86).
When an enlarged fluid-filled duodenum and stomach are identified, duodenal obstruction is present. Duodenal
atresia is the most common cause, but the differential diagnosis includes annular pancreas, or duodenal
stenosis. Up to 30% of fetuses with isolated duodenal atresia have trisomy 21, so chromosome analysis and
fetal echocardiography are indicated. Serial sonograms will indicate development of polyhydramnios or IUGR.
In a normal fetus, small bowel is only occasionally visualized, and loops measuring over 7 mm in diameter are
considered abnormal and suggestive of obstruction (87) (Fig. 12-17). Differential diagnosis includes a
duplication cyst, mesenteric cyst, ovarian cyst, and hydroureter. Polyhydramnios is not always present and
depends on the site of obstruction. Because a fetus with a large bowel obstruction or anal atresia typically does
not develop polyhydramnios or dilated bowel, these entities are less readily diagnosed in utero. At times, bowel
obstruction is associated with perforation and meconium spilling into the peritoneum, otherwise known as
meconium peritonitis. Sonographic findings include fetal ascites, peritoneal calcifications, and meconium
pseudocysts (88). The most common cause of small bowel obstruction is atresia felt to be secondary to an in
utero vascular infarct. Fifteen percent of infants with cystic fibrosis present with meconium ileus and sonographic
findings similar to those of bowel atresia. Echogenic bowel, meconium peritonitis, and pseudocyst formation
occur in both meconium ileus and isolated bowel atresia. Echogenic bowel has also been associated with
cytomegalovirus (CMV) and with chromosomal abnormalities such as trisomy 21 (89). Amniocentesis allows for
karyotyping, CMV cultures, and analysis of amniocytes for the common cystic fibrosis (CF) mutations. In cases
of suspected bowel abnormalities, serial sonograms will detect new onset of polyhydramnios and meconium
peritonitis.
Fetal bowel is well visualized by MRI. Meconium is low signal on T2w and high signal on T1w. Proximal bowel
loops are high signal on T2w and low on T1w. MR can provide additional information regarding level of
obstruction. Anal atresia, typically missed by US may be identified by MRI (90).
Ventral Abdominal Wall Defects
The most common abdominal wall defects identified sonographically include omphalocele and gastroschisis (91).
Through maternal serum alpha-fetoprotein screening, these defects can be diagnosed early in pregnancy.
Omphaloceles result from failure of intestines to return to the abdomen during the tenth week of gestation with
herniation of bowel or liver into the umbilical cord. A
P.176
surrounding peritoneal membrane is present, and the cord insertion is central (Figs. 12-18, 12-19).
Gastroschisis, on the other hand, is a paraumbilical defect located to the right of the umbilicus and is a fullthickness abdominal wall defect without a covering membrane. It is important to distinguish between the two
entities for associated diagnosis and prognosis. Gastroschisis is classically an isolated entity felt to be caused by
a vascular event and is not associated with chromosomal anomalies but is complicated by bowel fibrosis (92).
Figure 12-17 A: Transverse image of a 33-week gestation fetal abdomen demonstrates multiple dilated loops
of bowel (black arrowheads). Echogenic focus (white arrow) is consistent with a peritoneal calcification. B:
Following delivery at term, a water-soluble enema demonstrates a normal-caliber colon with dilated air-filled
small bowel loops (arrow). Jejunal atresia was found at surgery.
Figure 12-18 Omphalocele. Transverse sonographic image at the level of the umbilical cord insertion
demonstrates a large outpouching (arrows) with a covering membrane containing liver and bowel.
Omphaloceles, although not as likely to have bowel fibrosis, are more at risk for associated chromosomal
abnormalities (30%-50%) and syndromes such as Beckwith-Wiedemann syndrome (macroglossia, organo-
megaly) and pentalogy of Cantrell (ectopia cordis) (93).

Normal fetal bowel migrates into the base of the cord by 12 menstrual weeks. Thus, the sonographic diagnosis
of abdominal wall defects should be made only with caution before the second trimester. Chromosomal analysis
and fetal echocardiography should be offered, and consultation with a pediatric surgeon is helpful. Serial
sonograms will diagnose the development of polyhydramnios and IUGR (94,95).
Genitourinary Tract
Many asymptomatic genitourinary abnormalities are now being identified prenatally by sonography. Most renal
lesions that are identified are cystic or obstructive. Increased detection and earlier diagnosis aid in minimizing
P.177
the risk of further renal damage after birth. Detection of the fetus with irreversible or lethal renal disease also
assists in obstetric and perinatal management. Renal anomalies that can be diagnosed prenatally include renal
agenesis, dilated obstructed or nonobstructed collecting systems, renal cystic disease, and tumors. It is
estimated that 1 in 1,000 fetuses have a renal anomaly (96,97). In a series of 17,000 women screened
prospectively at 16 to 18 weeks, renal anomalies were identified in 313 cases, 55 of which were significant.
Upper tract dilation was the most common finding (298 cases), but it was transient in two-thirds of these cases.
Obstruction was noted in 23 infants on follow-up, with 15 of these requiring surgery. Eight infants had unilateral
multicystic kidney disease, and three had posterior urethral valves (98). Mortality after prenatal identification of
fetal uropathy ranges from 20% to 50%.
Figure 12-19 Omphalocele. Sagittal T2 weighted imaged demonstrates liver outpouching at the level of the
cord insertion.
A systematic approach to urinary tract abnormalities includes an assessment of amniotic fluid, characterization
of the urinary tract abnormality, and a search for additional abnormalities. The kidneys contribute little amniotic
fluid until 16 weeks of gestation, making it difficult to assess renal function before the second trimester. The
presence of oligohydramnios secondary to a urinary abnormality carries a poor prognosis. Rarely,
polyhydramnios may be present in a fetus with a mesoblastic nephroma, incomplete ureteropelvic junction (UPJ)
obstruction, or associated cranial or GI abnormalities.
Fetal kidneys can be visualized as early as 14 weeks of gestation. Standard renal measurements include length
[age in weeks = fetal kidney length (mm)], AP diameter, and renal circumference to abdominal circumference
ratio (0.270.33). If a fetal renal pelvis is dilated and fluid-filled, one must determine if it is physiological,
obstructive, secondary to reflux, or simply nonobstructive megacystis. In most cases, if the pelvic diameter
measures less than 1 cm, the finding is nonpathological. When the pelvic diameter measures over 1 cm, the
ratio of pelvic diameter to renal diameter measures greater than 0.5, or caliectasis or hydroureter is present, a
pathological process is likely (99).
Obstructive Uropathy
The most common cause of fetal pyelectasis is UPJ obstruction (98). Differential includes extrarenal pelvis,
vesicoureteral reflux, or a multicystic dysplastic kidney. When bilateral, the severity of obstruction is usually
asymmetric. Rarely, dysplasia, urinoma, or urine ascites develops. MRI is helpful in defining anatomy of complex
renal anomalies such as bladder extrophy and cloacal anomalies (100). If associated nonrenal anomalies are
identified, chromosome analysis is indicated. The frequency of follow-up sonograms depends on whether both
kidneys are involved and the severity of obstruction. Outcome is variable, with 10% increasing in dilation, 50%
remaining stable, and 40% improving (101). Prophylactic antibiotics should be provided to the infant after
delivery, until work up, including renal scans can document whether the dilatation is truly obstructive.
Bladder Outlet Obstruction

If hydroureter and pyelectasis are bilateral, the differential diagnosis includes urethral obstruction, bilateral
reflux, or bilateral megacystis. In the male fetus, the differential also includes posterior urethral valves and
prune belly syndrome. Accurate diagnosis of posterior valves includes the finding of a dilated, thick bladder with
keyhole posterior urethral expansion, dilated tortuous ureters, and caliectasis. Fetal MRI may aid in the
assessment of ureteral and urethral dilatation particularly if oligohydramnios is present. Urinary ascites may be
present (Fig. 12-20). Findings that suggest a poor outcome include dysplastic echogenic renal parenchyma and
oligohydramnios (102). If fetal bladder aspirate demonstrates elevated fetal urine electrolytes and elevated
osmolarity, a poor outcome is also suggested.
Frequent sonograms will detect progressive oligohydramnios and IUGR. If severe oligohydramnios is present, up
to 95% die from pulmonary hypoplasia. Those who survive require an emergency voiding cystourethrogram to
confirm the diagnosis and valvuloplasty or cystostomy to relieve the obstruction. Up to 40% of survivors may
develop renal failure at a later date.
Cystic Renal Disease

Cystic changes of the fetal kidneys may be secondary to hereditary diseases (autosomal recessive infantile
polycystic
P.178
kidneys), dysmorphology (multicystic dysplastic kidney disease), or severe obstruction (Potter type II).
Differentiating the various diagnoses, at times, may be difficult. When renal cystic disease is identified, the most
important factors to evaluate are amniotic fluid volume and whether one or both kidneys are affected. If the
cysts are unilateral and amniotic fluid is normal, prognosis typically is good.
Figure 12-20 Longitudinal image of a right kidney in a 23-week gestation male fetus with posterior urethral
valves demonstrates a kidney with a large upper pole cortical cyst (black arrow), caliectasis (arrowheads), and
urinary ascites (white arrow).
Multicystic dysplastic kidneys are felt to be an early error in development of mesonephric blastema or early
obstructive uropathy. Up to 75% of multicystic dysplastic kidneys are associated with other renal abnormalities
in the contralateral kidney, especially UPJ obstruction and vesicoureteral reflux. Sonographically, multiple cysts
are identified in various sizes that do not connect to a renal pelvis. No normal renal parenchymal tissue is seen.
Following delivery, the diagnosis is confirmed by renal radionuclide scan, and vesicoureteral reflux is excluded by
voiding cystourethrogram.
Autosomal recessive infantile polycystic kidneys typically present with bilateral large echogenic kidneys with a
large abdominal circumference. If severe oligohydramnios develops, outcome is poor because of pulmonary
hypoplasia (Fig. 12-21).
Skeletal Dysplasia
Skeletal dysplasias include a heterogeneous group of over 160 disorders. The prevalence is 2.4 per 10,000
births (103). Up to one-fourth of affected infants are stillborn, with another one-third dying by 1 week of age.
The most common lethal dysplasias include thanatophoric dysplasia, osteogenesis imperfecta type II, and
achondrogenesis. The most common nonlethal skeletal dysplasia is achondroplasia (104).
Prenatally, a systematic approach is needed to analyze skeletal anomalies. Long bones should be evaluated for
size, shape, bowing, and symmetry. Shortening of the extremities can involve the proximal segment
(rhizomelic), midsegment (mesomelic), distal segment (acromelic), or the entire limb (micromelic). Fractures
appear as bones that are irregular, angled, or bowed sonographically (Fig. 12-22). The skull should be evaluated
for frontal bossing or cloverleaf deformity. Decreased skull echogenicity may be noted with osteogenesis
imperfecta and hypophosphatasia. Hyper-telorism, micrognathia, and abnormally shaped ears may be identified.
Hands and feet should be evaluated for polydactyly, missing digits, or equinovarus deformities. The presence of
hemivertebrae, scoliosis, and platyspondyly should be explored. Although short ribs may be difficult to recognize
sonographically, the thoracic circumference can be measured and compared to normal values for gestational
age. If gestational age is unknown, the ratio of thoracic to abdominal circumference can be used. A small thorax
suggests a poor prognosis because chest restriction results in pulmonary hypoplasia and, at times, hydrops.
Additional findings such as cleft lip, cardiac, and renal anomalies help narrow the differential diagnosis.
Figure 12-21 Infantile polycystic kidney disease. Coronal image of a 24-week fetus demonstrates bilateral
enlarged echogenic kidneys (arrows). Severe oligohydramnios is present.
When there is a positive family history of a skeletal dysplasia, accurate prenatal diagnosis is possible. When a
skeletal abnormality is noted incidentally, a precise diagnosis is more difficult. Fetal radiographs help confirm a
diagnosis if fractures or joint calcifications are identified. Additional biochemical testing or karyotyping is useful.
Because dysplasias may progress with time (e.g., achondroplasia and osteogenesis imperfecta), serial
sonograms are useful for assessment of skeletal growth in cases at risk
P.179
(105). The potential development of polyhydramnios and/or hydrops should also be noted. Differentiating
intrauterine growth retardation or constitutional short stature from a true dysplasia is more difficult in the third
trimester. Assessment of fetal well-being including the biophysical profile and umbilical arterial flow patterns
aids in establishing the correct diagnosis.
Figure 12-22 Osteogenesis imperfecta type III. A: Longitudinal image of a fetal thigh at 38 weeks of gestation
demonstrates a short, bowed femur (cursors). B: Radiograph after delivery confirms the presence of short
bowed femurs, tibia, and fibula bilaterally.
DETECTION OF FETAL ASPHYXIA

Fetal asphyxia is a common cause of fetal injury or death. Fetal asphyxia is a spectrum of conditions ranging
from transient episodes of hypoxemia to sustained hypoxemia with resultant metabolic acidosis. There are
diverse signs of asphyxia with multiple organs potentially affected and a wide range in severity, duration, and
chronicity. When asphyxia develops, the fetus redistributes cardiac output from nonvital organs (lung, kidney,
skeleton) to vital organs (brain, adrenals, and myocardium. This protective redistribution is reflexive, resulting
from hypoxemic or acidemic stimulation of aortic body chemoreceptors (23). When sustained, redistribution can
be profound. Diminished amniotic fluid production develops as a result of decreased urine production.
Oligohydramnios further endangers the fetus because the umbilical cord is more at risk for compression,
especially during uterine contractions. Cerebral asphyxia results in alterations in biophysical activities and
responses. It is unclear whether increasing degrees of hypoxemia cause progressive loss of a biophysical
function or if all functions are lost simultaneously. Fetal breathing has been noted to disappear early, whereas
fetal movements disappear only with more severe disease (106,107).
Identification of an asphyxiated fetus is important because emergency delivery may improve outcome. Acute
fetal asphyxia is rare except in cases of cord prolapse or placental abruption. These obstetric complications can
be recognized sonographically and suggest appropriate intervention (108). Unfortunately, most asphyxiated
fetuses exhibit both subacute and chronic signs of asphyxia. The fetus suffering from uteroplacental insufficiency
typically has intermittent hypoxemic episodes induced during episodes of uterine contraction. Between these
periods, the fetus can be normally oxygenated or demonstrate minimal compensated hypoxemia. Signs of
hypoxemia thus depend on the duration, progression, and frequency of hypoxemic episodes.
The biophysical signs of asphyxia may be acute, such as loss of breathing movements and heart rate reactivity,
or they may be chronic, such as oligohydramnios. Biophysical responses will change during hypoxemic episodes.
Experiments in fetal lambs have demonstrated a decrease in fetal breathing and gross movements during
hypoxemia (109,110). Fetal muscle tone and heart rate response to movement (i.e., accelerations) also
decrease during hypoxemia (111). Maternal depressant medication and fetal cerebral anomalies can alter fetal
biophysical activities, so these causes should be excluded. Ultrasound findings of chronic effects of intermittent
hypoxemia include intrauterine growth retardation. Oligohydramnios should be considered a sign of fetal
asphyxia unless renal abnormalities or premature rupture of membranes is identified (22).
P.180
Fetal Biophysical Profile

Monitoring fetal activity and its responses has altered the practice of perinatology (112,113). Any fetus at risk
for uteroplacental insufficiency is a candidate for fetal evaluation in the third trimester. The first electronic fetal
heart rate evaluation developed was the contraction stress test (CST), which relies on the fetal heart rate
response to uterine contractions. The test records the fetal heart response as contractions are induced with
oxytocin. If no fetal decelerations are recorded, uteroplacental function is considered normal. Disadvantages
include the need to administer oxytocin intravenously and the length of the examination (average time of 90
minutes) (114).
The nonstress test (NST) evaluates fetal heart rate acceleration in response to fetal movement. As with a
negative CST, a reactive NST is predictive of intrauterine survival for approximately 7 days subsequent to the
test. This test has become popular because it requires relatively little time, cost, and technical skill (115).
Reproducibility of NST interpretation, however, is problematic (116). Past acute asphyxial events are not
excluded solely with a NST, and nonreactive patterns may exist for benign reasons.
The differentiation of the normal from the compromised fetus may be most accurate when several fetal and
environmental parameters are evaluated together. Manning proposed the combined use of five ultrasoundmonitored fetal variables to assess risk of asphyxia to reduce both false-positive and false-negative results.
These include fetal breathing movements, gross body movements, tone, amniotic fluid volume, and heart rate
reactivity (117). Each variable is evaluated according to tested criteria, and they are assigned scores of 2 when
normal and 0 when abnormal (Table 12-2). Observation of each variable is continued until normal criteria are
met or 30 minutes of continuous observation has been completed. The total score, termed the biophysical profile
score (BPS), is an accurate method of differentiating if a fetus is normal or compromised (Table 12-3). Manning
et al. reported that in a series of over 19,000 pregnancies, there was a false-normal test rate of approximately
one per 1,000, with more that 97% of pregnancies tested having normal test results (118). Clinical studies from
other centers have reported similar results with this method (119). In a follow up study, Manning and associates
reported on 493 fetuses in which BPS were performed just before umbilical venous blood pH values were
obtained via cordocentesis. A BPS of zero was associated with significant fetal acidemia, whereas scores of 8 or
greater were associated with a normal pH 120). A score of 6 was a poor predictor of abnormal outcome, but a
decrease from a score of 2 to 4 to a score of zero was an accurate predictor of abnormal outcome. In a similar
study, for each of the four acute variables of the BPS, a substantial difference in mean umbilical venous pH was
noted for the normal and abnormal results (121). Manning and associates prospectively studied the incidence of
cerebral palsy in 22,336 high-risk pregnancies managed with serial BPS compared with 30,224 low-risk
pregnancies that did not receive antepartum testing. Cerebral palsy (CP) diagnosed by 3 years of age was
significantly associated with a low BPS. Incidence of CP was 0.8 per 1,000 when the BPS score was normal
(10/10) and 250 per 1,000 when the BPS was very abnormal (0/10) (122).
TABLE 12-2 BIOPHYSICAL PROFILE SCORING: TECHNIQUE AND

INTERPRETATION
Biophysical Variable
FBMs
Normal (score 2)
Abnormal (score 0)
At least one episode of
Absent FBM or no
FBM of at least 30-sec

duration in a 30-min
episode of 30 sec in 30
min
observation period
Gross body movement At least three discrete
Two or fewer episodes of
body or limb movements body or limb movements

in 30 min (episodes of
in 30 min
active continuous
movement are
considered a single
movement)
Fetal tone
At least one episode of

active extension with
Either slow extension

with return to partial
return to flexion of fetal flexion or movement of

limb(s) or trunk; opening limb in full extension or
and closing of hand
absent fetal movement
considered normal tone
Reactive FHR
At least two episodes in

30 min of FHR
acceleration of 15 bpm
of at least 15-sec
Qualitative AFV
Less than two episodes

of acceleration of FHR or
acceleration of <15 bpm
in 30 min
duration associated with

fetal movement
At least one pocket of AF Either no AF pockets or a
that measures at least 2 pocket <2 cm in two

cm in tow perpendicular perpendicular planes
planes
AF, amniotic fluid; AFV, amniotic fluid volume; bpm, beats per minute;
FBM, fetal breathing movement; FHR, fetal heart rate.
From Manning FA, Morrison I, Lange IR, et al. Fetal assessment based on
fetal biophysical profile scoring: experience in 12,620 referred high risk
pregnancies. I. Perinatal mortality by frequency and etiology. Am J Obstet
Gynecol 1985;151:343, with permission.
Because the BPS is labor intensive, abbreviated screening tests have been
evaluated. When normal, the four ultrasound monitored variables appear equal to
that achieved by the addition of the NST component. The NST should be
performed when one or more variables are abnormal

P.181
(123). Provided amniotic fluid volume is normal, a BPS of 8 to 10 by whatever
combination can be considered normal. All equivocal or abnormal BPS results
should be based on a complete assessment including a NST (124).
TABLE 12-3 BIOPHYSICAL PROFILE SCORING: MANAGEMENT PROTOCOL
Score
10
Interpretation
Management
Normal infant, low risk for
-Repeat testing at weekly
chronic asphyxia and patients

over 42 weeks of gestation
intervals; repeat twice weekly in

diabetics
Normal infant, low risk for

-Repeat testing at weekly
chronic asphyxia-diabetics and intervals; repeat testing twice
patients over 42 weeks of
weekly in oligohydramnios is an
gestation;
Suspicion of chronic asphyxia
indication for delivery

Repeat testing in 46 hours;
deliver if oligohydramnios present
Suspicion of chronic asphyxia
-If past 36 weeks of gestation and

favorable, deliver; if less than 36
-weeks of gestation and L/S

0-2
<2, repeat test in 24 hrs;
repeat score 4, deliver
Strong suspicion of chronic

asphyxia regardless of
gestational age
-Extend testing time to 120 min;

if persistent score 4, deliver
L/S, lecithin-sphingomyelin ratio.

From Manning FA, Morrison I, Lange IR, et al. Fetal assessment based on
pregnancies. I. Perinatal mortality by frequency and etiology.
Am J Obstet Gynecol 1985;151:343, with permission.
Biophysical evaluation should be performed in patients with recognized high-risk
factors. It allows for accurate differentiation of the normal fetus from the
compromised one, information crucial in the timing of intervention. The variables
used in the fetal BPS were selected because of ease and rapidity of measurement.
Additional variables such as swallowing, fine hand movement, rapid-eyemovement state, peristalsis of fetal gut, urine output, and umbilical vessel flow
rates may be found to add to the more precise evaluation of the fetal condition
(125).
DOPPLER ASSESSMENT OF BLOOD VELOCITIES

IN UMBILICAL AND FETAL VESSELS
Doppler ultrasound velocimetry can provide the clinician with important
information on the hemodynamics of various vascular regions. Flow waveforms
provide important information 12 through 40 weeks gestation from maternal
vessels, placental circulation, and fetal systemic vessels. Gestational age-related
reference values have been established for maternal uterine and arcuate artery,
umbilical artery, fetal descending aorta, and fetal cerebral, renal, and femoral
arteries (126). Values are obtained by directing an ultrasound beam toward a
vessel and recording the frequency shift (Doppler effect) in the returning echoes.
The Doppler shift principle states that echoes returning from moving structures
are altered in frequency, and the amount of shift is directly proportional to the
velocity of the moving structure. The frequencies of echoes returning from blood
moving toward the transducer are increased, whereas the frequencies of echoes
returning from blood moving away from the transducer are decreased. This
method may not accurately measure velocity, but it does accurately evaluate the
relative resistance to flow (127). The frequency spectrum can be analyzed by
comparing the systolic to diastolic flow velocities. There are several different ratios
for measurement of flow impedance. These include the difference between peak
systolic and diastolic flow velocity over the mean flow velocity (pulsatility index),
the difference between peak systolic and diastolic flow over the peak systolic flow
(resistive index), and the ratio of the peak systolic to diastolic flow (S/D ratio). All
indices are independent of the angle of insonation.
Measurement of flow velocity waveforms in the uterine artery, which depicts
maternal vascular effects of the invading placenta, has been shown to be a
method for early recognition of the pregnancy at risk for preeclampsia and IUGR
(128,129). Uterine arterial remodeling occurs following successful placentation.
Prepregnancy, uterine arteries show high resistance, elastic recoil with early
diastolic notches and low diastolic flow. Successful placental invasion removes
intimal muscle with increasing diastolic flow and loss of elasticity (130). Indices
decrease as term approaches, with diastolic velocity increasing with advancing
gestation. Fetal growth retardation has been reported when indices increase or
notching in the waveform at end-systole develops at 2224 weeks gestation
(130,131). Uterine flow velocity waveforms are useful in predicting the frequency
and severity of preeclampsia and intrauterine growth retardation
(128,129,130,131). A major clinical obstetric application has been in the
measurement of umbilical arterial flow using the S/D ratio (Fig. 12-23). In a
normal pregnancy, the placental resistance declines slowly with advancing
gestational age. The umbilical artery peak systolic flow velocity gradually
decreases although diastolic flow velocity increases. Thus, the umbilical artery S/D
ratio declines with advancing gestational age and should be less than 3 after 30
weeks of gestation (132). In growth-retarded fetuses and fetuses developing

intrauterine distress, the umbilical artery blood velocity waveform changes as
placental vascular resistance increases.
P.182
Abnormalities are progressive with reduction, loss, and finally reversal of diastolic
flow. Absent or reversed end-diastolic flow (AREDV) is associated with a high risk
of morbidity and mortality. Studies have compared the prediction of fetal acidosis
by Doppler and cord blood gases. A statistically significant correlation has been
found between an increased umbilical S/D ratio and fetal hypoxia particularly
when diastolic blood flow is absent or reversed (133). In the preselected
population of high-risk pregnancies, this method has a high predictive value
regarding diagnosing fetal compromise and can be used for monitoring fetal
health (128,134,135). If an abnormal umbilical velocimetry is identified in an SGA
fetus, fetal heart rate variability should be tested. Fetuses that remain normal
after a nonstress test have a low morbidity and can be closely followed (136,137).
Figure 12-23 Umbilical artery velocimetry. S, systolic velocity; D,

diastolic velocity. Note S/D ratio of approximately 3.
Applied as a screening test in an unselected pregnant population, umbilical

velocimetry has currently not been found to be cost-effective (138). A
significant change in indices does not develop until more than 60% of the
terminal placental arteries are obliterated (129). On the other hand, in
preselected populations of high-risk pregnancies, Doppler of the umbilical
artery has a high predictive value regarding diagnosing fetal compromise
(135). These abnormal Doppler velocities are associated with higher
mortality and morbidity in the neonatal period. Long term follow-up studies
have shown an association between the abnormal intrauterine umbilical and
fetal blood flow and subsequent postnatal neurodevelopmental impairment
including mental retardation and severe motor impairment (139,140,141).
Absent diastolic flow velocity of the umbilical artery has been noted in up to
25% of fetuses with anomalies, typically trisomy 13 or 18 (136). Because
abnormal velocimetry may indicate a chronic fetal abnormality and not
acute distress, delivery should not be undertaken based on umbilical artery
Doppler velocimetry alone (130,137). In the fetus with IUGR and equivocal
umbilical blood flow studies, Doppler studies of fetal systemic vessels can
be useful.
The fetus with an abnormal umbilical artery velocimetry will redistribute
blood flow detectable as vasodilation of the main arteries of the fetus
including the heart and brain (23). Measurable fetal flow velocities include
the middle cerebral, descending aorta, renal arteries and ductus venosus
(142,143). These measures provide a means of assessing the distribution of
fetal blood flow and can be useful in assessing the fetal cardiovascular
adaptive responses to chronic hypoxemia. Doppler studies of the fetal
circulation in intrauterine growth retardation and hypoxia have
demonstrated a compensatory redistribution of arterial blood flow with
increased flow to the brain and myocardium with decreased flow to the
periphery (144,145). Redistribution of cardiac output reflects rising
placenta resistance and results in brain sparing when cerebrovascular
dilatation occurs (130).
Doppler waveforms of the middle cerebral artery (MCA) can be obtained at
the level of the circle of Willis with a high degree of reproducibility.
Normally, the MCA impedance decreases with advanced gestational age but
remains higher than that of the umbilical artery (144). Fetal brain sparing
during hypoxia is characterized by increase in diastolic and mean blood flow
velocity in the MCA. If vasodilation of the MCA is lost, the fetus will begin
to enter the acidotic stage. At term, evidence of fetal hemodynamic
redistribution may exist in the presence of normal umbilical artery indices,
so ratios such as the MCA/uric acid (UA) ratio can be useful. Normal
cerebral to placental resistance ratio (CPR) is typically greater than 1
(146,147,148).
In the compromised IUGR fetus, precordial veins can illustrate fetal cardiac
function. Which venous waveforms provide the best data are still being
investigated (130,131). The ductus venosus regulates venous control, is a
conduit of right atrial retrograde pulse waves, and is responsive to changes
in oxygenation. In fetuses compromised by anemia or hypoxia, changes in
pulsatile venous flow are being evaluated
(130,131,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38
,39,40,41,42,43,44,45,46,47,48,49,150). When umbilical artery flow
becomes abnormal, further evaluation of the fetal systemic Doppler may
improve analysis of fetal distress. Doppler studies of fetal intracranial
arteries and ductus venosus may thus result in improved timing of
intervention to minimize morbidity (145,146,147,148,149).
SUMMARY
Ultrasonography has profoundly influenced the practice of perinatal
medicine. Doppler and fetal MRI have also become useful adjuncts in the
evaluation of the fetus. The ability to distinguish the normal from abnormal
pregnancy has many applications. Serious complications such as
developmental anomalies, intrauterine asphyxia, and growth
P.183
abnormalities are now identified with greater frequency and accuracy as a
result of these advances. Recognition of high-risk conditions such as
placental abruption and cord prolapse and more chronic conditions such as
intrauterine asphyxia allow for optimal preventive care. The improved
accuracy in identifying fetal anomalies by ultrasound and MRI has increased
the potential for prenatal intervention and optimizing the care and delivery
of the fetus.
REFERENCES
1. Dewhurst CJ, Beasley JM, Campbell S. Assessment of fetal maturity and
dysmaturity. Obstet Gynecol 1972;113:141.
2. Donald I. Ultrasound in obstetrics. Br Med Bull 1968;24:71.
3. Silva PD, Mahairas G, Schaper AM, et al. Early crown-rump length a good
predictor of gestational age. J Reprod Med 1990;35: 641.
4. Selbing A. Gestation age and U.S. measurement of gestational sac, CRL,
and BPD during the first 15 weeks of pregnancy. Acta Obstet Gynecol Scand
1982;61:233.
5. Campbell S, Newman GB. Growth of the fetal biparietal diameter during
normal pregnancy. Br J Obstet Gynaecol 1971;78:513.
6. Sabbagha RE, Hughey M. Standardization of sonar cephalometry and
gestational age. Obstet Gynecol 1978;52:402.
7. Hadlock FP, Deter RL, Carpenter RL, et al. The effect of head shape on
the accuracy of BPD I estimating fetal gestational age. AJR Am J Roentgenol
1981;137:83.
8. Sabbagha RE, Turner JH, Rockette H, et al. Sonar BPD and fetal age:
definition of the relationship. Obstet Gynecol 1974;43:7.
9. Queenan JT, O'Brien GD, Campbell S. Ultrasound measurement of fetal
limb bones. Am J Obstet Gynecol 1980;138:297.
10. Hadlock FP, Deter RL, Harrist RB. Computer assisted analysis of fetal
age using multiple fetal growth parameters. J Clin Ultra-sound
1983;11:313.
11. Deter RL, Harrist RB, Birnholz JC, et al. Evaluation of fetal dating
studies. In: Hedlock F, Deter RL, eds. Qualitative obstetrical
ultrasonography. New York: John Wiley & Sons, 1986:33.
12. Hadlock FP, Deter RL, Harrist RB, et al. A date independent predictor of
IUGR:FL/AC ratio. Am J Roentgenol 1983;141:979.
13. Campbell S, Wilkin D. Ultrasound measurement of fetal abdominal
circumference in the estimation of fetal weight. Br J Obstet Gynaecol
1975;82:689.
14. Warsof SL, Gohari P, Berkowitz RL, et al. The estimation of fetal weight
by computer assisted analysis. Am J Obstet Gynecol 1977; 128:881.
15. Shepard MJ, Richards VA, Berkowitz RL, et al. An evaluation of two
equations for predicting fetal weight by ultrasound. Am J Obstet Gynecol
1982;142:47.
16. Kubik-Huch RA, Wildermuth S, Cettuzzi L, et al. Fetus and
uteroplacental unit fast MRI and 3D volumetry feasibility study. Radiology
2001;219:567.
17. Deter RL, Harrist RB, Hadlock FP, et al. Longitudinal studies of fetal
growth with the use of dynamic image ultrasonography. Am J Obstet
Gynecol 1982;143:545.
18. Morrison I. Perinatal Mortality. Semin Perinatol 1985;9:144.
19. O'Brien GD, Queenan JT. Ultrasound fetal femur length in relation to
intrauterine growth retardation. Am J Obstet Gynecol 1982;144:33.
20. Tamura RK, Sabbagha RE. Percentile ranks of sonar fetal abdominal
circumference measurements. Am J Obstet Gynecol 1980; 138:475.
21. Sabbagha RE. Intrauterine growth retardation: antenatal diagnosis by
ultrasound. Obstet Gynecol 1978;52:252.
22. Wladimiroff JW, Bloemsma CA, Wallenburg HCS. Ultrasound assessment
of fetal head and body sizes in relation to normal and retarded fetal growth.
Am J Obstet Gynecol 1978;131:857.
23. Cohn HE, Sacks EJ, Heyman MA, et al. Cardiovascular responses to
hypoxemia and acidemia in fetal lambs. Am J Obstet Gynecol 1974;120:817.
24. Phelan JP, Smith CV, Broussard P, et al. Amniotic fluid volume
assessment with the four quadrant technique at 3642 weeks gestation. J
Reprod Med 1987;32:540.
25. Manning FA, Hill LM, Platt LD. Qualitative amniotic fluid volume
determination by ultrasound: antepartum detection of intrauterine growth
retardation. Am J Obstet Gynecol 1981;1139:254.
26. Marks AD, Divon MY. Longitudinal study of the amniotic fluid index in
postdated pregnancy. Obstet Gynecol 1992;79:229.
27. Chamberlain PF, Manning FA, Morrison I, et al. Ultrasound evaluation of
amniotic fluid volume: I. The significance of marginal and decreased
amniotic fluid volume to perinatal outcome. Am J Obstet Gynecol
1984;150:245.
28. Manning FA, Morrison I, Lange IR, et al. Fetal assessment based on
pregnancies: I. Perinatal mortality by frequency and etiology. Am J Obstet

Gynecol 1985;151:343.
29. Boulet LS, Alexander GR, Salihu HM et al. Macorsomic births in the
United States: Determinants, outcomes, and proposed grades of risk. Am J
30. Petrikovsky B, Gelertner N, Oleschuk C. Fetal abdominal fat line: can
macrosomia be diagnosed. Am J Gynecol 1996;174:428.
31. Goldstein P, Gershenson D, Hobbins JC. Fetal biparietal diameter as a
predictor of mature L/S ratio. Obstet Gynecol 1976;438: 667.
32. Strassner HT, Plat LD, Whittle M, et al. Amniotic fluid
phosphatidylglycerol and real-time ultrasonic cephalometry. Am J Obstet
Gynecol 1979;135:804.
33. Golde SH, Platt LD. The use of ultrasound in the diagnosis of fetal lung
maturity. Clin Obstet Gynecol 1984;27:391.
34. Harman CR, Manning FA, Stearns E, et al. The correlation of ultrasonic
placental grading and fetal pulmonic maturity in five hundred and sixtythree pregnancies. Am J Obstet Gynecol 1982; 143:941.
35. Lee K, Khoshnood B, Chen L, et al. Infant mortality from congenital
malformations in the United States 19701997. Obstet Gynecol
2001;98:620.
36. Hubbarb AM, Adzick NS, Crombleholme TM, et al. Leftsided congenital
diaphragmatic hernia: value of prenatal MR imaging in preparation for fetal
surgery. Radiology 1997;203:636.
37. Levine D, Barnes PD, Madsen JR, et al. Fetal central nervous system
anomalies: MR imaging augments sonographic diagnosis. Radiology
1997;204:635.
38. Yip YP, Capriotti C, Tlagala, et al. Effects of MR exposure at 1.5 T on
early embryonic development of the chick. J Magn Reson Imaging
1994;4:742.
39. Gu Y, Hasegawa T, Yamamato Y, et al. The combined efects of MRI and
Xrays on ICR mouse embryos during organogenesis. J Radiat Res
2001;42:265.
40. Ginsberg JS, Hirsh J, Rainbow AJ, et al. Risks to the fetus of radiologic
procedures uses in the diagnosis of maternal venous thromboembolic
disease. Thromb Haemost 1989;61:189196.
41. Okazaki R, Ootsuyama A, Uchida S, et al. Effects of a 4.7 T statid

magnetic field on fetal development in ICR mice. J Radiat Res 2001;42:273.
42. Baker PN, Johnson IR, Harvey PR, et al. A three year follow up children
imaged in utero with echoplanar magnetic resonance. Am J Obstet Gynecol
1994;170:32.
43. Mevissen M, Buntenkotter S, Loscher W. Effect of static and time
varying magnetic field on reproduction and fetal development in rats.
Teratology 1994;50:229237.
44. Siedler DE, Filly RA. Relative growth of the higher brain structures. J
Ultrasound Med 1987;6:573.
45. Pretorius D, Davis K, Manco-Johnson M, et al. Clinical course of fetal
hydrocephalus: 40 cases. ANJR Am J Neuroradiol 1985;6:23.
46. Bilaniuk LT. MRI of the fetal brain. Semin Roentgenol 1999; 32:4861.
47. Girard N, Raybaud C, Ponset M. In vivo MR study of brain maturation in
normal fetuses. AJNR Am J Neuroradiol 1995;16: 407413.
48. Levine D, Barnes PD. Cortical maturation in normal and abnormal
fetuses as assessed with prenatal MR imaging Radiology 1999;210:751
758.
49. Drugan A, Kraus B, Canady A, et al. The natural history of prenatally
diagnosed cerebral ventriculomegaly. JAMA 1989;261: 1785.
50. Chervenak FA, Berkowitz RL, Tortora M, et al. The management of fetal
hydrocephalus. Am J Obstet Gynecol 1985;151:933.
51. Brown IM, Bannister CM, Rimmer S, et al. The outcome for infants
diagnosed prenatally as having cerebral ventriculomegaly. J Matern Fetal
Neonatal Med 1995;5:13.
P.184
52. Patel MD, Filly AL, Hersh DR, et al. Isolated mild fetal cerebral
ventriculomegaly: Clinical course and outcome. Radiology 192:759:1994.
53. Levitsky DB, Mack LA, Nyberg DA, et al. Fetal aqueductal stenosis
diagnosed sonographically: How grave is the prognosis? Am J Roentgenol
1995;164:725.
54. Benaceraff BR, Stryker J, Frigoletto FD. Abnormal US appearance of the
cerebellum (banana sign). Indirect sign of spina bifida. Radiology
1989;171:151.
55. Van de Hof MC, Nicolaides KH, Campbell J, et al. Evaluation of the
lemon and banana signs in one hundred thirty fetuses with open spina
bifida. Am J Obstet Gynecol 1990;162:322.
56. Bulas DI, Johnson D, Allen J, et al. Fetal hemangioma. Sonographic and
color flow Doppler findings. J Ultrasound Med 1992;11:499.
57. Benson JT, Dillard RG, Burton BK. Open spina bifida: does c-section
delivery improve prognosis? Obstet Gynecol 1988;71: 532.
58. Songster GS, Gray DL, Crane JP. Prenatal prediction of lethal pulmonary
hypoplasia using US fetal chest circumference. Obstet Gynecol
1989;73:261.
59. Estroff JA, Parak R, Frigoletto FD, et al. The natural history of isolated
fetal hydrothorax. Ultrasound Obstet Gynecol 1992;2: 162.
60. Lein JM, Colmorgen GHC, Gehret JF, et al. Spontaneous resolution of
fetal pleural effusion diagnosed during the second trimester. J Clin
Ultrasound 1990;18:54.
61. Adzick NS, Harrison MR, Glick PL, et al. Fetal cystic adenomatoid
malformation: prenatal diagnosis and natural history. J Pediatr Surg
1985;20:483.
62. Saltzman DH, Adzick NS, Benacerraf BR. Fetal cystic adenomatoid
malformation of the lung: apparent improvement in utero. Obstet Gynecol
1988;71:1000.
63. Benya EC, Bulas DI, Selby DM, et al. Cystic sonographic appearance of
extralobar pulmonary sequestration. Pediatr Radiol 1993;23:605.
64. Albright EB, Crane JP, Shackelford GD. Prenatal diagnosis of a
bronchogenic cyst. J Ultrasound Med 1988;7:91.
65. Coakley FV, Lopoo JB, Ying LU, et al Normal and hypoplastic fetal lungs:
volumetric assessment with prenatal single shot rapid acquisition with
relaxation enhancement MRI. Radiology 2000;216:107.
66. Duncan KR, Gowland PA, Moore RJ, et al. Assessment of fetal lung
growth in utero with echoplanar MR imaging. Radiology 1999;210:197.
67. Hubbard AM, Cromplehome TM. Anomalies and malformation affecting
the fetal neonatal chest. Semin Roentgenol 1998;33:117.
68. Hubbard AM. Magnetic resonance imaging of fetal thoracic
abnormalities. Top Magn Reson Imaging 2001;12:18.
69. Blott M, Nicolaides KH, Greenough A. Pleuroamniotic shunting for
decompression of fetal pleural effusions. Obstet Gynecol 1988;71:798.
70. Bromley B, Parad R, Estroff JA, et al. Fetal lung masses: prenatal
course and outcome. J Ultrasound Med 1995;14:927.
71. Songster GS, Gray DL, Crane JP. Prenatal prediction of lethal pulmonary
hypoplasia using US fetal chest circumference. Obstet Gynecol
1989;73:261.
72. Hubbard AM, Crombleholme TM, Adzick NS. Prenatal MRI Evaluation of
Giant Neck Masses in Preparation for the Fetal EXIT procedure. Am J
73. Kathary N, Bulas D, Newman K, et al. Fetal MRI evaluation of cervical
neck masses. Pediatr Rad 2001;31:72774.
74. Bulas DI, Saal HM, Fonda J, et al. Cystic hygroma and CDH: early
prenatal evaluation of Fryns syndrome. Prenat Diag 1992;12:867.
75. Adzick NS, Harrison MR, Glick PR. Diaphragmatic hernia in the fetus:
prenatal diagnosis and outcome in 94 cases. J Pediatr Surg 1985;20:357.
76. Wilson JM, Fauza DO, Lund DP, et al. Antenatal diagnosis of isolated
CDH is not an indicator of outcome. J Pediatr Surg 1994;29:815.
77. Metkus AP, Filly RA, Stringer MD, et al. Sonographic predictors of
survival in fetal diaphragmatic hernia. J Pediatr Surg 1996;31:148.
78. Harrison MR. The fetus with a diaphragmatic hernia: pathophysiology,
natural history and surgical management. In Harrison MR, Bolvus MS, Filly
RA, eds. The unborn patient, 2nd ed. Philadelphia: WB Saunders, 1990:295.
79. VanderWall KJ, Skarsgard ED, Filly RA, et al. Fetendoclip: a fetal
endoscopic tracheal clip procedure in a human fetus. J Pediatr Surg
1997;32:970.
80. Perone N. A practical guide to fetal echocardiography. Contemp Obstet
Gynecol 1988;1:55.
81. Ott WJ. The accuracy of antenatal fetal echocardiography screening in
high and low risk patients. Am J Obstet Gynecol 1995;172:1741.
82. Crawford DC, Chita SK, Allan LD. Prenatal detection of congenital heart
disease: factors affecting obstetric management and survival. Am J Obstet
Gynecol 1988;159:352.
83. Hertzberg BS. Sonography of the fetal gastrointestinal tract: anatomic
variants, diagnostic pitfalls and abnormalities. Am J Roentgenol
1994;162:1175.
84. McKenna KM, Goldstein RB, Stringer MD. Small or absent fetal stomach:
prognostic significance. Radiology 1995;197:729.
85. Pretorius DH, Drose JA, Dennis MA, et al. Tracheoesophageal fistula in
utero: twenty-two cases. J Ultrasound Med 1986;6:509.
86. Langer JC, Hussain H, Khan A, et al. Prenatal diagnosis of esophageal
atresia using MRI. J Pediatr Surg 2001;36:804.
87. Nyberg DA, Mack LA, Patten RM, et al. Fetal bowel, normal sonographic
findings. J Ultrasound Med 1987;6:257.
88. Foster MA, Nyberg DA, Mahony BS, et al. Meconium peritonitis prenatal
sonographic findings and their clinical significance. Radiology
1987;165:661.
89. Nyberg DA, Dubinsky TJ, Resta RG, et al. Echogenic fetal bowel during
the second trimester: clinical importance. Radiology 1993;188:527.
90. Benachi A, Sonig P, Jounnic JM, et al. Detemination of the anatomical
location of an antenatal intestinal occlusion by magnetic resonance imaging.
Ultrasound Obstet Gynecol 2001;18:164165.
91. Hertzberg BS, Bowie JD. Fetal gastrointestinal abnormalities. Radiol Clin
North Am 1990;28:101.
92. Babcock CJ, Hedrick MH, Goldstein RB, et al. Gastroschisis: can
sonography of the fetal bowel accurately predict postnatal outcome. J
Ultrasound Med 1993;13:701.
93. Nicolaides KH, Snifders RJM, Cheng HH, et al. Fetal gastrointestinal and
abdominal wall defects: associated malformation and chromosomal
abnormalities. Fetal Diagn Ther 1992;7:102.
94. Chescheir NC, Azizkhan RG, Seeds JW, et al. Counseling and care for
the pregnancy complicated by gastroschisis. Am J Perinatol 1991;8:323.
95. Fitzsimmons J, Nyberg DA, Cey DR, et al. Perinatal management of
gastroschisis. Obstet Gynecol 1988;71:910.
96. Estes J, Harrison M. Fetal Obstructive Uropathy. Semin Pediatr Surg
1993;2:129.
97. Cusick E, Didier F, Droulle P, et al. Mortality after an antenatal
diagnosis of fetal uropathy. J Pediatr Surg 1995;30:463.
98. Gunn TR, Moral JD, Pease P. Antenatal diagnosis of urinary tract
abnormalities by ultrasonography after 28 weeks: incidence and outcome.
99. Fernbach SK, Maizels M, Conway JJ. Ultrasound grading of
hydronephrosis; introduction to the system used by the Society for Fetal
Urology. Pediatr Radiol 1993;23:478.
100. Hubbard A, Harty M, Ruchelli E, et al. Prenatal MRI of the fetal urinary
tract: normal and abnormal anatomy with US and pathological correlation.
Radiology 1998;209:259.
101. King LR, Hatcher PA. Natural history of fetal and neonatal
hydronephrosis. Pediatr Urol 1990;35:433.
102. Hutton KE, Thomas DF, Arthur RJ, et al. Prenatally detected posterior
urethral valves: is gestational age at detection a predictor of outcome? J
Urol 1994;152:698.
103. Camera G, Mastroizcovo P. Birth prevalence of skeletal dysplasia in the
Italian multicentric monitoring system for birth defects. In: Papadatos CJ,
Bartsocas CCS, eds. Skeletal dysplasias. New York: Alan R Liss, 1982:441.
104. Clark RN. Congenital dysplasias and dwarfism. Pediatr Rev
1990;12:149.
105. Bulas DI, Stern HJ, Rosenbaum KN, et al. Variable prenatal appearance
of osteogenesis imperfecta. J Ultrasound Med 1994;13:419.
106. Vintzileos AM, Rippert LSM, Sniders RJM, Nicolaides KH, et al.
Relationship of fetal biophysical profile score and blood gas values in
severly growth retarded fetuses. Am J Obstet Gynecol 1990; 163:569.
107. Manning FA, Platt LD. Maternal hypoxemia and fetal breathing
movement. Obstet Gynecol 1979;53:758.
108. Lange IR, Manning FA, Morrison I, et al. Cord prolapse: is antenatal
diagnosis possible? Am J Obstet Gynecol 1985;151:1083.
P.185
109. Boddy K, Dawes GS. Fetal breathing. Br Med Bull 1976;31:3.

110. Natale R, Clewlow F, Dawes GS. Measurement of fetal forelimb
movements in the lamb in utero. Am J Obstet Gynecol 1981; 140:545.
111. Brown R, Patrick J. The non-stress test: how long is enough? Am J
112. Manning FA. Dynamic ultrasound based fetal assessment: the fetal
biophysical profile score. Clinical Obstet Gynecol 1995; 3826:44.
113. William K. Amniotic fluid assessment. Obstet Gynecol Surv
1993;46:795.
114. Ray M, Freeman R, Pine S, et al. Clinical experience with oxytocin
challenge test. Am J Obstet Gynecol 1972;114:1.
115. Freeman RK. The use of oxytocin challenge test for antepartum clinical
evaluation of uteroplacental respiratory function. Am J Obstet Gynecol
1975;121:481.
116. Hage ML. Interpretation of nonstress tests. Am J Obstet Gynecol
1985;153:490.
117. Manning FA, Platt LD, Sipos L. Antepartum fetal evaluation:
development of a fetal biophysical profile score. Am J Obstet Gynecol
1980;136:787.
118. Manning FA, Morrison I, Harman CR, et al. Fetal assesment based on
fetal biophysical prophile score. Experience with 19,221 referred high risk
pregnancies. Am J Obstet Gynecol 1987;157:880.
119. Baskett TF, Gray JH, Prewett ST, et al. Antepartum fetal assessment
using a fetal biophysical profile score. Am J Obstet Gynecol 1983;148:630.
120. Manning FA, Snijder R, Harman CR, et al. Fetal BPS correlation with
antepartum umbilical venous pH. Am J Obstet Gynecol 1993;169:755.
121. Vintzileos AM, Gaffney SE, Salinger LM, et al. The relationship between
fetal biophysical profile and cord pH in patients undergoing C section.
122. Manning FA, Harman C, Menticoglou S. Fetal BPS and cerebral palsy at
age 3 years. Am J Obstet Gynecol 1996;174:319.
123. Manning FA, Harman CR, Lange IR, et al. Modified fetal BPS by
selective use of the NST. Am J Obstet Gynecol 1987;156:709.
124. Manning FA. Dynamic ultrasound based fetal assessment: the fetal
biophysical profile score. Clin Obstet Gynecol 1995;38:26.
125. Marsal K. Rational use of Doppler ultrasound in perinatal medicine. J
Perinat Med 1994;22:463.
126. Johnson T. Maternal perception and Doppler detection of fetal
movement. Clin Perinatol 1994;21:765.
127. Gill RW. Measurement of blood flow by ultrasound accuracy and source
of error. Ultrasound Med Biol 1985;11:625.
128. Fleischer A, Schumlman H, Farmakides G. Uterine artery Doppler
velocimetry in pregnant women with hypertension. Am J Obstet Gynecol
1986;154:806.
129. Thompson RS, Trudinger BJ. Doppler waveform pulsatility index and
resistance pressure and flow in the unbilical placental circulation. Ultasound
Med Biol 1990;16:449.
130. Harman C, Baschet A. Comprehensive assessment of fetal wellbeing:

which Doppler tests should be performed? Curr Opin Obstet Gynecol
2003;15:147.
131. Detti L, Akiyama M, Mari G. Doppler blood flow in obstetrics. Curr Opin
132. Giles WB, Trudinger BJ, Cook CM. Umbilical artery velocity time
waveforms in pregnancy. J Ultrasound Med 1982;1:9.
133. Okamura K, Watanabe T, Ando J, et al. Blood gas profiles of fetuses
with abnormal Doppler flow in the umbilical artery. Am J Perinatol
1996;13:297.
134. Weiner CP. The relationship between umbilical artery S/D ratio and
umbilical blood gas measurements in specimens obtained by cordocentesis.
135. Westergaard HB, Langhoff-Roos J, Lingman G, et al. A critical appraisal
of the use of umbilical artery Doppler ultrasound in high risk pregnancies;
use of metaanalyses in evidence based obstetrics. Ultrasound Obstet
Gynecol 2001;17:46.
136. Rochelson B, Schulman H, Farmakides G, et al. The significance of
absent end diastolic velocity in umbilical artery velocity waveforms. Am J
137. Giles W. Clinical use of Doppler US in pregnancy. Information from six
randomised trials. Fetal Diagn Ther 1993;8:247.
138. Bricker L, Neilson JP. Routine Doppler ultrasound in pregnancy.
Cochrane Database Syst Rev 2;2003.
139. Vossbeck S, de Camargo OD, Grab D, et al. Neonatal and
neurodevelopmental outcome in infants born before 30 weeks of gestation
with absent or reversed end diastolic flow velocities in the umbilical artery.
Eur J Pediatr 2001;160:128.
140. Wienerroither H, Steiner H, Tomaselli J, et al. Intrauterine blood flow
and long term intellectual neurologic and social development. Obstet
Gynecol 2001;97:449.
141. Kurschera J, Tomaselli J, Urlesberger B, et al. Absent or reversed enddiatolic blood flow in the umbilical artery and abnormal Doppler
cerebroplacental ratio-cognitive, neurological, and somatic development at
36 years. Early Hum Dev 2002;29:47.
142. Wladimiroff JW, Wijingaard JAGW, Degani S. Cerebral and umbilical

blood flow velocity waveforms in normal and growth retarded pregnancies.
143. Veille JC, Kanaan C. Duplex Doppler ultrasonographic evaluation of the
fetal lamb artery in normal and abnormal fetuses. Am J Obstet Gynecol
1989;161:1502.
144. Dubiel M, Gudmundsson S, Gunnarson G, et al. Middle cerebral artery
velocimetry as a predictor of hypoxemia in fetuses with increased resistance
to blood flow in the umbilical artery. Early Hum Dev 1997;47:177.
145. Donofreio MT, Bremer YA, Schieken RM, et al. Autoregulation of
Cerebral Blood Flow in Fetuses with Congenital Heart Disease: the brain
sparing effect. Pediatr Cardiol 2003;28.
146. Harrington K, Carpenter RG, Nguyen M, et al. Changes observed in
Doppler studies of the fetal circulation in pregnancies complicated by
preeclampsia or SGA baby. Ultrasound Obstet Gynecol 1995;6:19.
147. Devine PA, Bracero LA, Lyskiewicz A, et al. Middle cerebral to umbilical
artery Doppler ratio in post date pregnancies. Obstet Gynecol 1994;84:856.
148. Zimmermann P, Alback T, Koskinen J, et al. Doppler flow velocimetry
of the umbilical artery, uteroplacental artery and fetal MCA in prolonged
pregnancy. Ultrasound Obstet Gynecol 1995;5:189.
149. Hecher K, Campbell S, Doyle P, et al. Assessment of fetal compromise
by Doppler ultrasound investigation of the fetal circulation. Circulation
1995;91:129.
150. Galan HL, Ferrazzi E, Hobbins JC. Intrauterine growth restriction:
biometric and Doppler assessment. Prenatal Diagn 2002; 22:331.
Chapter 13
Fetal Therapy
Mark I. Evans
Mark P. Johnson
Alan W. Flake
Yuval Yaron
Michael R. Harrison
Over the past three decades, physicians from multiple specialties have developed numerous
methods for the diagnosis of structural and physiological fetal abnormalities (1,2). When they are
severe or lethal, pregnancy termination is viewed by many as a reasonable consideration. For
couples in countries that permit its availability and in cultures in which the fetus does not have
more rights than the mother, a variable portion of patients chose this option (3,4). With more
moderate fetal anomalies, obstetrical care can be modified to optimize outcomes and prevent
secondary complications. In some instances, prenatal treatments of the underlying problem have
become possible. In general, structural malformations are more logically approached with surgery,
although metabolic disorders may benefit from pharmacological or genetic therapies (2).
Fetal therapy has evolved into four major areas: open surgical approaches, closed endoscopic
surgical approaches, pharmacological therapy, and stem cell/gene therapy. Advances in the field
have been characterized by alternating exuberance at spectacular successes, but also periods of
intense frustration at technical challenges to be overcome to bring new approaches on line.
Moving goal posts have also been a common problem, secondary to improvements in ancillary
care that keep raising the bar to show intervention as having a positive benefit to risk ratio.
Even after four decades since the first transfusions by Lilley, the single most misunderstood and
continuous issue about fetal therapy continues to be why before and not after birth. There is no
one single answer; rather there are multiple disorder-specific considerations. If something cannot
be treated safely postnatally, then there is generally justification for prenatal intervention.
However, for many conditions profound and irreparable damage occurs before birth, making fetal
intervention the best or sometimes only way to ameliorate the damage. Some procedures have
been quite rare. Others are more common. The expectation is that with improvements and
increasing utilization of prenatal diagnosis, more women will choose to consider the opportunities
to treat fetuses before birth.
SURGICAL THERAPY
In Utero Closed Fetal Surgery
The most successful in utero fetal surgery has been for the evaluation and treatment of obstructive
uropathy (5,6). Lower urinary tract obstruction (LUTO) is a heterogeneous entity that affects 1:500
8,000 newborn males (5,6,7,8). Posterior urethral valves or urethral atresias are the most
common causes, although stenosis of the urethral meatus, anterior urethral valves, ectopic
insertion of a ureter and tumors of the bladder have also been observed. Massive distention of the
bladder can be seen with compensatory hypertrophy and hyperplasia of the smooth muscle within
the bladder wall. Loss of compliance and elasticity, and poor postnatal function generally require
post natal surgical reconstruction (9). Elevated intravesicular pressures prevent urine inflow from
the ureters, eventually distortion of the ureterovesical angles contributes to reflux hydro-nephrosis
(5,6,7). Progressive pyelectasis and calyectasis compress the delicate renal parenchyma within the
encasing serosal capsule, leading to functional abnormalities within the medullary and eventually
the cortical regions (5,6,7,8,9,10). Focal compressive hypoxia likely contributes to the progressive
fibrosis and perturbations in tubular function resulting in urinary hypertonicity. Obstructive
processes can eventually lead to type IV cystic dysplasia and renal insufficiency (8,9).
The effects extend beyond the genitourinary tract. Progressive oligo/anhydramnios leads to
compressive deformations as seen in Potter sequence, including extremity contractures, facial
dysmorphology, and disruptions of abdominal wall musculature, as in prune belly. Absence of
normal amniotic fluid volume profoundly impedes pulmonary growth and development. Constant
compressive pressure on the fetal thorax leads to restriction of expansion of the chest through
normal physiological breathing movements. Babies born with LUTO mostly die because of
pulmonary complications. They do not live long enough to die of renal failure.
Figure 13-1 Oligohydramnios and dilated bladder in a fetus at 17 weeks with good electrolytes.
P.187
Sonographic findings in LUTO include dilated and thickened walls of the bladder, hydronephrosis,
and oligohydramnios (Fig. 13-1). Urethral strictures or atresia, urethral agenesis, megalourethra,
ureteral reflux, and cloacal anomalies may be present and have a very similar appearance on
ultrasound. The typical keyhole sign of proximal urethral dilation is secondary to urethral
obstruction present in posterior urethral valves or atresia. However, the precise diagnosis can only
be made after birth (9).
The prenatal evaluation and management of fetuses with the sonographic findings of LUTO require
multiple steps (5,6,7,8). Ruling out other congenital anomalies such as cardiac and neural tube
defects is necessary before intervention can be considered.
Karyotyping is essential to confirm a normal male chromosomal status. Most are isolated
problems, but the incidence of aneuploidy is higher than the general population. Female fetuses,
however, almost always have more complex syndromes of cloacal malformations and do not
benefit from in utero shunt therapy. Because of the presence of oligo/anhydramnios, we commonly
obtain karyotypes by transabdominal chorionic villus sampling, which gives reliable results within
several days during which the remainder of the prenatal evaluation is underway. Fluorescence in
situ hybridization is now commonly used to get rapid status of chromosomes 13, 18, 21, X & Y
(11,12).
Essential to the prenatal workup is the evaluation of underlying renal status in the fetus. Over the
past 15 years, a multicomponent approach has developed for the analysis of fetal urine that
evaluates proximal tubular and possible glomerular status using sodium, chloride, osmolality,
calcium, -2 microglobulin, albumin, and total protein concentrations (6,8). It has been shown to
be significantly improved by sequential samplings at 48- to 72-hour intervals to be useful
approach. The degree of impaired renal function and damage with the extent of urinary
hypertonicity and proteinuria can then be directly correlated. The ability to counsel patients about
the renal status of their fetus and the long-term prognosis has been dramatically improved as a
result.
Figure 13-2 Vesicoamniotic shunt with the proximal portion lying within the fetal bladder
although the distal portion lies within the amniotic fluid space allowing diversionary draining of
urine into the appropriate space.
Vesicoamniotic catheter shunts bypass the urethral obstruction diverting the urine into the
amniotic space to allow appropriate drainage of the upper urinary tract and prevention of
pulmonary hypoplasia and physical deformations (Fig. 13-2). In fetuses with isolated LUTO, a
normal male karyotype, and progressively improving urinary profile that meet threshold
parameters (Table 13-1), intervention has been very successful in salvaging fetuses using
percutaneous vesicoamniotic shunt therapy.
Subsequent experience in humans has been widely variable and appears to be related to the
extent of prenatal evaluation prior to shunt placement, and the etiology of obstruction. Freedman
and associates found that prune
P.188
belly infants, without complete urethral obstructions have very good renal outcomes following
vesicoamniotic shunt therapy (9). They also found significant improvement in survival and renal
function in infants with posterior urethral valves treated by shunting. However, many such children
develop mild-to-moderate renal insufficiency at birth and several of these have progressed to renal
failure, requiring dialysis, and transplantation. The worst prognosis appears to be for those with
urethral atresia. There have been survivors, including some with urethral atresia, following early
shunt intervention. Animal studies have indicated that early onset, complete obstructions result in
more severe renal damage than later onset or partial obstructions. Such data emphasizes the
necessity of early diagnosis, evaluation, and intervention to achieve the best outcomes in such
cases.
TABLE 13-1 UPPER THRESHOLD VALUES FOR SELECTING FETUSES THAT MIGHT BENEFIT
FROM PRENATAL INTERVENTION FOR URINARY OBSTRUCTION
Sodium
<100 mg/dl
Chloride
Osmolality
Calcium
<90 mg/dl
<190 mOsm/L
<8 mg/dl
-2 microglobulin
<6 mg/L
Total protein
<40 mg/dl
Data from our experience over the past 15 years suggest that patients having bladder shunts had
a 91% survival, but that long-term renal function was not guaranteed. Just under half had normal
renal function, and about a quarter had mild impairments. The experience depended largely on
the exact etiology of the disorder with posterior urethral value having the best outcomes and
urethral atresia the worst. Our experience suggests that close pediatric urological/renal function
assessment is essential to maximize outcomes. The Paris group has found, consistent with our
experience, that about 25% of children had serious, long-term renal impairments and about 15%
actually developed end-stage renal disease requiring trans-plant (10).
Although vesicoamniotic shunting has certainly improved survival and renal function in cases of
early obstructive uropathy, complications of this procedure remain unacceptably high. We found in
the eighties and nineties that in 40% of our cases, the shunts became physically displaced into the
amniotic or intraperitoneal space, or have become blocked causing loss of drainage function and
necessitating replacement. On balance, intervention for LUTO has clearly saved fetuses, who would
otherwise have surely died. Many have normal to moderately impaired renal function. A carefully
balanced approach in counseling is required for patients to determine what is right for them.
Other Shunt Procedures

Shunting was attempted for obstructive hydrocephalus, in the late 1970s (13). The results were
almost uniformly dismal. Most of the patients operated on were, in retrospect, very poor
candidates for intervention. Many had multisystem syndromic disorders, including aneuploidy, and
hopeless congenital anomalies such as holoprosencephaly.
Ventriculoamniotic shunts were abandoned the early eighties. However, with a better
understanding of the poor natural history of the anomaly, and improved, more accurate diagnostic
techniques, there may eventually be some limited applications for prenatal neurological shunting.
It may eventually be useful in cases of early onset, isolated, progressive obstructive hydrocephaly.
The other use of percutaneously placed shunts has been for thoracic abnormalities (14,15). The
macrocystic form of congenital cystic adenomatoid malformation (CCAM) can present with a very
large intrathoracic mass with a dominant macrocyst which causes cardiac and mediastinal shifts.
Potential hemodynamic changes can result along with pulmonary compression and risk of lung
hypoplasia. Such dominant cysts can be approached using pleuroamniotic shunts to chronically
drain these structures, reducing their volume, and diminishing their space occupying effects within
the thoracic cavity.

Isolated pleural effusions can enlarge, causing hemodynamic changes and onset of generalized
hydrops and pulmonary compression. The mass interferes with normal lung development
increasing the risk of hypoplasia. Small, unilateral effusions generally do not warrant intervention,
but cannot be ignored as they have the potential for rapid progression (15).
As with all fetal interventions, prenatal evaluation prior to intervention is critical for appropriate
case selection. Seemingly isolated effusions may be associated with cardiac malformations,
aneuploidy, anemia, or an infectious process. Thoracoamniotic shunting is effective in carefully
evaluated cases when the risk of pulmonary hypoplasia from large effusions early in gestation is
present, or early signs of progressive hydrops (unilateral or bilateral effusion, skin or scalp edema,
ascites, pericardial effusion) appear. Fetal anemia, per se, is not an indication for thoracoamniotic
effusion shunting. Hydropic changes will usually resolve with timely fetal transfusion therapy. By
the time the fetus develops significant ascites, the prognosis even with successful shunt
intervention, diminishes considerably (15). However, several cases with suddenly progressive
pleural effusions and onset of generalized hydrops have been treated with complete resolution of
all hydropic complications and normal postnatal infants.
Open Surgery
Open fetal surgery has been performed for a limited number of indications for nearly two decades
(3). There are appropriate concerns for maternal risks, rigorous selection criteria, and somewhat
frustrating results. There has been continuing innovation and development of instruments and
techniques, motivated by the clinical necessity to improve the safety of open fetal surgery for both
the fetus and the mother, which have turned the field upside down in many instances.
Congenital Diaphragmatic Hernia

The surgical approach to congenital diaphragmatic hernia (CDH) has evolved considerably since
the first attempted CDH repair in 1986 and the first success in 1989 (16,17,18). Definitive repair
of CDH by reduction of viscera from the chest, diaphragmatic patch placement, and abdominal silo
construction (to reduce intraabdominal pressure) satisfied the desire for a single step approach but
had unacceptable
P.189
mortality, particularly after a clinical trial showed that open fetal repair was no better than
postnatal repair when the liver was herniated into the chest (19,20). The definitive repair was then
abandoned and in utero tracheal occlusion took its place (Fig. 13-3) (20,21,22,23,24,25,26).
Figure 13-3 Surgical isolation of the fetal trachea prior to placement of hemoclips in a tracheal
occlusion procedure for congenital diaphragmatic hernia. (See color plate)
Tracheal occlusion produces increased lung size through accumulation of pulmonary secretions.
The herniated viscera are reduced from the chest and therefore decrease the risk for lung
hypoplasia. The technique of achieving reliable, complete and reversible tracheal occlusion has
evolved. Initially, it could only be accomplished by open fetal surgery and fetal neck dissection
(taking care to avoid the recurrent laryngeal nerves) and placement of occlusive hemoclips. Then,
a fetoscopic technique was developed to accomplish the same neck dissection and tracheal clip
(the Fetendo Clip Procedure) (24). While successful, it proved difficult, with a significant learning
curve. Attempts to simplify the procedure by developing an appropriate polymer to use as a
tracheal plug inserted through the fetal mouth have been generally unsuccessful. Unless there was
complete occlusion, the pulmonary secretions would leak, thereby defeating the purpose of the
plug. Finally, a relatively simple technique was developed in which a fetoscope passed through a
single port is advanced into the fetal trachea (fetal bronchoscopy) and a detachable silicone
balloon is inflated to occlude the trachea.
CDH has a prototype of rapid changes in technology, further clouding any simple attempts to
understand the role for fetal surgery. With increasing sophistication of the surgical approach and
concomitant improvements in neonatal care using extracorporeal membrane oxygenation, it was
impossible to accurately determine the relative benefits of each approach without a prospective,
randomized comparison that held all other details constant. Thus, after much debate, a
randomized trial of surgery for CDH vs. optimal postnatal care was funded by National Institute of
Child Health and Human Development (NICHD) (21). The principal component of the trial was that
patients in the postnatal care arm (control group) would receive the same neonatal care by the
same center as the surgical arm.
It was originally expected that patients having the surgery would have a survival rate of about
70%. The best data on controls showed survivals of about 35%. The surgically treated patients
achieved the expected survival rate. However, by having the controls cared for at the same
tertiary specialty centers as the surgical group, survival in the controls was essentially the same as
the surgical group. Therefore, the trial was stopped prematurely. Such data show dramatically the
principle of the moving target and how our use of technology must continually adapt to changing
conditions (27).
Congenital Cystic Adenomatoid Malformation

CCAM is a space occupying congenital cystic lesion of the lung. Hydrops evolves by causing
mediastinal shift which compromises venous return to the heart. When fetuses with CCAM develop
hydrops, the fetal mortality approaches 100% (Fig. 13-4) (28,29,30). Fetal resection of CCAM can
reverse hydrops and has improved survival dramatically (17). The fetal operation is performed by
exposure of the arm and chest wall on the side of the lesion through the maternal hysterotomy. A
large muscle sparing thoracotomy is performed through the midthorax of the fetus and the lobe
containing the CCAM is isolated. The attachments of
P.190
the lobe to adjacent lung tissue are bluntly divided and the lobar hilum is divided by use of a
stapler or a bulk ligature. During the remainder of the pregnancy, the remaining normal lung
shows compensatory growth to fill the space left following removal of the mass.
Figure 13-4 Microcystic congenital cystic adenomatous malformation (microcystic) with

beginnings of hydrops.
Sacrococcygeal Teratoma
Fetal sacrococcygeal teratoma (SCT) arises from the presacral space, which may grow to massive
proportions and in some fetuses induces high output congestive heart failure from tumor vascular
steal. Fetal SCT with high output physiology and associated placentomegaly or hydrops uniformly
results in fetal demise (Fig. 13-5) (31,32). The pathophysiological rationale for fetal surgery is to
ligate the vascular connections to the tumor, remove the vascular shunt, and reverse the high
output physiology. The fetal operation is performed by exteriorization of the fetal buttocks with
attached tumor (31). The head, torso, and lower extremities of the fetus are kept in utero if at all
possible. Since the tumor can sometimes be larger than the fetus, significant loss of uterine
volume occurs, and the uterus may contract increasing the risk for placental abruption, placental
dysfunction because of compression, or postoperative preterm labor. Once exteriorized, the anus
is identified, and the fetal skin is incised posterior to the anorectal sphincter complex to avoid
injury to the continence mechanism. A tourniquet is then applied at the base of the tumor and
brought down gradually as the tumor is finger fractured down to its vascular pedicle. The vascular
pedicle is then ligated or stapled depending on the width of the pedicle. The entire fetal procedure
can be performed in less than 15 minutes with minimal blood loss. Because of the increase in
afterload following ligation of the low resistance tumor circuit, the fetal hemodynamic status must
be monitored by fetal echocardiography during and in the immediate period following the ligation.
Figure 13-5 Saccrococcygeal teratoma in a fetus. (See color plate)
Neural Tube Defects

Neural tube defects (NTDs) result from abnormal closure of the neural tube which normally occurs
between the third and fourth week of gestational age. The etiology is complex, with both genetic
and environmental factors involved. There are historical data in humans suggesting increased NTD
frequencies in subjects with poor dietary histories or with intestinal bypasses. Analysis of
recurrence patterns within families and of twin-twin concordance data provides evidence of a
genetic influence in nonsyndromal cases. However, factors such as socioeconomic status,
geographic area, occupational exposure, and maternal use of antiepileptic drugs are also
associated with variations in the incidence of NTDs (33). In 1980, Smithells and associates
suggested that vitamin supplementation containing 0.36 mg folate could reduce the frequency of
NTD recurrence by sevenfold (33,34,35). For almost a decade, there has been a great deal of
controversy regarding the benefit of folate supplementation for the prevention of NTDs3
(36,37,38,39). In 1991, a randomized double-blinded trial designed by the Medical Research
Council Vitamin Study Research Group demonstrated that preconceptual folate reduces the risk of
recurrence in high risk patients (40). Subsequently, it was shown that preparations containing
folate and other vitamins also reduce the occurrence of first time NTDs (41). In response to these
findings,
P.191
guidelines were issued calling for consumption of 4.0 mg/day folic acid by women with a prior child
affected with an NTD, for at least 1 month prior to conception through the first 3 months of
pregnancy. Additionally, 0.4 mg/day folic acid is recommended to all women planning a pregnancy
to be taken preconceptually. The data on NTD recurrence prevention is now very well established,
and has become routine practice for high-risk cases. As of January 1998, the United States Food
and Drug Administration have mandated that breads and grains be supplemented with folic acid.
The impact of food fortification with folic acid on NTDs birth prevalence during the years 19901999 was evaluated by assessing birth certificate reports before and after mandatory fortification
(42). It was found that the birth prevalence of NTDs reported decreased by 19%. It is important to
note that the continuing decline in NTDs rates are estimated to be as a result of the introduction
and increased utilization of prenatal diagnosis in addition to the recommendation for multivitamin
use in women of childbearing age and the population-wide increases in blood folate levels because
food fortification was mandated (43). Recently, Evans and associates have shown a 32% drop in
high maternal serum alpha-fetoprotein (MSAFP) values in the United States comparing 2000
values versus 1997 before the introduction of folic acid supplementation (44).
Folate plays a central part in embryonic and fetal development because of its role in nucleic acid
synthesis mandatory for the widespread cell division that takes place during embryogenesis. Folate
deficiency can occur because of low dietary folate intake or because of increased metabolic
requirement as seen in particular genetic alterations such as the polymorphism of the thermolabile
enzyme methyltetrahydrofolate reductase (MTHRF). However, evidence regarding its role in NTD is
unsupported, except in certain populations, suggesting that these variants are not large
contributors to the etiology of NTDs (45,46). Additional candidate genes other than MTHFR may be
responsible for an increased risk for NTDs (47). It has recently been reported that methionine
synthase polymorphisms are associated with increased risk for NTDs, that is not influenced by
maternal preconception folic acid intake at doses of 0.4 mg/day (48). Other candidate genes
include the mitochondrial membrane transporter gene UCP2 (49). Despite previous studies
suggesting zinc deficiency to play a role in the etiology of NTDs (50,51), further studies were
inconclusive (52,53). Because methionine deficiency may be involved in NTDs, it may be beneficial
in NTD risk reduction (54). Preconception folic acid intake as a sole vitamin or as multivitamin
supplementation reduces the risk of recurrence and first time NTDs.
Babies with meningomyeloceles have impaired lower motor function, loss of bowel, and bladder
control. A significant percentage develop obstructive hydrocephalus, which requires
ventriculoperitoneal shunting (55,56). Experience from the 1970s and 1980s showed that babies
with meningomyelocele delivered atraumatically by cesarean section had a better level of motor
function for the given level of anatomic defect, than those babies delivered through the vaginal
canal (55). Such data suggest that compression and trauma to the cord in the delivery process can
have permanent long-term sequelae to motor function. In theory, trauma to the spinal cord in
utero, either from banging into the uterine wall or the toxic effects of the amniotic fluid in the third
trimester, could be detrimental to the function of the spinal cord. Traditional dogma held that the
pathogenesis of meningomyelocele was that an abnormally developed spinal cord, which did not
engender the proper development of the bony spinal column, may not be the whole story. It is
possible that the primary defect is in the bony spinal column, which exposes a presumptively
undamaged spinal cord. The cord is then damaged by the toxic affects of amniotic fluid and trauma
from the uterine environment and repeated contact with the uterine wall. Thus, the rationale for
attempts to cover and protect the spinal cord in utero, to minimize the seque-lae (57).
Three groups (58,59,60) have done most of the work in this area and have attempted to repair
meningomyeloceles in utero, both as an open surgical procedure, and endoscopically, with the
stated attempt to reduce long-term morbidity and mortality. The principal benefit of the surgery is
likely secondary, i.e., a significant reduction in the number of babies requiring ventriculoperitoneal
shunting for obstructive hydrocephalus (58,61). There is still much controversy surrounding the
data (54). A randomized, prospective trial comparing fetal to postnatal neurosurgical closure
began in 2003 but will take several years to be completed. A major milestone of this trial has been
the agreement among the participating centers not to perform any cases outside the trial, and
other centers around the country have agreed not to start programs until the trial is completed.
The Ex Utero Intrapartum Therapy Procedure

The ex utero intrapartum therapy procedure (EXIT) procedure is a modified cesarean delivery and
may be applied to deliver fetuses after fetal surgical procedures such as tracheal ligation, or for
fetuses with difficult airway problems such as massive cervical teratomas or cystic hygromas
(62,63). The most important component of the EXIT procedure is maintenance of uteroplacental
perfusion until the fetal airway is secured and ventilation is established. In direct contrast to
cesarean section in which uterine contraction for hemostasis is encouraged, uterine relaxation is
maintained by deep general anesthesia. The fetal manipulations are then performed with maternal
support via the placenta. Clips can be removed, chest masses removed, bronchoscopy performed,
and stable airway access established in otherwise very difficult circumstances with this approach.
Once the fetus is ready for transport to the nursery or adjacent operating suite following these
preliminary steps, the cord is clamped, cut and the cesarean delivery completed.
P.192
ENDOCRINE DISORDERS
Adrenal Disorders
Congenital Adrenal Hyperplasia
Congenital adrenal hyperplasia is actually not a single disorder but a group of autosomal recessive
metabolic disorders, characterized by enzymatic defects in the steroidogenic pathway (64). A
compensatory increase in adrenocorticotropic hormone secretion leads to overproduction of the
steroid precursors in the adrenal cortex resulting in adrenal hyperplasia. Excess precursors often
are converted to androgens that may result in virilization of female fetuses. The phenotype is
determined by the severity of the cortisol deficiency and the nature of the steroid precursors which
accumulate proximal to the enzymatic block. The most common abnormality, responsible for
greater than 90% of patients with congenital adrenal hyperplasia (CAH) is caused by a deficiency
of the 21-hydroxylase (21-OH) enzyme. Other, less common causes for CAH, include deficiencies
in 11-hydroxylase, 17-hydroxylase, and 3-hydroxysteroid-dehydrogenase. Reduced 21-OH
activity results in accumulation of 17-hydroxyprogesterone (17-OHP) as a result of its decreased
conversion to 11-deoxycorticosterone. Excess 17-OHP is then converted via androstenedione to
androgens, the levels of which increase by as much as several hundred-fold (Fig. 13-6). The
excess androgens cause virilization of the undifferentiated female external genitalia. The degree of
virilization may vary from mild clitoral hypertrophy to complete formation of a phallus and
scrotum. In contrast, genital development in male fetuses is normal. The excess androgens cause
postnatal virilization in both genders and may manifest in precocious puberty (64).
Figure 13-6 STEROIDOGENIC PATHWAYPathway of conversion from Cholesterol to Cortisol is

vulnerable to enzymatic errors. Blockage at 21 hydroxylase leads to over production of 17
hydroxyprogesterone which ultimately leads to excess androgens that produce masculinization of
the external genitalia.
The classical form of CAH involves a severe enzyme deficiency or even a complete block of
enzymatic activity, which is associated in two-thirds to three-fourths with salt-loss that may be lifethreatening. The classical form is easy to recognize in female newborns but may be overlooked in
males, which may present at a later stage with severe dehydration and even demise. The
nonclassical attenuated form of 21-OH deficiency results in partial blockade of the enzymatic
activity and is usually clinically apparent as simple virilization in women only later in life. It is
estimated to occur in about 3.5% in Ashkenazi Jews and about 2% in Hispanics (65). In the late
seventies and early eighties, diagnosis of CAH was made on amniocentesis by the finding of
elevated levels of 17-OHP in the supernatant. In the eighties, with the development of chorionic
villus sampling (CVS) linkage based molecular diagnosis in the first trimester became available,
because the gene for 21-OH was found to be linked to the human leukocyte antigen (HLA)
complex on chromosome 6 (66). The gene for 21-OH (CYP21B) was later mapped, allowing direct
mutation analysis in informative families (67).
It has been known for two decades that the fetal adrenal gland can be pharmacologically
suppressed by maternal replacement doses of dexamethasone (68). Suppression can prevent
masculinization of affected female fetuses in couples who are carriers of classical CAH. Evans, and
associates were first to administer dexamethasone to a carrier
P.193
mother beginning at 10 weeks of gestation in an attempt to prevent masculinization (68). Serial
maternal estriol and cortisol levels indicated that adrenal gland suppression had been achieved.
The female fetus was born at 39 weeks gestation with normal external genitalia. A similar
approach was employed successfully by Forrest and New (69,70). Differentiation of the external
genitalia begins at about 7 weeks of gestation. Thus for a carrier couple, pharmacological therapy
has to be initiated prior to diagnosis. Direct deoxyribonucleic acid (DNA) diagnosis or linkage
studies are then performed by CVS in the first trimester and therapy is continued only if the fetus
is found to be an affected female. Detailed inclusion criteria for treatment have been issued by the
European Society for Pediatric Endocrinology and Wilkins Pediatric Endocrine Society (71).
Thyroid Disorders
Hyperthyroidism
Neonatal hyperthyroidism is rare with an incidence of 1:4,000 to 1:40,000/live births (72). Fetal
thyrotoxic goiter is usually secondary to maternal autoimmune disease, most commonly, Graves'
disease or Hashimoto's thyroiditis. As many as 12% of infants of mothers with a known history of
Graves' disease are affected with neonatal thyrotoxicosis, which may occur even if the mother is
euthyroid (73). The underlying mechanism is the transplacental passage of maternal IgG
antibodies. In this case the antibodies, known as TSAb, are predominantly directed against the
thyroid-stimulating hormone (TSH) receptor. Often in these cases, the fetal goiter is first
diagnosed on ultrasound in patients with elevated thyroid stimulating antibodies. In some cases,
fetal goiters are incidentally detected on routine ultrasonography. Others may be discovered in
patients referred for scan because of polyhydramnios. Untreated fetal hyperthyroidism may be
associated with a mortality rate of 12%-25% as a result of high-output cardiac failure (73).
Once the diagnosis of fetal hyperthyroidism is confirmed, fetal treatment should be initiated.
Authors have attempted treating fetal hyperthyroidism with maternally administered antithyroid
drugs. Porreco has reported maternal treatment of fetal thyrotoxicosis with propylthiouracil (PTU),
which lead to a good outcome (74). The initial dose used was 100 mg p.o. three times a day,
which was later decreased to 50 mg p.o. three times a day. A favorable outcome was shown using
maternal methimazole to treat fetal hyperthyroidism in a patient who could not tolerate PTU (75).
Hatjis also treated fetal goiterous hyperthyroidism with a maternal dose of 300 mg PTU (75). This
patient however, required supplemental Synthroid to remain euthyroid. There was good fetal
outcome in this case as well.
Hypothyroidism
Congenital hypothyroidism is relatively rare affecting about 1:3,000 to 1:4,000 infants (76). About
85% of the cases are the result of thyroid dysgenesis, a heterogeneous group of developmental
defects characterized by inadequate amount of thyroid tissue. Congenital hypothyroidism is only
rarely associated with errors of thyroid hormone synthesis, TSH insensitivity, or absence of the
pituitary gland. Fetal hypothyroidism may not necessarily manifest in a goiter before birth because
maternal thyroid hormones may cross the placenta. Congenital hypothyroidism presenting with a
goiter can be found in only about 10%-15% of cases (77).
Fetal goiterous hypothyroidism is usually secondary to maternal exposure to thyrostatic agents
such as PTU, radioactive I131, or iodide exposure used to treat maternal hyperthyroidism (78).
Maternal ingestion of amiodarone or lithium may also cause hypothyroidism in the fetus. Finally,
fetal hypothyroidism may follow transplacental passage of maternal blocking antibodies (known as
TBIAb or TBII). Rarely it may be a result of rare defects in fetal thyroid hormone biosynthesis (72).
An enlarged fetal goiter may cause esophageal obstruction, polyhydramnios, leading to preterm
delivery or premature rupture of membranes. Rarely, a goiter may even lead to high-output heart
failure (78). A large fetal goiter can also cause extension of the fetal neck leading to dystocia. The
effects of the fetal hypothyroidism itself may be devastating. Without treatment, postnatal growth
delay and severe mental retardation may ensue. Even with immediate diagnosis and treatment at
birth, long-term follow-up of children with congenital hypothyroidism has demonstrated that they
have lower scores on perceptual-motor, visuospatial, and language tests (79).
In suspicious cases, an extensive maternal and family history should be obtained. In patients with
a positive history, maternal thyroid hormone levels, and blocking immunoglobulin levels should be
measured. Additionally, all women with a history of any thyroid disease (both hypothyroidism and
hyperthyroidism) are advised to have monthly fetal ultrasound scans to screen for fetal goiter,
polyhydramnios, or fetal tachycardia (79).
Occasionally, fetal goiterous hypothyroidism may be identified by a routine ultrasound performed
as a result of increased uterine size caused by polyhydramnios secondary to esophageal
obstruction and impaired swallowing. Sometimes, a fetal goiter may incidentally be discovered on
a routine scan. Before the advent of cordocentesis, amniotic fluid levels of TSH and free thyroxine
were used as potential indicators of fetal thyroid function. However, these proved to be
inconsistent (80). With cordocentesis, fetal thyroid status can be directly and accurately evaluated;
fetal response to therapy can therefore be reliably measured using available appropriate
nomograms for fetal serum levels of free T4, total T4, free T3, total T3, and TSH (81,82). In utero
treatment was initially suggested by Van Herle and associates using i.m. injection of levothyroxine
sodium (83). Subsequent studies however, have indicated that intraamniotic (IA) administration of
thyroxine may be superior and can lead to resolution of the polyhydramnios as well. The dose of
the
P.194
injected drug may be refined using the fetal thyroid profile in the amniotic fluid and the thyroid
size (84). The doses commonly used for treatment range from 200-500 mg IA every week (23).
With this regimen, fetal goiters have been shown to regress, the hyperextension of the fetal head
have been shown to resolve and fetal and newborn TSH levels have normalized (84).
INBORN ERRORS OF METABOLISM

Methylmalonic Acidemia
The methylmalonic acidemias (MMA) are a group of autosomal recessive enzyme-deficiency
diseases. Some cases are caused by mutations in the gene encoding methylmalonyl-coenzyme-A
mutase. Others are a result of a defect that reduces the biosynthesis of adenosylcobalamin from
vitamin B12. The disease has a wide clinical spectrum ranging from a benign condition to a fatal
neonatal disease characterized by severe metabolic acidosis, developmental delay and biochemical
abnormalities that include methylmalonic aciduria, long chain ketonuria, and intermittent
hyperglycinemia. Neurological abnormalities may result from diminution of myelin content and of
ganglioside N-acetylneuraminic acid in the cerebrum (85). Patients with defects in
adenosylcobalamin biosynthesis may respond to administration of large doses of B12, which may
enhance the amount of active holoenzyme (mutase apoenzyme plus adenosylcobalamin).
Ampola and associates first attempted prenatal diagnosis and treatment of a B12-responsive
variant of MMA (86). Their patient previously had a child who died of severe MMA. In the
subsequent pregnancy, amniocentesis at 19 weeks revealed an elevated methylmalonic acid level.
Cultured amniocytes also demonstrated defective propionate oxidation and undetectable levels of
adenosylcobalamin. When adenosylcobalamin was added, normal succinate oxidation and
methylmalonyl-coenzyme-A mutase activity were noted, confirming that the fetus also suffered
from MMA. Late in the pregnancy, cyanocobalamin (10 mg/day) was orally administered to the
mother, only marginally altering maternal serum B12 level. They noted a slight reduction of
maternal urinary methylmalonic acid excretion that remained several-fold above normal. At 34
weeks gestation, 5 mg of cyanocobalamin per day was administered intramuscularly. The maternal
serum B12, level then rose gradually to more than sixfold above normal and was accompanied by
a progressive decrease in urinary methylmalonic acid excretion. Maternal urinary methylmalonate
was only slightly above the normal range when delivery occurred at 41 weeks. Amniotic fluid
methylmalonic acid concentrations were three times the normal mean at 19 menstrual weeks and
four times the normal mean at term, despite prenatal treatment. Postnatally, the diagnosis of
methylmalonic acidemia was confirmed. The infant suffered no acute neonatal complications and
had an extremely high serum B12 level. Long-term postnatal management involved protein
restriction but continuous B12 treatment was not required.
It is not clear whether in utero treatment actually resulted in an improved outcome but it is likely
that correction of the biochemical abnormality in the fetus had some beneficial effect on fetal
development. In a cohort of 8 children with MMA, Andersson and associates described congenital
malformations, probably caused by prenatally abnormal cyanocobalamin metabolism (87). Growth
was significantly improved in most cases after initiation of therapy postnatally and in one case
microcephaly resolved. However, developmental delay of variable severity was always present
regardless of treatment onset. These data suggest that prenatal therapy of MMA may be effective
and perhaps ameliorate some of the prenatal effects. Evans and associates have documented the
changing dose requirements necessary over the course of pregnancy to maintain adequate levels
of B12. They sequentially followed maternal plasma and urine levels in a prenatal treated
pregnancy (88).
Multiple Carboxylase Deficiency

Biotin-responsive multiple carboxylase deficiency is an inborn error of metabolism caused by
diminished activity of the mitochondrial biotin-dependent enzymes (pyruvate carboxylase,
propionyl-coenzyme-A carboxylase and 3-methylcrotonyl-coenzyme-A carboxylase). The condition
may arise from mutations in the holocarboxylase synthetase gene (HCS) or the biotinidase gene
(89,90,91,92,93). Affected patients present with dermatitis, severe metabolic acidosis, and a
characteristic pattern of organic acid excretion. Normal metabolism can be restored toward normal
levels by biotin supplementation. Prenatal diagnosis is based on the demonstration of elevated
levels of typical organic acids (3-hydroxyisovalerate, methylcitrate) in the amniotic fluid or in the
chorionic villi. However, the existence of a mild form of HCS deficiency can complicate prenatal
diagnosis as organic acids level in amniotic fluid might be normal (94). Therefore, prenatal
diagnosis must be performed by enzyme assay in cultured fetal cells in biotin-restricted medium.
Roth and colleagues treated a fetus whose two previous siblings died of multiple carboxylase
deficiency (95). The mother was first seen at 34 weeks gestation and prenatal diagnosis was not
performed. Oral administration was begun at a dose of 10 mg/day of biotin. There were no
apparent untoward effects. Maternal urinary biotin excretion increased by 100. Nonidentical twins
were delivered at term. Cord blood and urinary organic acid profiles were normal, and cord blood
biotin concentrations were 4-7 times greater than normal. The neonatal course for both twins was
unremarkable. Subsequent study of the cultured fibroblasts of both twins indicated that the cells of
one twin had complete deficiency of carboxylase activities. This was later confirmed by genetic
studies. Other cases of such treatment were also reported (94,96,97).
P.195
Smith-Lemli-Optiz Syndrome
Smith-Lemli-Optiz syndrome (SLOS) is an autosomal recessive disorder characterized by multiple
anomalies, dysmorphic features, growth and mental retardation. Males with SLOS frequently have
ambiguous genitalia (98). The severe form is associated with a high rate of neonatal mortality
(99). SLOS is estimated to occur in 1:20,000-40,000 live births with an estimated carrier
frequency of 1:70 (100,101). SLOS is caused by an inborn error of cholesterol biosynthesis as a
result of a deficiency of the enzyme dehydrocholesterol-7 reductase leading to reduced
cholesterol levels and elevated 7- and 8- dehydrocholesterol levels (7-DHC and 8-DHC
respectively) in all body fluids and tissues including amniotic fluid and chorionic villi
(100,102,103,104). The diagnosis is based on elevated levels of 7-DHC (100-1,000 times the
normal value). Clinical manifestations correlate with cholesterol levels. Prenatal diagnosis of SLOS
has been available since 1994 by either amniocentesis or chorionic villus sampling (105,106,107).
Since the identification of the cholesterol metabolic defect in SLOS, a treatment protocol has been
attempted providing exogenous cholesterol. This form of therapy has now been provided to many
patients with SLOS for the past several years in many centers in the United States and
internationally (108,109,110), with the goal of raising cholesterol levels and decrease the
precursors, 7-DHC and 8-DHC. It has been shown that dietary cholesterol supplementation can
restore a normal growth pattern in children and adolescents with SLOS, alleviate behavioral
abnormalities, and improve general health (108,109,110). Since significant development of the
central nervous system and myelination occurs prior to birth it is reasonable to assume that
providing cholesterol to the fetus, as early as possible would result in the most clinical benefit.
However, providing cholesterol to the mother is of no use because cholesterol does not cross the
placenta well in the second trimester and there is lack of evidence that it crosses the placenta in
the third trimester. It is impractical to inject cholesterol into the amniotic fluid because it would
precipitate. However, cholesterol can be given to the fetus by giving fresh frozen plasma in the
form of low-density lipoprotein-cholesterol. A group at Tufts University has attempted treatment
antenatally in several affected fetuses. In cases in which treatment was started late in pregnancy,
the results were inconclusive. Although few descriptions of fetal therapy for SLOS exist, the latest
report of antenatal treatment comes from that same group of investigators (111). Therapy was
begun at 34 weeks of gestation and resulted in increased fetal cholesterol levels and red blood cell
mean corpuscular volume with subtle improvement in fetal growth. However, no significant change
in 7-DHC and 8-DHC levels was observed, further emphasizing the inconclusiveness of that
treatment.
Galactosemia
Galactosemia is an autosomal recessive disorder caused by decreased activity of galactose-1phosphate uridyltransferase (GALT). Clinical manifestations include cataracts, growth deficiency,
and ovarian failure. Clinical symptoms appear in the neonatal period and can be largely
ameliorated by elimination of galactose from the diet. Cellular damage in galactosemia is thought
to be mediated by accumulation of galactose-1-phosphate intracellularly and of galactitol in the
lens. Several disease-causing mutations in the GALT gene have been reported in classical
galactosemia (112). Galactosemia can also be diagnosed prenatally by biochemical studies of
cultured amniocytes and chorionic villi.
There are suggestions that even the early postnatal treatment of galactosemic individuals with a
low galactose diet may not be sufficient to ensure normal development. Some have speculated
that prenatal damage to galactosemic fetuses could contribute to subsequent abnormal
neurological development and to lens cataract formation. Furthermore, it has been recognized that
female galactosemics, even when treated from birth with galactose deprivation, have a high
frequency of primary or secondary amenorrhea because of ovarian failure. This is because oocytes
have already been damaged irreversibly long before birth (113,114). There also may be some
subtle abnormalities of male gonadal function as well. Thus, galactose restriction during pregnancy
may be beneficial in affected fetuses. In humans, ovarian meiosis begins at 12 weeks and ovarian
damage may occur prior to prenatal diagnosis. Thus, anticipatory treatment in pregnancies at risk
for having a galactosemic fetus might best be initiated very early in gestation or even
preconceptually. There are no studies that adequately assess the impact of prenatal administration
of a low-galactose diet to galactosemic infants. Nevertheless, prenatal galactose restriction is
probably desirable in galactosemia and should be harmless.
PRENATAL HEMATOPOIETIC STEM CELL

TRANSPLANTATION
The engraftment and clonal proliferation of a relatively small number of normal hematopoietic
stem cells (HSCs) can sustain normal hematopoiesis for a lifetime. This observation provides the
compelling rationale for bone marrow transplantation (BMT) and is now supported by thousands of
long-term survivors of BMT who otherwise would have succumbed to lethal hematological disease
(115,116). Realization of the full potential of BMT, however, continues to be limited by a critical
shortage of immunologically compatible donor cells, the inability to control the recipient or donor
immune response, and the requirement for recipient myeloablation to achieve engraftment. The
price of HLA mismatch remains high: the greater the mismatch, the higher the incidence of graft
failure, graft-versus-host disease (GVHD), and delayed immunological reconstitution. Current
methods of myeloablation have high morbidity and mortality. In combination, these problems
remain prohibitive for most patients who might benefit from BMT. A theoretically attractive
alternative, which potentially can address many of the limitations of
P.196
BMT, is in utero transplantation of HSC. This approach is potentially applicable to any congenital
hematopoietic disease that can be diagnosed prenatally and can be cured or improved by
engraftment of normal HSCs.
Rationale for In Utero Transplantation

The rationale for in utero transplantation is to take advantage of the window of opportunity
created by normal ontogeny of hematopoietic cell lines. There is a period, prior to population of the
bone marrow and prior to thymic processing of self-antigen, when the fetus theoretically should be
receptive to engraftment of foreign HSC without rejection and without the need for myeloablation.
In the human fetus, the ideal window would appear to be prior to 14 weeks' gestation, before
release of differentiated T lymphocytes into the circulation and although the bone marrow is just
beginning to develop sites for hematopoiesis (117). It certainly may extend beyond that in
immunodeficiency states, particularly when T-cell development is abnormal. During this time,
presentation of foreign antigen by thymic dendritic cells theoretically should result in clonal
deletion of reactive T cells during the negative selection phase of thymic processing. Recent
advances in prenatal diagnosis have made possible the diagnosis of a large number of congenital
hematological diseases during the first trimester. Technical advances in fetal intervention make
transplantation feasible by 12 to 14 weeks' gestation. The ontological window of opportunity falls
well within these diagnostic and technical constraints, making application of this approach a
realistic possibility.
Because of the unique fetal environment, prenatal HSC transplantation could theoretically avoid
many of the current limitations of postnatal BMT. There would be no requirement for HLA
matching, resulting in expansion of the donor pool. Transplanted cells would not be rejected, and
space would be available in the bone marrow, eliminating the need for toxic immunosuppressive
and myeloablative drugs. The mother's uterus may ultimately prove to be the ultimate sterile
isolation chamber, eliminating the high risk and costly 2 to 4 months of isolation required after
postnatal BMT and prior to immunological reconstitution. Finally, prenatal transplantation would
preempt the clinical manifestations of the disease, avoiding the recurrent infections, multiple
transfusions, growth delay, and other complications that cause immeasurable suffering for the
patient and often compromise postnatal treatment.
Source of Donor Cells

Identifying the best source of donor cells may ultimately prove to be the most critical factor for the
success of engraftment. The most obvious advantage of the use of fetal hematopoietic stem cells
is the minimal number of mature T cells in fetal liver-derived populations prior to 14 weeks'
gestation. This alleviates any concern about GVHD and avoids the necessity of T cell depletion
processes, which can negatively impact potential engraftment (115,116,117,118,119).
Although there may be important homing, proliferative, and developmental advantages to the use
of fetal cells, there are practical and ethical advantages to the use of cord blood or postnatal HSC
sources. Legitimate ethical concerns regarding the use of fetal tissue for transplantation must be
addressed by the medical and lay community. Fetal tissue obtained by the usual methods at the
time of elective abortion has a high degree of microbial contamination (120). The transplantation
of transmissible viral, fungal, or bacterial disease could have disastrous consequences for the
recipient fetus or mother. Although the fetal liver is a rich source of HSC, small size limits total cell
yield, and current technology does not yet allow undifferentiated expansion of donor cells. In
contrast, the use of adult-derived cells would allow a renewable, relatively infection-free, ethically
acceptable source of donor cells. Tolerance induction by the in utero transplantation of highly
purified adult bone marrow HSC from a living related donor, followed by a single or multiple
postnatal booster injections offers an intriguing approach in situations in which postnatal BMT is
necessary but an HLA-matched donor is not available (121,122,123).
Diseases Amenable to Prenatal Treatment

Generally speaking, any disease that can be diagnosed early in gestation, which is improved by
BMT, and for which postnatal treatment is not entirely satisfactory, is a target disease. Some
diseases, however, are far more likely to benefit from prenatal transplantation than others. The list
can be divided into three general categories: hemoglobinopathies, immunodeficiency disorders,
and inborn errors of metabolism. Each of the diseases has unique considerations for treatment,
and in fact each disease may respond differently (Table 13-2). Issues such as availability of
engraftment sites within the bone marrow at time of transplantation, and capacity of a needed
enzyme to cross the blood-brain barrier at a particular gestational age need to be considered. Of
particular relevance to the prenatal approach, in which experimental levels of engraftment have
been relatively low, is the observation that in many of the target diseases, engrafted normal cells
would be predicted to have a significant survival advantage over diseased cells. This would have
the clinical effect of amplification of the level of engraftment in the peripheral circulation.
Additionally, even with minimal levels of engraftment, specific tolerance for donor antigen should
be induced, allowing additional cells from the same donor to be given to the tolerant recipient after
birth.
Hemoglobinopathies
The sickle cell anemia and thalassemia syndromes make up the largest patient groups potentially
treatable by prenatal stem cell transplantation (122,123,124,125). Both groups can be diagnosed
within the first trimester. Both have been cured by postnatal BMT, but BMT is not recommended
routinely because of its prohibitive morbidity and mortality,
P.197
and the relative success of modern medical management. In both diseases the success of BMT is
indirectly related to the morbidity of the disease, that is, the younger the patient, the fewer
transfusions received, and the less organ compromise from iron overload, the better the results.
With both diseases the primary questions relevant to prenatal transplantation are: (a) what levels
of normal peripheral cell expression are necessary to alleviate clinical disease, and (b) can
adequate levels of donor cell engraftment be achieved by in utero HSC transplantation? At present
only indirect evidence exists to answer these questions.
TABLE 13-2 POTENTIAL CANDIDATES FOR IN UTERO STEM CELL FETAL THERAPY
Hematopoietic Disorders
Disorders affecting lymphocytes

SCID (sex linked)
SCID (adenosine deaminase deficiency)
Ommen syndrome
Agammaglobinemia
Bare Lymphocyte syndrome
Disorders affecting granulocytes
Chronic granulomatous disease
Infantile agranulocytosis
Neutrophil membrane GP-180
Lazy leukocyte syndrome
Disorders affecting erythrocytes
Sickle cell disease
-Thalassemia
-Thalassemia
Hereditary spherocytosis
Fanconi anemia
Mannosidosis
-Mannosidosis
-Thalassemia
Mucolipidoses
Gaucher disease
Metachromatic leukodystrophy
Krabbe disease
Niemann-Pick disease
-Glucoronidase deficiency
Fabry disease
Adrenal leukodystrophy
Mucopolysaccharidoses
MPS I
(Hurler Disease)
MPS
MPS
MPS
MPS
II
IIIB
IV
VI
(Hunter Disease)
(Sanfilippo B)
(Morquio)
(Maroteaux-Lamy)
In sickle cell disease (SCD) the pathophysiology is directly related to the concentration of
hemoglobin S (HbS) within red cells, which results in marked rheological abnormality, including
hyperviscosity, cellular adherence, and sickling, with a result of vaso-occlusion and tissue
ischemia. In examining the in vitro relationships between hematocrit (HCT) and viscosity using
mixtures of sickle and normal red blood cells (RBCs), Schmalzer observed that the primary
determinant of viscosity is the sickle HCT (fraction of RBCs that contain HbS) (126). Adverse
effects of HCT on viscosity were seen at a sickle HCT level in the low twenties. Oxygen delivery, as
gauged by the maximal point on the HCT vs. viscosity curve, was markedly improved by
exchanging normal for sickle RBCs (even when the total HCT was held constant). The clinical
correlate of this in vitro information presently is chronic exchange transfusion therapy with its
inherent discomfort and potential complications. However, prenatal transplantation of normal
hematopoietic stem cells was thought to offer a solution to this problem. Normal cells were
expected to have a developmental advantage over HbS cells such that overall production of Hb-
normal RBCs may exceed that of HbS cells. This could potentially decrease the overall HbS
fraction, reducing the risk of hyperviscosity and vaso-occlusive/ischemic complications. However,
experimental data have not yet born this out.
The clinical manifestations of thalassemia are secondary to hypoxia related to severe anemia and
ineffective erythropoiesis. It is now standard therapy to transfuse patients with thalassemia major
chronically from an early age, which suppresses endogenous erythropoiesis and maintains oxygen
delivery. When instituted at an early age this effectively prevents the bone marrow expansion and
secondary bony changes, and the hemodynamic and cardiac manifestations of the disease. The
necessary normal hemoglobin (Hb) level required is controversial, but good results have been
achieved with maintenance of an Hb of 9 g/dL.
Although these levels of normal Hb are higher than have been achieved experimentally (30%
donor Hb is maximal), there would be a significant survival advantage of normal cells in both
diseases. In SCD, erythrocytes have a circulating half-life of 10 to 20 days (normal half-life = 120
days) prior to destruction. In thalassemia, most cells (80%) never leave the bone marrow and also
have shortened survival in the periphery. Therefore, engraftment of even a relatively small
number of normal stem cells could result in significantly increased levels of peripheral donor cell
expression. The problem, however, is creating spaces in the bone marrow for engraftment, which
is an unsolved dilemma.
Immunodeficiency Diseases
These represent an extremely heterogeneous group of diseases, which differ in their likelihood of
cure by their capacity to develop hematopoietic chimerism (115,116). Once again, the most likely
to benefit from even low levels of donor cell engraftment are those diseases in which a survival
advantage exists for normal cells. The best example of this situation is severe combined
immunodeficiency syndrome (SCID). Several different molecular causes of SCID have been
identified, with approximately two-thirds of cases being of X-linked recessive inheritance (X-SCID).
The genetic basis of X-SCID has been defined recently (127) as a mutation of the gene encoding
the common -Y chain (- c), which is a common component of several members of the cytokine
receptor superfamily, including those for interleukin-2 (IL-2), IL-4, IL-7, IL-9, IL-15, and possibly
IL-13. Children affected with X-SCID have simultaneous disruption of multiple cytokine systems,
resulting in
P.198
a block in thymic T-cell development and diminished T-cell response. B cells, although present in
normal or even increased numbers, are dysfunctional, either secondary to the lack of helper T-cell
function or an intrinsic defect in B-cell maturation. Another form of SCID is secondary to adenosine
deaminase (ADA) deficiency. Clinical experience with HLA-matched sibling bone marrow, fetal liver
or thymus transplantation generally has been successful without myeloablative therapy,
suggesting that the lymphoid progeny of relatively few engrafted normal HSC have a selective
growth advantage in vivo over genetically defective cells (128). The competitive advantage of
normal cell populations in X-SCID is best supported by the discovery of skewed cross-inactivation
in female carriers (129). Only T cells containing the normal X chromosome were found to be
present in the circulation of recipients. Evidence that ADA production confers a survival advantage
derives from the early experience with gene therapy for ADA deficiency SCID. ADA-gene-corrected
autologous T cells have persisted for prolonged periods despite discontinuation of the T-cell
infusions (130). Transfer of ADA-gene-corrected cells versus uncorrected cells from the same SCID
patient into immunodeficient beige, nude, x-linked (BNX) mouse results in survival of the corrected
cells and death of the uncorrected cells, confirming a survival advantage for ADA-producing cells
even when there is normal ADA production in the surrounding environment. Unfortunately, other
diseases such as chronic granulomatous disease would not be expected to provide a competitive
advantage for donor cells. Nevertheless, in all these conditions even a partial engraftment and
expression of normal cell phenotype might at least partially ameliorate the clinical manifestations
of the disease and should result in donor-specific tolerance for later transplantation. If higher
levels of engraftment are needed, further HSC transplants from the same donor could be
performed after birth without fear of rejection.
Flake and associates reported the successful treatment of a fetus with X-SCID in a family in which
a previously afflicted child died at 7 months of age (131). Diagnosis by chorionic villous sampling
at 12 weeks in the second pregnancy showed another affected male. For this couple, abortion was
not an option. After lengthy informed consent, paternal bone marrow was harvested, T cells
depleted, and enriched stem cell populations injected intraperitoneally into the fetus beginning
about 16 weeks of gestation. Subsequent injections were performed at 17 and 18 weeks. The baby
presently shows a split chimerism with all of his T cells being his father's and the majority of B
cells being his own. He has achieved developmentally normal milestones and immune progress
through 8 years of age (123). Other cases have been recently tried using less T cell-depleted
populations resulting in higher T cell concentrations that have ended in fetal demise
(117,132,133,134).
Inborn Errors of Metabolism

An even more heterogeneous group of diseases, inborn errors of metabolism, can be caused by a
deficiency of a specific lysosomal hydrolase, which results in the accumulation of substrates such
as mucopolysaccharide, glycogen, or sphingolipid. Depending on the specific enzyme abnormality
and the compounds that accumulate, certain patterns of tissue damage and organ failure occur
(116). These include central nervous system (CNS) deterioration, growth failure, dysostosis
multiplex and joint abnormalities, hepatosplenomegaly, myocardial or cardiac disease, upper
airway obstruction, pulmonary infiltration, corneal clouding, and hearing loss. The potential
efficacy of prenatal HSC transplantation for the treatment of these diseases must be considered on
an individual disease basis. The purpose of BMT in these diseases is to provide HSC-derived
mononuclear cells that can repopulate various organs in the body, including the liver (Kupffer
cells), skin (Langerhans' cells), lung (alveolar macrophages), spleen (macrophages), lymph nodes,
tonsils, and the brain (microglia). Patients who have been corrected by postnatal BMT, such as
Gaucher's disease or Maroteaux-Lamy syndrome (minimal CNS involvement), are reasonable
candidates for prenatal treatment. In many cases postnatal BMT has corrected the peripheral
manifestations of the disease and has arrested the neurological deterioration. However, postnatal
BMT has not reversed neurological injury that is present in such disorders as metachromatic
leukodystrophy and Hurler's disease (115,116). In these cases the neurological injury may begin
well before birth. Postnatal maturation of the blood-brain barrier restricts access to the CNS of
transplanted cells or the deficient enzyme. These considerations suggest that prenatal treatment
may be beneficial and offer the possibility for a cure. The primary unanswered question is whether
donor HSC-derived microglial elements would populate the CNS, providing the necessary metabolic
correction inside the blood-brain barrier.
To date, the only definitively successful transplants have been for SCID (115,116). All others have
either failed to take or were afflicted with graft versus host disease. Much more work needs to be
done to determine the optimal dosing, timing, and specific protocols for different conditions.
As reviewed recently by Shields and associates (116), as of 2002 there have been 39 published
cases of in utero stem cell transplants. Since the first successful case performed in 1995, only
immunodeficiency cases continued to be successful. Of 14 cases of and -thalassemia, none
have been successful. Three sickle cell, 3 Rh isoimmunization, and 7 storage disease cases have
likewise been unsuccessful (116). Of 12 immunodeficiency cases, 6 have shown at least split
chimerism or alive and well. The dosages used, timing, fresh vs. frozen tissue, and methods of
evaluation before and after transplantation have been widely varied. While no definitive
conclusions are possible at this time, it is clear that our lack of understanding of the mechanics or
development of immunocompetency are persuasive. Until a better understanding of the processes
obtained, better theories on treatment to achieve better outcomes may be limited.
P.199
CONCLUSION
There are an increasing number of congenital and genetic abnormalities for which in utero
treatment is possible, and are in some cases, relatively routine now. Advances in therapies have
progressed at different paces for different disorders, but there is great hope and enthusiasm that
progress will continue to expand the number of disorders for which therapy can be effective (60).
References
1. Reproductive Risks and Prenatal Diagnosis. Norwalk, CT: Appleton & Lange, 1992.
2. Harrison MR, Evans MI, Adzick NS, et al. The Unborn Patient. Philadelphia: WB Saunders,
2000.
3. Pryde PG, Isada NB, Hallak M, et al. Determinants of parental decision to abort or continue
after non-aneuploid ultrasound-detected fetal abnormalities. Obstet Gynecol 1992;80(1)1:52-56.
4. Evans MI, Sobiecki MA, Krivchenia EL, et al. Parental decisions to terminate/continue
following abnormal cytogenetic prenatal diagnosis: what is still more important than when.
Am J Med Genet 1996;61(4):353-355.
5. Johnson MP, Flake AW, Quintero RA, et al. Invasive outpatient procedures in reproductive
medicine. New York: Raven Press, 1997.
6. Evans MI, Sacks AJ, Johnson MP, et al. Sequential invasive assessment of fetal renal function
and the intrauterine treatment of fetal obstructive uropathies. Obstet Gynecol 1991;77(4):545550.
7. Wilson RD, Johnson MP. Prenatal ultrasound guided percutaneous shunts for obstructive
uropathy and thoracic disease. Semin Pediatr Surg 2003;12(3):182-189.
8. Johnson MP, Corsi P, Bradfield W, et al. Sequential urinalysis improves evaluation of fetal
renal function in obstructive uropathy. Am J Obstet Gynecol 1995;173(1):59-65.
9. Freedman AL, Bukowski TP, Smith CA, et al. Fetal therapy for obstructive uropathy: diagnosis
specific outcomes [corrected]. J Urol 1996;156(2 Pt 2):720-723, discussion 723-724.
10. Dommergues M, Meuller F. Experience of 100 Obstructive Uropathies. International Fetal

Medicine and Surgery Society Meeting, 2003.
11. Feldman B, Aviram-Goldring A, Evans MI. Interphase FISH for prenatal diagnosis of common
aneuploidies. Methods Mol Biol 2002;204:219-241.
12. Evans MI, Henry GP, Miller WA, et al. International, collaborative assessment of 146,000
prenatal karyotypes: expected limitations if only chromosome-specific probes and fluorescent insitu hybridization are used. Hum Reprod 1999;14(5):1213-1216.
13. Drugan A, Krause B, Canady A, et al. The natural history of prenatally diagnosed cerebral
ventriculomegaly. JAMA 1989; 261(12):1785-1788.
14. Ahmad FK, Sherman SJ, Hagglund KH, et al. Isolated unilateral fetal pleural effusion: the
role of sonographic surveillance and in utero therapy. Fetal Diagn Ther 1996;11(6):383-389.
15. Nicolaides KH, Azar GB. Thoraco-amniotic shunting. Fetal Diagn Ther 1990;5:153-164.
16. Adzick NS, Harrison MR, Flake AW, et al. Automatic uterine stapling devices in fetal surgery:
experience in a primate model. Surgical Forum 1985;34:479.
17. Jennings RW, Adzick NS, Longaker MT, et al. Radiotelemetric fetal monitoring during and
after open fetal operation. Surg Gynecol Obstet 1993;176(1):59-64.
18. Harrison MR, Adzick NS, Longaker MT, et al. Successful repair in utero of a fetal
diaphragmatic hernia after removal of herniated viscera from the left thorax. N Engl J Med
1990;322(22): 1582-1584.
19. Harrison MR, Adzick NS, Flake AW, et al. Correction of congenital diaphragmatic hernia in
utero. VI. Hard-earned lessons. J Pediatr Surg 1993;28(10):1411-1417, discussion 1417-1418.
20. Harrison MR, Keller RL, Hawgood SB, et al. A randomized trial of fetal endoscopic tracheal
occlusion for severe fetal congenital diaphragmatic hernia. N Engl J Med 2003;349(20):19161924.
21. Harrison MR, Sydorak RM, Farrell JA, et al. Fetoscopic temporary tracheal occlusion for
congenital diaphragmatic hernia: prelude to a randomized, controlled trial. J Pediatr Surg
2003;38(7): 1012-1020.
22. Heerema AE, Rabban JT, Sydorak RM, et al. Lung pathology in patients with congenital
diaphragmatic hernia treated with fetal surgical intervention, including tracheal occlusion.
Pediatr Dev Pathol 2003;6(6):536-546.
23. Sydorak RM, Harrison MR. Congenital diaphragmatic hernia: advances in prenatal therapy.
Clin Perinatol 2003;30(3):465- 479.
24. Danzer E, Sydorak RM, Harrison MR, et al. Minimal access fetal surgery. Eur J Obstet
Gynecol Reprod Biol 2003;108(1):3-13.
25. Evans MI, Harrison MR, Flake AW, et al. Fetal therapy. Best Pract Res Clin Obstet Gynaecol
2002;16(5):671-683.
26. Paek BW, Coakley FV, Lu Y, et al. Congenital diaphragmatic hernia: prenatal evaluation with
MR lung volumetrypreliminary experience. Radiology 2001;220(1):63-67.
27. Wenstrom KD. Fetal surgery for congenital diaphragmatic hernia. N Engl J Med 2003;349
(20):1887-1888.
28. Adzick NS, Kitano Y. Fetal surgery for lung lesions, congenital diaphragmatic hernia, and
sacrococcygeal teratoma. Semin Pediatr Surg 2003;12(3):154-167.
29. Adzick NS, Harrison MR, Crombleholme TM, et al. Fetal lung lesions: management and
outcome. Am J Obstet Gynecol 1998; 179(4):884-889.
30. Crombleholme TM, Coleman B, Hedrick H, et al. Cystic adenomatoid malformation volume
ratio predicts outcome in prenatally diagnosed cystic adenomatoid malformation of the lung. J
Pediatr Surg 2002;37(3):331-338.
31. Holterman AX, Filiatrault D, Lallier M, et al. The natural history of sacrococcygeal teratomas
diagnosed through routine obstetric sonogram: a single institution experience. J Pediatr Surg
1998;33(6):899-903.
32. Paek BW, Vaezy S, Fujimoto V, et al. Tissue ablation using high-intensity focused ultrasound
in the fetal sheep model: potential for fetal treatment. Am J Obstet Gynecol 2003;189(3)(Sep):
702- 705.
33. Lemire RJ. Neural tube defects. JAMA 1988;259(4):558-562.
34. Frey L, Hauser WA. Epidemiology of neural tube defects. Epilepsia 2003;44[Suppl 3]:4-13.
35. Smithells RW, Sheppard S, Schorah CJ, et al. Possible prevention of neural-tube defects by
periconceptional vitamin supplementation. Lancet 1980;1(8164):339-340.
36. Smithells RW, Nevin NC, Seller MJ, et al. Further experience of vitamin supplementation for
prevention of neural tube defect recurrences. Lancet 1983;1(8332):1027-1031.
37. Younis JS, Granat M. Insufficient transplacental digoxin transfer in severe hydrops fetalis.
Am J Obstet Gynecol 1987;157(5): 1268-1269.
38. Mills JL, Rhoads GG, Simpson JL, et al. The absence of a relation between the
periconceptional use of vitamins and neural-tube defects. National Institute of Child Health and
Human Development Neural Tube Defects Study Group. N Engl J Med 1989;321(7):430-435.
39. Mulinare J, Cordero JF, Erickson JD, et al. Periconceptional use of multivitamins and the
occurrence of neural tube defects. JAMA 1988;260(21):3141-3145.
40. Prevention of neural tube defects: results of the Medical Research Council Vitamin Study.
MRC Vitamin Study Research Group. Lancet 1991;338(8760):131-137.
41. Czeizel AE, Dudas I. Prevention of the first occurrence of neural-tube defects by
periconceptional vitamin supplementation. N Engl J Med 1992;327(26):1832-1835.
42. Honein MA, Paulozzi LJ, Mathews TJ, et al. Impact of folic acid fortification of the US. food
supply on the occurrence of neural tube defects. JAMA 2001;285(23):2981-2986.
43. Olney RS, Mulinare J. Trends in neural tube defect prevalence, folic acid fortification, and
vitamin supplement use. Semin Perinatol 2002;26(4):277-285.
44. Evans MI, Llurba E, Landsberger EJ, et al. Impact of folic acid fortification in the United
States: markedly diminished high maternal serum alpha-fetoprotein values. Obstet Gynecol
2004; 103(3):474-479.
45. Parle-McDermott A, Mills JL, Kirke PN, et al. Analysis of the MTHFR 1298A C and 677C T
polymorphisms as risk factors for neural tube defects. J Hum Genet 2003;48(4):190-193.
P.200
46. Finnell RH, Shaw GM, Lammer EJ, et al. Does prenatal screening for 5,10methylenetetrahydrofolate reductase (MTHFR) mutations in high-risk neural tube defect
pregnancies make sense? Genet Test 2002;6(1):47-52.
47. Rampersaud E, Melvin EC, Siegel D, et al. Updated investigations of the role of
methylenetetrahydrofolate reductase in human neural tube defects. Clin Genet 2003;63(3):210214.
48. Zhu H, Wicker NJ, Shaw GM, et al. Homocysteine remethylation enzyme polymorphisms and
increased risks for neural tube defects. Mol Genet Metab 2003;78(3):216-221.
49. Volcik KA, Shaw GM, Zhu H, et al. Risk factors for neural tube defects: associations between
uncoupling protein 2 polymorphisms and spina bifida. Birth Defects Res Part A Clin Mol Teratol
2003;67(3):158-161.
50. Sever LE. Zinc deficiency in man. Lancet 1973;1(7808):887.
51. McMichael AJ, Dreosti IE, Gibson GT, et al. A prospective study of serial maternal serum zinc
levels and pregnancy outcome. Early Hum Dev 1982;7(1):59-69.
52. Stoll C, Dott B, Alembik Y, et al. Maternal trace elements, vitamin B12, vitamin A, folic acid,
and fetal malformations. Reprod Toxicol 1999;13(1):53-57.
53. Hambidge M, Hackshaw A, Wald N. Neural tube defects and serum zinc. Br J Obstet
Gynaecol 1993;100(8):746-749.
54. Shoob HD, Sargent RG, Thompson SJ, et al. Dietary methionine is involved in the etiology of
neural tube defect-affected pregnancies in humans. J Nutr 2001;131(10):2653-2658.
55. Adzick NS, Walsh DS. Myelomeningocele: prenatal diagnosis, pathophysiology and
management. Semin Pediatr Surg 2003; 12(3):168-174.
56. Evans ML, Holzgreve W, Johnson MP, et al. Fetal cell testing: societal and ethical
speculations. Ann N Y Acad Sci 1994;731: 257-261.
57. Meuli M, Meuli-Simmen C, Hutchins GM, et al. In utero surgery rescues neurological function
at birth in sheep with spina bifida. Nat Med 1995;1(4):342-347.
58. Bruner JP, Tulipan N, Paschall RL, et al. Fetal surgery for myelomeningocele and the
incidence of shunt-dependent hydrocephalus. JAMA 1999;282(19):1819-1825.
59. Sutton LN, Adzick NS, Bilaniuk LT, et al. Improvement in hindbrain herniation demonstrated
by serial fetal magnetic resonance imaging following fetal surgery for myelomeningocele. JAMA
1999;282(19):1826-1831.
60. Tulipan N, Hernanz-Schulman M, Bruner JP. Reduced hindbrain herniation after intrauterine
myelomeningocele repair: a report of four cases. Pediatr Neurosurg 1998;29(5):274-278.
61. Johnson MP, Sutton LN, Rintoul N, et al. Fetal myelomeningocele repair: short-term clinical
outcomes. Am J Obstet Gynecol 2003;189(2):482-487.
62. Liechty KW, Crombleholme TM, Flake AW, et al. Intrapartum airway management for giant
fetal neck masses: the EXIT (ex utero intrapartum treatment) procedure. Am J Obstet Gynecol
1997;177(4):870-874.
63. Hedrick HL. Ex utero intrapartum therapy. Semin Pediatr Surg 2003;12(3):190-195.
64. MacLaughlin DT, Donahoe PK. Sex determination and differentiation. N Engl J Med 2004;350
(4):367-378.
65. Speiser PW, Dupont B, Rubinstein P, et al. High frequency of nonclassical steroid 21hydroxylase deficiency. Am J Hum Genet 1985;37(4):650-667.
66. Dupont B, Oberfield SE, Smithwick EM, et al. Close genetic linkage between HLA and
congenital adrenal hyperplasia (21-hydroxylase deficiency). Lancet 1977;2(8052-828053):1309-1312.
67. White PC, Grossberger D, Onufer BJ, et al. Two genes encoding steroid 21-hydroxylase are
located near the genes encoding the fourth component of complement in man. Proc Natl Acad
Sci U S A 1985;82(4):1089-1093.
68. Evans MI, Chrousos GP, Mann DW, et al. Pharmacologic suppression of the fetal adrenal
gland in utero. Attempted prevention of abnormal external genital masculinization in suspected
congenital adrenal hyperplasia. JAMA 1985;253(7):1015-1020.
69. Forrest M, David M. Prenatal treatment of congenital adrenal hyperplasia due to 21hydroxylase deficiency. Paper presented at 7th International Congress of Endocrinology 1984;
Quebec, Canada.
70. New MI, Carlson A, Obeid J, et al. Prenatal diagnosis for congenital adrenal hyperplasia in
532 pregnancies. J Clin Endocrinol Metab 2001;86(12):5651-5657.
71. Clayton PE, Miller WL, Oberfield SE, et al. Consensus statement on 21-hydroxylase
deficiency from the European Society for Paediatric Endocrinology and the Lawson Wilkins
Pediatric Endocrine Society. Horm Res 2002;58(4):188-195.
72. Fisher DA, Klein AH. Thyroid development and disorders of thyroid function in the newborn.
N Engl J Med 1981;304(12):702- 712.
73. Bruinse HW, Vermeulen-Meiners C, Wit JM. Fetal treatment for thyrotoxicosis in nonthyrotoxic pregnant women. Fetal Ther 1988;3(3):152-157.
74. Porreco RP, Bloch CA. Fetal blood sampling in the management of intrauterine
thyrotoxicosis. Obstet Gynecol 1990;76(3 Pt 2): 509-512.
75. Hatjis CG. Diagnosis and successful treatment of fetal goitrous hyperthyroidism caused by
maternal Graves disease. Obstet Gynecol 1993;81(5 Pt 2):837-839.
76. Fisher DA. Neonatal thyroid disease of women with autoimmune thyroid disease. Thyroid
Today 1986;9:1-7.
77. Volumenie JL, Polak M, Guibourdenche J, et al. Management of fetal thyroid goitres: a report
of 11 cases in a single perinatal unit. Prenat Diagn 2000;20(10):799-806.
78. Morine M, Takeda T, Minekawa R, et al. Antenatal diagnosis and treatment of a case of fetal
goitrous hypothyroidism associated with high-output cardiac failure. Ultrasound Obstet Gynecol
2002;19(5):506-509.
79. Rovet J, Ehrlich R, Sorbara D. Intellectual outcome in children with fetal hypothyroidism. J
Pediatr 1987;110(5):700-704.
80. Sack J, Fisher DA, Hobel CJ, et al. Thyroxine in human amniotic fluid. J Pediatr 1975;87
(3):364-368.
81. Thorpe-Beeston JG, Nicolaides KH, McGregor AM. Fetal thyroid function. Thyroid 1992;2
(3):207-217.
82. Ballabio M, Nicolini U, Jowett T, et al. Maturation of thyroid function in normal human
foetuses. Clin Endocrinol (Oxf) 1989;31(5):565-571.
83. Van Herle AJ, Young RT, Fisher DA, et al. Intra-uterine treatment of a hypothyroid fetus. J
Clin Endocrinol Metab 1975;40(3):474- 477.
84. Gruner C, Kollert A, Wildt L, et al. Intrauterine treatment of fetal goitrous hypothyroidism
controlled by determination of thyroid-stimulating hormone in fetal serum. A case report and
review of the literature. Fetal Diagn Ther 2001;16(1):47-51.
85. Brusque A, Rotta L, Pettenuzzo LF, et al. Chronic postnatal administration of methylmalonic
acid provokes a decrease of myelin content and ganglioside N-acetylneuraminic acid
concentration in cerebrum of young rats. Braz J Med Biol Res 2001;34(2):227-231.
86. Ampola MG, Mahoney MJ, Nakamura E, et al. Prenatal therapy of a patient with vitamin-B12responsive methylmalonic acidemia. N Engl J Med 1975;293(7):313-317.
87. Andersson HC, Marble M, Shapira E. Long-term outcome in treated combined methylmalonic
acidemia and homocystinemia. Genet Med 1999;1(4):146-150.
88. Evans MI, Duquette DA, Rinaldo P, et al. Modulation of B12 dosage and response in fetal
treatment of methylmalonic aciduria (MMA): titration of treatment dose to serum and urine
MMA. Fetal Diagn Ther 1997;12(1):21-23.
89. Leon-Del-Rio A, Leclerc D, Akerman B, et al. Isolation of a cDNA encoding human

holocarboxylase synthetase by functional complementation of a biotin auxotroph of Escherichia
coli. Proc Natl Acad Sci U S A 1995;92(10):4626-4630.
90. Suzuki Y, Aoki Y, Ishida Y, et al. Isolation and characterization of mutations in the human
holocarboxylase synthetase cDNA. Nat Genet 1994;8(2):122-128.
91. Aoki Y, Suzuki Y, Sakamoto O, et al. Molecular analysis of holocarboxylase synthetase

deficiency: a missense mutation and a single base deletion are predominant in Japanese
patients. Biochim Biophys Acta 1995;1272(3):168-174.
92. Dupuis L, Leon-Del-Rio A, Leclerc D, et al. Clustering of mutations in the biotin-binding

region of holocarboxylase synthetase in biotin-responsive multiple carboxylase deficiency. Hum
Mol Genet 1996;5(7):1011-1016.
93. Pomponio RJ, Hymes J, Reynolds TR, et al. Mutations in the human biotinidase gene that
cause profound biotinidase deficiency in symptomatic children: molecular, biochemical, and
clinical analysis. Pediatr Res 1997;42(6):840-848.
P.201
94. Suormala T, Fowler B, Jakobs C, et al. Late-onset holocarboxylase synthetase-deficiency:
pre- and post-natal diagnosis and evaluation of effectiveness of antenatal biotin therapy. Eur J
Pediatr 1998;157(7):570-575.
95. Roth KS, Yang W, Allan L, et al. Prenatal administration of biotin in biotin responsive
multiple carboxylase deficiency. Pediatr Res 1982;16(2):126-129.
96. Packman S, Cowan MJ, Golbus MS, et al. Prenatal treatment of biotin responsive multiple
carboxylase deficiency. Lancet 1982; 1(8287):1435-1438.
97. Thuy LP, Jurecki E, Nemzer L, et al. Prenatal diagnosis of holocarboxylase synthetase
deficiency by assay of the enzyme in chorionic villus material followed by prenatal treatment.
Clin Chim Acta1 999;284(1):59-68.
98. Smith DW, Lemli L, Opitz JM. A Newly Recognized Syndrome of Multiple Congenital
Anomalies. J Pediatr 1964;64:210-217.
99. Curry CJ, Carey JC, Holland JS, et al. Smith-Lemli-Opitz syndrome-type II: multiple
congenital anomalies with male pseudohermaphroditism and frequent early lethality. Am J Med
Genet 1987;26(1):45-57.
100. Opitz JM. RSH/SLO (Smith-Lemli-Opitz) syndrome: historical, genetic, and developmental
considerations. Am J Med Genet 1994;50(4):344-346.
101. Kelley RI. A new face for an old syndrome. Am J Med Genet 1997;68(3):251-256.
102. Kelley RI. Diagnosis of Smith-Lemli-Opitz syndrome by gas chromatography/mass

spectrometry of 7-dehydrocholesterol in plasma, amniotic fluid and cultured skin fibroblasts. Clin
Chim Acta 1995;236(1):45-58.
103. Tint GS, Irons M, Elias ER, et al. Defective cholesterol biosynthesis associated with the
Smith-Lemli-Opitz syndrome. N Engl J Med 1994;330(2):107-113.
104. Waterham HR, Wijburg FA, Hennekam RC, et al. Smith-Lemli-Opitz syndrome is caused by
mutations in the 7-dehydrocholesterol reductase gene. Am J Hum Genet 1998;63(2):329-338.
105. Johnson JA, Aughton DJ, Comstock CH, et al. Prenatal diagnosis of Smith-Lemli-Opitz
syndrome, type II. Am J Med Genet 1994;49(2):240-243.
106. Hobbins JC, Jones OW, Gottesfeld S, et al. Transvaginal ultrasonography and
transabdominal embryoscopy in the first-trimester diagnosis of Smith-Lemli-Opitz syndrome,
type II. Am J Obstet Gynecol 1994;171(2):546-549.
107. Sharp P, Haan E, Fletcher JM, et al. First-trimester diagnosis of Smith-Lemli-Opitz

syndrome. Prenat Diagn 1997;17(4):355-361.
108. Irons M, Elias ER, Tint GS, et al. Abnormal cholesterol metabolism in the Smith-Lemli-Opitz
syndrome: report of clinical and biochemical findings in four patients and treatment in one
patient. Am J Med Genet 1994;50(4):347-352.
109. Irons M, Elias ER, Abuelo D, et al. Treatment of Smith-Lemli-Opitz syndrome: results of a
multicenter trial. Am J Med Genet 1997;68(3):311-314.
110. Nowaczyk MJ, Whelan DT, Heshka TW, et al. Smith-Lemli-Opitz syndrome: a treatable
inherited error of metabolism causing mental retardation. CMAJ 1999;161(2):165-170.
111. Irons MB, Nores J, Stewart TL, et al. Antenatal therapy of Smith-Lemli-Opitz syndrome.
Fetal Diagn Ther 1999;14(3):133-137.
112. Elsas LJ. Prenatal diagnosis of galactose-1-phosphate uridyltransferase (GALT)-deficient

galactosemia. Prenat Diagn 2001; 21(4):302-303.
113. Chen YT, Mattison DR, Feigenbaum L, et al. Reduction in oocyte number following prenatal
exposure to a diet high in galactose. Scienc e 1981;214(4525):1145-1147.
114. Bandyopadhyay S, Chakrabarti J, Banerjee S, et al. Prenatal exposure to high galactose

adversely affects initial gonadal pool of germ cells in rats. Hum Reprod 2003;18(2):276-282.
115. Flake AW. Stem cell and genetic therapies for the fetus. Semin Pediatr Surg 2003;12
(3):202-208.
116. Shields LE, Lindton B, Andrews RG, et al. Fetal hematopoietic stem cell transplantation: a
challenge for the twenty-first century. J Hematother Stem Cell Res 2002;11(4):617-631.
117. Flake AW, Zanjani ED. In utero hematopoietic stem cell transplantation: ontogenic
opportunities and biologic barriers. Blood 1999;94(7):2179-2191.
118. Liechty KW, MacKenzie TC, Shaaban AF, et al. Human mesenchymal stem cells engraft and
demonstrate site-specific differentiation after in utero transplantation in sheep. Nat Med 2000;6
(11):1282-1286.
119. Guidos CJ, Danska JS, Fathman CG, et al. T cell receptor-mediated negative selection of
autoreactive T lymphocyte precursors occurs after commitment to the CD4 or CD8 lineages. J
Exp Med 1990;172(3):835-845.
120. Touraine JL, Raudrant D, Laplace S. Transplantation of hemopoietic cells from the fetal liver
to treat patients with congenital diseases postnatally or prenatally. Transplant Proc 1997;29(121-2): 712-713.
121. Flake AW, Zanjani ED. In utero hematopoietic stem cell transplantation. A status report.
JAMA 1997;278(11):932-937.
122. Hayward A, Ambruso D, Battaglia F, et al. Microchimerism and tolerance following

intrauterine transplantation and transfusion for alpha-thalassemia-1. Fetal Diagn Ther 1998;13
(1):8-14.
123. Flake AW, Zanjani ED. Treatment of severe combined immunodeficiency. N Engl J Med
1999;341(4):291-292.
124. Westgren M, Ringden O, Eik-Nes S, et al. Lack of evidence of permanent engraftment after
in utero fetal stem cell transplantation in congenital hemoglobinopathies. Transplantation
1996;61(8): 1176-1179.
125. Touraine JL, Raudrant D, Royo C, et al. In utero transplantation of hemopoietic stem cells
in humans. Transplant Proc 1991;23(1 Pt 2):1706-1708.
126. Schmalzer EA, Lee JO, Brown AK, et al. Viscosity of mixtures of sickle and normal red cells
at varying hematocrit levels. Implications for transfusion. Transfusion 1987;27(3):228- 233.
127. Noguchi M, Yi H, Rosenblatt HM, et al. Interleukin-2 receptor gamma chain mutation
results in X-linked severe combined immunodeficiency in humans. Cell 1993;73(1):147-157.
128. Buckley RH, Schiff SE, Schiff RI, et al. Haploidentical bone marrow stem cell
transplantation in human severe combined immunodeficiency. Semin Hematol 1993;30(4)[Suppl
4]:92- 101, discussion 102-104.
129. Puck JM, Stewart CC, Nussbaum RL. Maximum-likelihood analysis of human T-cell X
chromosome inactivation patterns: normal women versus carriers of X-linked severe combined
immunodeficiency. Am J Hum Genet 1992;50(4):742-748.
130. Slavin S, Naparstek E, Ziegler M, et al. Clinical application of intrauterine bone marrow
transplantation for treatment of genetic diseases-feasibility studies. Bone Marrow Transplant
1992;9[Suppl 1]:189-190.
131. Flake AW, Roncarolo MG, Puck JM, et al. Treatment of X-linked severe combined
immunodeficiency by in utero transplantation of paternal bone marrow. N Engl J Med 1996;335
(24): 1806-10.
132. Bambach BJ, Moser HW, Blakemore K, et al. Engraftment following in utero bone marrow
transplantation for globoid cell leukodystrophy. Bone Marrow Transplant 1997;19(4)1:399- 402.
133. Peranteau WH, Hayashi S, Kim HB, et al. In utero hematopoietic cell transplantation: what
are the important questions? Fetal Diagn Ther 2004;19(1):9-12.
134. Porta F, Mazzolari E, Zucca S, et al. Prenatal transplant in a fetus affected by Omenn
Syndrome. Bone Marrow Transplant 2000; 25(Suppl):S43.
Chapter 14
The Impact of Maternal Illness on the
Neonate
Helain J. Landy
Progress in obstetric and neonatal care has directly contributed to improvements
in neonatal outcome. The infant mortality rate (deaths in the first year of life per
1,000 live births), an established indicator of a nation's health status and wellbeing, has declined exponentially in the twentieth century with a drop of 46%
since 1980 (1,2). The infant mortality rate, directly related to birth weight, has
declined in spite of the increase in the percentage of low-birth-weight infants (2).
These encouraging data largely reflect neonatal and pediatric advances in
combination with regionalization of perinatal services and delivery of high-risk
mothers in tertiary centers (2,3,4,5).
The mother's well-being during pregnancy has direct relevance for the newborn.
Potential complications such as preterm delivery, growth disturbances
(intrauterine growth restriction [IUGR] or macrosomia), congenital malformations,
or chronic maternal illness may be important factors. This chapter discusses the
impact of maternal illness on fetal development and well-being.
PRETERM DELIVERY
Preterm delivery is responsible for the majority of neonatal deaths and a major
proportion of perinatal morbidity; in 1999, for the first time, prematurity
constituted the leading cause of infant deaths in the first month of life (5,6).
Approximately 11% of all deliveries in the United States occur prior to term (7)
and these children are at higher risks of lifelong problems, including cerebral
palsy, deafness, blindness, learning disabilities, and developmental delay (8).
Encouraging data reveal increasingly effective care for pregnant women and
neonates. This translates into a better prognosis for preterm infants, including
increased survival rates (greater than 80%, compared with 74% in 1988) without
an associated increase in morbidity (9).
Despite intense research efforts and technological advances, data demonstrate a
steady rise in the preterm delivery rate in the United States over the past 20
years. In 1981, the preterm delivery rate was 9.4% and rose to 10.6% in 1990
(2). Recent data show a minor drop in the preterm birth rate from 11.8% in 1999
to 11.6% in 2000 (the latest year for which information exists); this represents
the first such decline since 1992 (2). Compared with other industrialized nations,
the United States ranks poorly (2,10). A number of confounding variables may
explain these facts. These include a racial disparity regarding persistently higher
mortality statistics for black infants, a lack of agreement in the diagnosis of
preterm labor, an overlap in the characteristics of actual and threatened preterm
labor, controversies over the effectiveness of available screening techniques and
therapies for preterm labor, and varying dosages and administration of tocolytic
medications (2).
Preterm Labor and Premature Rupture of

Membranes
A significant proportion of spontaneous preterm deliveries result from preterm
labor and premature rupture of the membranes (PROM). Approximately 70% of
preterm births may be associated with clinical evidence of ruptured membranes or
underlying maternal, obstetric, or fetal conditions (Table 14-1) (10,11,12,13).
Predisposing factors for preterm labor remain largely unknown. Epidemiologic
studies consistently report an association with nonwhite race (African American
relative risk [RR] = 3.3), lower socioeconomic status (RR = 1.83-2.65), low
prepregnancy weight (odds ratio [OR] = 2.72), and maternal age younger than 17
years or older than 40 years (RR = 1.47-1.95) (14). Lack of prenatal care is
strongly associated with higher rates of preterm delivery, low birth weight, and
maternal and infant mortality (15). Other particularly important risk factors
include maternal history of preterm birth, especially during the second trimester,
with or without rupture of membranes, and vaginal bleeding in at least two
trimesters (14,16,17,18). Prostaglandins, thought to be
P.203
important mediators in the onset of labor at term, may have a role in preterm
labor; in the absence of intrauterine infection, however, the data are not
convincing (19).
Although PROM occurs in only 3% of pregnancies before term (20), it is associated
with 20% to 50% of preterm deliveries (21) and represents the foremost
predisposing factor for admissions to neonatal intensive care units (22). The
pathophysiology underlying PROM appears to be multifactorial; evidence suggests
that inflammatory weakening of the membranes or choriodecidual infection may
be involved, especially at early gestational ages (20,23). Some clinical
characteristics seen with PROM are also common to cases with preterm labor (e.
g., lower socioeconomic status, young maternal age, sexually transmitted
diseases, cigarette smoking, and vaginal bleeding) (20). Many studies suggest an
association between bacterial vaginosis, a relatively common condition in which
the normally predominant lactobacilli are overgrown by anaerobes and other
peroxide-producing bacteria, and complications of premature birth, preterm labor,
TABLE 14-1 UNDERLYING CONDITIONS ASSOCIATED WITH PRETERM

BIRTHa
Maternal Conditions
Anatomic conditions
Fetal Conditions
Fetal growth restriction
Uterine malformations
Unicornuate or bicornuate uterus
Multiple gestation
Fetal anomaly
Myomas
Cervical incompetence
Placenta previa or abruption
Fetal death
Ruptured membranes
Medical conditions
Systemic medical or obstetric illness
Trauma
Exogenous substance use
Tobacco
Cocaine
Maternal in utero exposure to
diethylstilbestrol (DES)
Infection (subclinical or clinical)
Urogenital tract
Amniotic cavity
Systemic
a
From Singh GK, Yu SM. Infant mortality in the United States: trends,
differentials, and projections, 1950 through 2010. Am J Public Health
1995;85:957964; Amon E, Anderson GD, Sibai BM, et al. Factors
responsible for a preterm delivery of the immature newborn infant (less
than or equal to 1000 gm). Am J Obstet Gynecol 1987;156:11431148;
Meis PJ, Ernest JM, Moore ML. Causes of low birth weight births in public
and private patients. Am J Obstet Gynecol 1987;156:11651168; and
Tucker JM, Goldenberg RL, Davis RO, et al. Etiologies of preterm birth in an
indigent population: is prevention a logical expectation? Obstet Gynecol
1991;77:343347.
and PROM (14,24,25,26); however, conflicting data exist (27).

In most patients, spontaneous labor begins after the membranes rupture. The
latency period, defined as the time between membrane rupture and onset of
contractions, is shorter in patients close to term. More than half of preterm
patients will be in labor within 24 hours of ruptured membranes, and 50% to 60%
will deliver within 1 week despite conservative management (7,20,28,29). PROM
remote from viability is associated with significant rates of maternal and fetal
morbidity and low rates of neonatal survival (7,8,19,29). Neonatal respiratory
distress syndrome (RDS) is the most common complication at all gestational ages
(7,8). Fetal demise resulting from intrauterine infection, umbilical cord
compression, or placental abruption occurs in approximately 1% of cases of PROM
at 30 to 36 weeks of gestation and up to 21.7% of cases of PROM in the midtrimester (7,8). Prolonged oligohydramnios can also be associated with fetal
pulmonary hypoplasia, limb-positioning deformities, and abnormal facies (e.g.,
low-set ears and epicanthal folds) (19,30).
Pharmacologic tocolysis, employed in an attempt to prevent progression of
preterm labor, is implemented most often before 34 weeks of gestation. The
various agents used in pharmacologic tocolysis, which are administered orally,
subcutaneously, or parenterally, can have potentially serious maternal and/or
fetal side effects (Table 14-2) (31). Despite the widespread use and evaluation of
tocolysis over the past 20 years, data fail to demonstrate benefits of tocolysis in
terms of medical costs, improvements in neonatal survival, or improvements in
other long-term outcome parameters (31). The major benefit of tocolysis is its
ability to provide short-term pregnancy prolongation in order that corticosteroids
can be administered (31).
Corticosteroids, which are given between 24 and 34 weeks of gestation to patients
who are at risk for preterm delivery, are used to induce fetal pulmonary maturity.
The recommended protocol includes a single course of betamethasone (12 mg
intramuscularly [i.m.]) given in 2 doses 24 hours apart) or dexamethasone (6 mg
i.m. given in 4 doses every 12 hours). Optimal benefit is seen within 24 hours and
through 7 days after administration (32,33). A recent meta-analysis evaluating
multiple randomized studies confirmed the decreased incidence and severity of
neonatal RDS with antenatal administration of corticosteroids (34). Additional
neonatal benefits include reductions in rates of intraventricular hemorrhage and
death, even with treatment of less than 24 hours' duration and in gestations
earlier than 30 to 32 weeks with ruptured membranes (33). Long-term data of
infants exposed in utero to a single course of corticosteroids failed to demonstrate
any significant adverse effects on physical development, growth, cognitive or
motor skills, or early school performance (31).
The use of antibiotic therapy, particularly aimed at treating bacterial vaginosis,
group B streptococcus, and other genital tract organisms, has been studied widely
in recent years. Initially shown to decrease rates of preterm delivery, preterm
labor, and neonatal sepsis in some studies, the data show mixed results
(31,35,36,37). Routine antibiotic administration is not recommended if used only
in an attempt at prevention of preterm delivery (31). While conservatively
managing preterm PROM remote from term, however, aggressive antibiotic
therapy with erythromycin and ampicillin or amoxicillin prolongs pregnancy and
P.204
decreases major neonatal morbidity (e.g., RDS, early sepsis, severe
intraventricular hemorrhage, severe necrotizing enterocolitis) (20,38,39).
TABLE 14-2 POTENTIAL COMPLICATIONS OF VARIOUS TOCOLYTIC
AGENTS
Agent
Magnesium sulfate
Side Effects
Maternal: Decreased respiratory,

cardiovascular, and renal function;
contraindicated in myasthenia gravis
Betamimetic agents (e.g.,

ritodrine, terbutaline)
Fetal: Alterations in fetal heart rate

patterns
Maternal: Pulmonary edema;
myocardial ischemia or infarction
(rare); hypotension; hyperglycemia;
hypokalemia
Fetal: Alterations in fetal heart rate
patterns
Prostaglandin synthetase inhibitors Maternal: Gastrointestinal bleeding

(e.g., indomethacin, sulindac)
Fetal: Constriction of ductus arteriosus;
reversible, dose-dependent
oligohydramnios from decreased fetal
urinary output; pulmonary
hypertension seen with prolonged use
(rare)
Calcium channel blockers (e.g.,
Well tolerated
nifedipine)
a
From Goldenberg RL. The management of preterm labor. Obstet Gynecol

2002;100:10201037.
Screening for Preterm Delivery

Because a woman's past obstetric history may be important for subsequent
pregnancies, several risk-scoring systems have been developed. The most widely
applied system was devised by Papiernik and modified by Creasy (40). Application
of these systems to various populations in the United States, however, has been
disappointing (41), and applicability to nulliparae has not been accepted. A 1996
study of almost 3,000 gravidas that attempted to develop a risk-assessment
system for predicting spontaneous preterm delivery by using clinical information
at 23 to 24 weeks of gestation was similarly disappointing (18). A recent study
demonstrated the utility of the preterm labor index, a clinical tool to assess the
likelihood of overall preterm delivery and delivery within 1 week (42). The
preterm labor index, originally proposed in 1973, combines four clinical
parameters (uterine contractions, PROM, vaginal bleeding, and cervical dilation)
and is comparable to newer biochemical markers for predicting preterm delivery
(42,43). Its utility across populations remains to be studied.
Other screening tools have been proposed to help identify the pregnancy at risk
for preterm delivery, including home uterine activity monitoring, screening for
fetal fibronectin in vaginal secretions, cervical examination, and screening for
genital tract colonization and/or infection.
Home Uterine Activity Monitoring

The working premise behind home uterine activity monitoring (HUAM) is that
uterine contractions may increase in frequency in the 24 hours before the
development of preterm labor (14,44). The use of HUAM involves
tocodynamometric recording of uterine contractions combined with interpretation
and daily telephone contact by health care providers to try to detect early preterm
labor. Oral tocolysis often is used in conjunction with HUAM. Data from many
randomized studies evaluating the efficacy of HUAM are conflicting and the quality
of supportive evidence is limited by study designs (14,31,45-47). Its use in clinical
practice is widespread in spite of the lack of endorsement by the U.S. Preventive
Services Task Force (46) and the American College of Obstetricians and
Gynecologists (ACOG) (14).
Fetal Fibronectin
Fetal fibronectin (fFN), a protein produced by the fetal membranes, most likely
functions to bind the placenta and membranes to the decidua. Disruption of this
normal interface, as with preterm labor or PROM, allows leakage of fibronectin into
cervicovaginal secretions. In normal pregnancy it is rarely present after 20 weeks
of gestation, and it has been found to be of value in predicting preterm delivery
(48,49,50). Positive tests for fFN can be seen in women with bacterial vaginosis or
subsequent maternal and fetal infections (48,51). The most useful aspect of the
fFN assay is its negative predictive value: Less than 1% of women with
questionable preterm labor will deliver within the next 2 weeks with a negative
test for fFN compared with approximately 20% of women with a positive test (31).
At this time, however, its use as a routine screening test for the general obstetric
population has not been endorsed by the ACOG Committee on Obstetric Practice
(52).
Cervical Examination
The value of routine digital cervical examination in otherwise uncomplicated
pregnancies is controversial. Asymptomatic cervical dilation may be a normal
anatomic
P.205
variant, and may represent the earliest sign of impending preterm delivery,
implying cervical incompetence or undiagnosed preterm labor (53,54,55).
Sonographic cervical examination is superior to digital assessment of cervical
dilation, length, and effacement, and confirms the association of cervical
shortening with preterm delivery (56,57,58,59,60). Predictive sonographic
findings for increased risks of preterm delivery include cervical shortening
(generally described as less than 2.5 to 3 cm), funneling or ballooning of
membranes at the level of the internal os, and cervical dilation (57,60,61).
Currently, routine use of sonographic cervical length in the prediction of preterm
delivery is not recommended because of the lack of proven treatments affecting
outcome (14). Combining transvaginal sonography with other biochemical
markers, such as fFN, may prove more predictive, however (14,62).
Genital Tract Screening

Routine screening for various infections (e.g., syphilis, hepatitis B, and rubella) is
recommended early in pregnancy. Additional testing for other sexually transmitted
disorders, however, is left to the discretion of the practitioner based on the
patient's history and physical examination (63). Infections with Neisseria
gonorrhoeae, Trichomonas vaginalis, Chlamydia trachomatis, and group B
streptococci, as well as urinary tract infections, asymptomatic bacteriuria, and
bacterial vaginosis, are associated with premature delivery (64,65,66). Some
investigators recommend more aggressive screening and treatment for these
conditions during gestation, as well as in women contemplating pregnancy (65).
Recently adopted guidelines from the Centers for Disease Control and Prevention
in conjunction with ACOG and the American Academy of Pediatrics recommend the
routine screening for group B streptococcus colonization in late pregnancy and
antibiotic treatment for carriers in labor (67). This strategy reflects attempts to
prevent early onset neonatal disease rather than impact on rates of prematurity
(31,67).
MATERNAL NUTRITION
Recognition of the importance of proper nutrition during pregnancy has varied
over the years. Earlier in this century, restrictions in maternal diet were
implemented to lessen fetal growth in order to decrease the rates of preeclampsia

and labor complications (68). Fetal growth, which is determined by complex
interactions of environmental, genetic, and physiologic factors, is influenced by
both maternal prepregnancy body mass index and pregnancy weight gain.
Whereas low birth weight and IUGR are seen in infants of underweight mothers
and mothers with poor weight gain, macrosomia is common in infants of
overweight mothers and mothers with excessive weight gain (68,69,70). Current
weight-gain recommendations for singleton pregnancies are 25-35 pounds for
normal weight gravidas, 28-40 pounds for underweight gravidas, and 15-25
pounds for overweight gravidas; in twin gestations, ideal weight gain is 35-45
pounds (71). Recent epidemiologic evidence suggests an independent association
between maternal obesity and fetal neural tube defects and other malformations
(72), as well as increased maternal risks of diabetes mellitus, preeclampsia, and
cesarean delivery (73).
Although it is common for multivitamins to be prescribed to pregnant women in
the United States, a balanced diet with appropriate weight gain should supply the
required vitamins and make supplementation unnecessary. In certain conditions,
however, specific recommendations are warranted. There are relatively few food
sources high in vitamin D, for example; in the United States, its major source is
vitamin D-fortified milk (74). Vitamin D synthesis in the skin depends on
ultraviolet light skin exposure; however, there is generally insufficient exposure in
winter, especially at higher latitudes (74). Vitamin D deficiency in pregnant
women varies based on the patient's racial and ethnic background as well as
dietary customs; a fairly high prevalence has been reported among Indian and
Pakistani women living in Britain (74). Maternal deficiency of vitamin D may be
seen with neonatal hypocalcemia and tetany. If supplementation is to be
administered, 400 to 500 IU doses daily have been reported as safe and
adequate; excessive supplementation can result in calcium hyperabsorption,
hypercalcemia, and calcification of soft tissues (74).
Vitamin supplementation of vitamin B12 and zinc are indicated for strict
vegetarians; additional folic acid is suggested for women taking anticonvulsant
medications, carrying multiple gestations, or with hemoglobinopathies. To prevent
anemia, iron supplementation in the second and third trimesters is recommended
(74). In contrast, excessive vitamin A, in daily doses of at least 25,000 to 50,000
IU, is teratogenic, producing malformations similar to those seen with maternal
exposure to 13-cis-retinoic acid (Accutane) (75).
High-dose folic acid supplementation (4 mg/day) reduces the risk of a subsequent
neural tube defect in women with a prior affected infant (76), and such doses,
beginning 1 month before conception and continuing through the first 3 months of
pregnancy, are advocated (77). Further studies led to the Centers for Disease
Control and Prevention statement that all women of child-bearing age in the
United States who are capable of becoming pregnant should consume 0.4 mg of
folic acid daily (78). Beginning January 1998, the U.S. Food and Drug
Administration ordered folic acid fortification of bread, flour, and other grain foods
to help prevent these birth defects (79).
MATERNAL ILLNESSES
Hypertension
Hypertension complicates 5% to 8% of pregnanciesand constitutes a major cause
of maternal and perinatal
P.206
morbidity and mortality (80,81,82). Potential fetal and neonatal complications
associated with chronic hypertension include prematurity, increased overall
perinatal morbidity, IUGR, and fetal death (80,81,82). Several studies document
that the risk of perinatal mortality is increased two to four times if the mother is
hypertensive compared with the general obstetric population (83). Maternal
complications include superimposed preeclampsia, placental abruption, cesarean
delivery, and potentially life-threatening complications such as pulmonary edema,
hypertensive en-cephalopathy, retinopathy, cerebral hemorrhage, and acute renal
failure (80,81,82,83,84,85). Low-risk patients have mild essential hypertension
and no organ involvement; those with severe hypertension or superimposed
preeclampsia are considered to be at high risk. The risks of significant
complications are especially increased in patients with uncontrolled severe
hypertension and underlying renal or cardiac disease prior to or early in gestation
(82).
Classification and terminology of the different subdivisions of the hypertensive
disorders in pregnancy is confusing. A recent recommendation by the National
High Blood Pressure Education Program Working Group replaces the term
pregnancy-induced hypertension with gestational hypertension to describe
situations in which elevated blood pressure without proteinuria develops after 20
weeks of gestation and blood pressure levels return to normal postpartum
(80,85). Up to 25% of women with gestational hypertension will develop
proteinuria or preeclampsia; with severe chronic hypertension, this rate can
approach 50% (84). Proteinuria is classified by at least 0.3 g of protein in a 24hour urine specimen, which usually corresponds to 1+ or greater on a urine
dipstick evalua-tion (83).
Preeclampsia is defined as hypertension with proteinuria in addition to other
possible symptoms of headache, edema, visual disturbances, and epigastric pain;
laboratory abnormalities may involve hemolysis, elevated liver enzymes, and low
platelet count (known by the acronym HELLP). HELLP syndrome can occur in up to
20% of women with severe preeclampsia and may have a variety of clinical
presentations (86). In past years, the diagnosis of preeclampsia included specific
blood pressure elevations above the patient's baseline blood pressure; these
clinical parameters have not been found to be a reasonable prognostic indicator of
outcome (85). Eclampsia involves the development of seizures and/or coma,
which represents central nervous system involvement, in a preeclamptic patient.
Severe preeclampsia can be defined by the criteria listed in Table 14-3 (80,85).
Thrombocytopenia (platelet count below 100,000/mL) is the most consistent
finding in patients with preeclampsia (87). Table 14-4 outlines the risk factors for
the development of preeclampsia (80,85).
The precise pathophysiologic factors involved in preeclampsia have been difficult
to elucidate. Trophoblastic invasion by the placenta appears to be important
because the severity of hypertension appears to be related to the degree of
trophoblastic invasion (80,88). Vascular changes, specifically vasospasm, seem
responsible for many of the serious clinical manifestations (e.g., hypertension and
diminished renal function). Earlier investigations demonstrated support for the
involvement of vasospasm in the etiology of preeclampsia. Specifically, the failure
of the blunted pressor response to angiotensin II present in normal pregnancy is
not seen in preeclampsia (89), and a progressive sensitivity to the pressor effects
of infused angiotensin can be demonstrated after 18 weeks in patients destined to
become preeclamptic (90). Other factors may involve an imbalance in the
production of prostacyclin, a potent vasodilator, relative to levels of thromboxane,
a vasoconstrictor, and alterations in the synthesis of nitric oxide and/or endothelin
1 (85).
TABLE 14-3 CLINICAL MANIFESTATIONS OF SEVERE PREGNANCYINDUCED HYPERTENSION
Systolic blood pressure 160 mm Hg or diastolic blood pressure 110 mm

Hg on two occasions at least 6 hours apart with the patient at bed rest
Proteinuria 5 g in a 24-hour collection or 3+ on two random urine
samples collected at least 4 hours apart
Oliguria (<500 mL in 24 hours)
Cerebral or visual disturbances
Pulmonary edema
Right upper quadrant or epigastric pain
Hepatocellular dysfunction
Thrombocytopenia
Intrauterine growth restriction
From National High Blood Pressure Education Program Working Group on

high blood pressure in pregnancy. Am J Obstet Gynecol 2000;183(1):S1
S22; and Sibai BM, Lindheimer M, Hauth J, et al. Risk factors for
preeclampsia, abruptio placentae, and adverse neonatal outcomes among
women with chronic hypertension. N Engl J Med 1998;339:667671.
TABLE 14-4 RISK FACTORS ASSOCIATED WITH THE DEVELOPMENT OF

PREECLAMPSIA
Nulliparity
Age >35 years
African American race
Family history of preeclampsia
Prior history of preeclampsia
Chronic hypertension
Chronic renal disease
Obesity
Vascular and connective tissue disease
Antiphospholipid syndrome
Thrombophilia
Pregestational diabetes mellitus

Multiple gestation
Gestational trophoblastic disease
Fetal hydrops
From National High Blood Pressure Education Program Working Group on

high blood pressure in pregnancy. Am J Obstet Gynecol 2000;183(1):S1
S22; and Sibai BM, Lindheimer M, Hauth J, et al. Risk factors for
preeclampsia, abruptio placentae, and adverse neonatal outcomes among
women with chronic hypertension. N Engl J Med 1998;339:667671.
P.207
Some fetal effects of preeclampsia reflect vasospasm with regard to placental
perfusion. A decrease in uteroplacental perfusion may result in abruption, IUGR,
oligohydramnios, or nonreassuring fetal status demonstrated on antepartum
testing. Placental abruption, which occurs in fewer than 2% of patients with
chronic hypertension, can be twice as high in those with superimposed
preeclampsia (82,83). In 1990, Burrows and Andrews (91) found that neonatal
thrombocytopenia occurred in more infants of hypertensive than in those born to
normotensive mothers (9.2% compared with 2.2%).
Prematurity significantly contributes to the increased rates of perinatal morbidity
and mortality associated with preeclampsia. Conservative management,
recommended in a tertiary care setting or in consultation with a maternal- fetal
medicine subspecialist, may be attempted for women with severe preeclampsia
remote from term (80,82). Delivery remains the only definitive treatment, and
ultimately may be required (92,93). In cases managed conservatively, patients
are hospitalized; bed rest with frequent clinical and laboratory assessments and
corticosteroids are administered to enhance fetal pulmonary maturity. Magnesium
sulfate, an agent long recognized for its anticonvulsant effects, is often
administered (85). Although phenytoin sulfate may be used, magnesium sulfate is

the preferred drug (94,95). Blood pressure control is achieved with the use of
antihypertensive agents such as labetalol, hydralazine, or nifedipine (80,81,82). A
protocol of high-dose corticosteroid administration stabilizes both clinical and
laboratory parameters in patients with HELLP syndrome before term (96,97).
Several different antihypertensive agents are used during pregnancy. Methyldopa is frequently used because it is a safe, effective agent that has been
studied extensively. Newer antihypertensive drugs that are safe in pregnancy
include beta-blockers, labetalol (a combination alpha- and beta-blocker), and
calcium channel blockers such as nifedipine. There is concern about using diuretics
during pregnancy because of an accompanying plasma volume reduction (83).
The National High Blood Pressure Education Program Working Group on High
Blood Pressure in Pregnancy suggests that diuretics can be used safely except in
situations in which there is already evidence of diminished uteroplacental function
(80,81). Angiotensin-converting enzyme (ACE) inhibitors, commonly used in
young, nonpregnant adults, are both teratogenic and fetotoxic if administered
during gestation. Skull defects (hypoplastic calvaria and encephalocele) and in
utero renal failure leading to oligohydramnios, pulmonary hypoplasia, longstanding neonatal anuria, and fetal or neonatal death have been reported
(98,99,100). In women using an ACE inhibitor, antihypertensive medication is
changed to another agent once pregnancy is confirmed.
Over the years, prevention of preeclampsia has been attempted. Initially, a
beneficial effect was demonstrated in high-risk groups that were given low-dose
aspirin (60 to 80 mg daily), but subsequent studies have been conflicting and a
similar effect was not found in low-risk women (101,102,103,104). Similarly,
other studies have obtained disappointing results in evaluating supplementation
with calcium, magnesium, zinc, fish oil, or use of different antihypertensive
medication (103). More recently, evaluation of the uteroplacental circulation via
transvaginal Doppler sonography at 12 to 16 weeks suggests an association with
abnormal waveforms and increased risks of developing preeclampsia later in
gestation (103,104,105,106,107). Further study is ongoing.
Diabetes Mellitus
Diabetes mellitus complicates nearly 4% of pregnancies (108,109). The disease is
classified based on the requirement for insulin therapy into type 1 (insulindependent) or type 2 (non-insulin-dependent) diabetes. The White classification
system for diabetes in pregnancy, developed in 1949, is based on age of onset
and duration of disease, as well as disease progression with respect to vascular
complications (110). With continued improvements in glucose control, assessment
of fetal well-being, and neonatal management, the White classification is no
longer as helpful as it once was in the management of the pregnant diabetic
(111). Instead, the distinction can be made between diabetes that preceded a
woman's pregnancy (pregestational diabetes) and diabetes first recognized during

pregnancy (gestational diabetes).
Fetuses of diabetic mothers may have growth disturbances at both ends of the
spectrum: IUGR and macrosomia. IUGR, fetal growth less than or equal to the
10th percentile for gestational age, is not an infrequent finding in pregnancies of
women with vascular complications of pregestational diabetes. In diabetic
pregnancies, IUGR often results from uteroplacental insufficiency, usually
associated with maternal hypertension, although fetuses with congenital
anomalies also may exhibit signs of IUGR.
Infants of mothers with both pregestational and gestational diabetes are at risk
for macrosomia, which refers to fetal growth beyond a specific weight regardless
of gestational age, usually above 4,000 g or 4,500 g (112). With the development
of a new national reference for fetal growth based on data from over 3.8 million
births (113), clinicians can distinguish between macrosomia and large-forgestational-age (above the 90th percentile for a given gestational age) (112,113).
Several large studies support the continued use of 4,500 g as a value above which
a fetus should be considered to be macrosomic (112,114,115,116,117). The
antenatal sonographic diagnosis of macrosomia remains inaccurate, especially in
diabetics (118,119,120). In women with normal glucose tolerance, 2% of infants
weighed more than 4,500 g compared with 6% of women with untreated
borderline gestational diabetes (112,121). Even in the absence of gestational
diabetes, however, higher infant weights are associated with higher maternal
glucose levels (112,122,123). In patients with untreated gestational diabetes, up
to 20% of infants may be macrosomic (112,124).
Macrosomia is accompanied by the additional risks of prolonged labor, fracture of
the clavicle, birth trauma from
P.208
shoulder dystocia, and instrumented or cesarean deliveries
(125,126,127,128,129). Shoulder dystocia occurs in only 1.4% of all vaginal
deliveries (130). Among deliveries of diabetic mothers, however, shoulder
dystocia occurs two to six times more frequently than in the nondiabetic
population (120,129,131). The complication has been reported to occur in 20% to
50% of infants of diabetic mothers with birth weights above 4,500 g
(114,115,116,117,120). Brachial plexus injury is the most serious complication
resulting from shoulder dystocia. Fortunately, it is rare, occurring in approximately
0.05% to 0.19% of vaginal deliveries (112,120), but that risk is increased 18- to
21-fold among vaginally delivered infants weighing more than 4,500 g
(114,120,132,133,134,135). The majority of cases of brachial plexus injury
resolve within 1 year (136,137), yet permanent injury is more commonly seen
among infants with birth weights greater than 4,500 g (138). Important
information for the practicing clinician to remember is that most cases of shoulder
dystocia and brachial plexus injuries occur among infants who are not macrosomic
(135). Clavicular fracture, which usually resolves without any permanent effect,
can be seen in 0.3% to 0.7% of all deliveries (134,139,140), and its occurrence is
increased up to ten times in the macrosomic infant (134).
Polyhydramnios, defined as excessive amniotic fluid, is not an unusual finding in
diabetic pregnancies. A four-quadrant sonographic assessment of the amniotic
fluid index (AFI) defines polyhydramnios as an AFI greater than the 95th
percentile for gestational age (141). In diabetic pregnancies, the etiology is not
clear, although fetal malformations or poor glucose control may be related. When
polyhydramnios complicates maternal diabetes, higher rates of perinatal morbidity
and mortality are reported (142).
Pregestational Diabetes Mellitus

Estimates from 1999 indicate that approximately 10,000 women in the United
States with pregestational diabetes delivered live births (143). Pregnancies of
women with pregestational diabetes are at significant risk for both spontaneous
abortion and fetal anomalies (144,145,146,147,148,149,150), the latter
representing the major cause of perinatal mortality in this group of patients
(144,149,150). The frequency of congenital anomalies in infants of diabetic
mothers occurs at two to three times the rate in the nondiabetic population (151).
These malformations, which occur before 7 weeks of gestation, commonly include
open neural tube defects, congenital heart defects, and the caudal regression
syndrome (144,151,152).
Hyperglycemia is responsible for the increased risks of both fetal malformations
and spontaneous abortion seen in diabetics (146,147,148,149). Poor glycemic
control, combined with derangements in amino acid and lipid concentrations, is
believed to underlie the development of fetal malformations (147,153).
Hypertrophic cardiomyopathy, which can cause cardiomegaly and congestive heart
failure, may result from elevated maternal glucose levels throughout pregnancy
(154,155). A patient's degree of hyperglycemia can be assessed by the level of
glycosylated hemoglobin (hemoglobin A1c), a retrospective marker of glucose
control (153). Intensive glycemic control in the periods before conception and
organogenesis can lower the frequency of congenital anomalies in infants of
diabetic mothers (144,156,157).
In the past, a major source of perinatal mortality among diabetic pregnancies was
fetal demise, often unexplained and sudden. Among well-controlled diabetics,
stillbirth is an uncommon event (144). With diabetic ketoacidosis, however,
perinatal mortality rates may be as high as 50% to 90% (158). Fetal blood
sampling has confirmed that these previously designated unexplained stillbirths
result from hyperglycemia, metabolic disturbances, polycythemia, and acidemia
(159,160). Efforts to prevent fetal demise at the Joslin Clinic in the 1960s resulted
in the strategy of scheduled preterm deliveries. Even though reductions in the
number of stillbirths occurred, neonatal deaths from RDS were prevalent,

predominantly as a result of errors in estimates of gestational age (161). Although
older data from the 1980s suggested a delay in fetal pulmonary maturation in
diabetic pregnancies (162,163,164) more recent studies have disproved this
assumption, especially in patients with good glucose control (165,166,167,168).
Other complications seen in women with pregestational diabetes mellitus,
especially those with end-organ dysfunction from long-standing vascular disease,
include preterm delivery and preeclampsia (85).
In the immediate neonatal period, infants of diabetic mothers are at higher risks
for a number of metabolic irregularities. Hypoglycemia may occur from a rapid
decline in neonatal glucose levels after delivery. This results from a combination of
removal of the continued placental source of glucose and fetal islet cell
hyperplasia from chronic maternal hyperglycemia. Close maternal glucose control
during labor with continuous insulin and glucose infusions can lessen the
development of neonatal hypoglycemia (169). Other neonatal metabolic
derangements include hypercalcemia, polycythemia, and hyperbilirubinemia
(160,170,171).
Gestational Diabetes Mellitus

In the United States, gestational diabetes mellitus (GDM) is diagnosed by at least
two abnormal values on a 3-hour 100-g oral glucose tolerance test subsequent to
an elevated 1-hour 50-g glucose challenge test (108). GDM occurs in 2% to 5% of
pregnancies but comprises the majority of cases of diabetes in pregnancy (108).
Recent estimates suggest rising prevalence rates as a result of increased detection
following the 1980s' recommendations for universal screening (172,173). In
women with GDM, the risk of developing diabetes mellitus 20 years later is
approximately 50% (174).
For most patients with GDM, glucose levels may be controlled with dietary therapy
alone. Insulin therapy is suggested when fasting glucose levels exceed 95 mg/dL,
1-hour postprandial values exceed 130 to 140 mg/dL, or 2-hour postprandial
value exceed 120 mg/dL (108). Regular exercise may help with glucose regulation
in this
P.209
group of patients (108). One randomized trial found similar results among
patients with GDM comparing use of insulin versus glyburide, a second-generation
sulfonylurea, although additional study is needed (175). In general, infants of
mothers with GDM are at risk for macrosomia, operative and cesarean deliveries,
birth trauma (including fracture of the clavicle, shoulder dystocia, and brachial
plexus injury), fetal death, and neonatal hyperglycemia and/or hyperbilirubinemia
(108,112). Patients diagnosed with GDM early in pregnancy and in whom glucose
levels require intensive therapy with diet and insulin behave much like those with
pregestational diabetes; these patients has a higher risk of fetal malformations

and stillbirth than do patients with mild GDM (111). Postpartum recommendations
for women diagnosed with GDM during pregnancy include evaluation of glucose
status using a 75-g glucose tolerance test (108).
Fetal Assessment
Several tools are used to assess fetal well-being in diabetic and in other high-risk
pregnancies. Obstetric ultrasonography can be of value in determining early fetal
viability and in screening for various anomalies. Fetal echocardiography is
employed generally after 20 weeks' gestation to evaluate fetal cardiac structure.
In the third trimester, sonography can assess fetal growth and aid in the diagnosis
of macrosomia, polyhydramnios, or septal hypertrophy. An ongoing sense of fetal
well-being can be obtained through maternal perception of fetal movement
counts. Serial fetal biophysical testing (e.g., nonstress tests and biophysical
profiles) is usually implemented by 32 weeks in most insulin-dependent diabetics;
pregnancies with well-controlled gestational diabetes, considered to be at low risk
for fetal demise, may not require testing except in the presence of other obstetric
factors (161).
Autoimmune Disorders
A number of disorders that involve circulating autoantibodies and/or deposition of
immune complexes may have direct effects on pregnancy. These include the
rheumatologic or connective tissue diseases and conditions associated with
circulating antiphospholipid antibodies.
Rheumatologic Disorders
Rheumatologic disorders are chronic inflammatory diseases usually affecting the
connective tissues and joints. The most common disorders occurring in young
women include systemic lupus erythematosus, rheumatoid arthritis, scleroderma,
and Sjgren's syndrome.
Systemic Lupus Erythematosus

Systemic lupus erythematosus (SLE) is the most common connective tissue
disorder seen among reproductive-age women. Its various clinical presentations
include polyarthritis, skin manifestations, Raynaud phenomenon, and nephritis;
laboratory abnormalities such as anemia, leukopenia, thrombocytopenia, and the
presence of autoantibodies are common. The disease is characterized by
remissions and exacerbations. Patients with SLE have a high prevalence of fetal
wastage: spontaneous abortion, IUGR, preterm delivery, stillbirth, and perinatal
death (176,177,178,179,180,181). A majority of the pregnancy losses result from
fetal death in the second and third trimesters (182). Many lupus patients have
pregnancies with high rates of preterm and term PROM (183). Fetal survival is
higher when the disease is in remission (176,177,181,182,183,184). Other
predictors of fetal wastage include active nephritis, hypertension, and circulating
antiphospholipid antibodies (e.g., lupus anticoagulant or anticardiolipin
antibodies), the latter likely the most important factor associated with pregnancy
loss (176,177,179,180,181,182,184,185).
Despite older data that supported the opinion of disease exacerbation in lupus
patients during pregnancy, most authorities now agree that pregnancy has no
effect on disease progression (176,186,187). Treatment with corticosteroids is
standard (178,179,180,185,188). Salicylates and other nonsteroidal
antiinflammatory agents (e.g., paracetamol) are commonly used, although high
doses are discouraged (188,189,190,191). Azathioprine, an immunosuppressive
that has been used predominantly in renal transplantation patients
(188,192,193,194), and antimalarial agents are used widely in SLE patients and
considered to be safe during pregnancy (179,188,195,196,197,198). In some
cases, plasmapheresis has been performed (199,200).
Infants of mothers with SLE are at risk for the neonatal lupus syndrome. This
constellation of findings consists of abnormalities in the heart and skin or
development of clinical features of SLE and occurs from transplacental passage of
maternal antibodies (201). Congenital complete heart block is the most frequently
seen heart abnormality (202), occurring in fewer than three percent of infants at
risk (189). The pathophysiology involves deposition of immunoglobulin,
specifically circulating IgG autoantibodies directed against ribosomal nucleoprotein
antigens (anti-Ro or SSA and, to a lesser extent, anti-La or SSB antibodies) in
fetal cardiac tissue (201,203). Anti-Ro (SSA) and anti-La (SSB) antibodies are
detectable in 40% to 50% of SLE patients (181,202). Most cases of congenital
complete heart block occur in fetuses whose mothers do not have overt clinical
SLE (204). Many studies, however, indicate that the majority of mothers with
affected children have detectable anti-Ro (SSA) antibodies (201,202). Heart block
in the absence of structural defects has been documented as early as 16 weeks of
gestation (205).
The presenting finding of congenital heart block is fetal dysrhythmia; sonography
may reveal a pericardial effusion or hydrops resulting from either congestive
failure or an immune mechanism (myocarditis) (203,206,207,208). Maternal
treatment with dexamethasone and/or plasmapheresis, used to lower circulating
antibody levels and to minimize inflammatory injury, has had some success in
reversing fetal heart block and improving fetal cardiac contractility
(207,208,209,210). Resolution of ascites has been
P.210
reported with corticosteroid use (209). The technique of in utero cardiac pacing,
demonstrated to be technically feasible in a hydropic 24-week fetus with heart
block (211), may be a future option in severe cases. After birth, heart block is
usually permanent; intermittent and incomplete cases, as well as unusual late
presentations, have been described (201,212,213,214). Cardiac pacemaking is
instituted for those infants that survive; however, mortality rates are high (12%
to 28%) (201,215). Coexistent cardiac anomalies among infants with congenital
heart block are common (214,216).
The skin is the other major organ system involved in the neonatal lupus
syndrome. Cutaneous lesions are frequently widespread macular rashes, although
a butterfly rash and discoid lesions are found occasionally (181,201). These
histologically inflammatory lesions generally appear within the first few weeks of
life and disappear spontaneously within 6 months, coexistent with the clearance of
maternal autoantibodies from the neonatal circulation (181,201). Hematologic
manifestations, such as anemia and thrombocytopenia, glomerulonephritis,
hepatosplenomegaly, and neurologic symptoms, are unusual (201,202).
Rheumatoid Arthritis
Rheumatoid arthritis (RA) is the most common inflammatory joint disease,
affecting approximately 1% of the population in North America and complicating
approximately 1 in 1,000 to 2,000 pregnancies (217). The disorder affects women
three times as often as men, with a peak incidence between the fourth and sixth
decades of life (217). There is a familial preponderance associated with the tissue
antigen HLA-DR4 (217). RA is characterized by chronic polyarthritis and
inflammatory synovitis, usually of the peripheral joints, resulting in bone and
cartilage destruction and joint deformities. Other clinical manifestations include
anorexia, weakness, fatigue, and vague musculoskeletal complaints. The diagnosis
is made based on specific criteria outlined by the American Rheumatism
Association (218). Despite much study, it is not known why the symptoms of RA
are ameliorated during pregnancy; studies evaluating cortisol levels and sex
hormone concentrations have not provided any solid answers (219,220). In
contrast to SLE, perinatal morbidity and mortality are not increased in patients
with RA (220).
The major therapies for RA are acetylsalicylic acid (aspirin) and nonsteroidal
antiinflammatory agents; concerns for adverse fetal and/or neonatal effects (such
as impaired hemostasis, premature closure of the fetal ductus arteriosus,
prolonged gestation, and long labor) have been largely theoretical (188,219).
Gold therapy, which lowers the levels of rheumatoid factor, has been used for
many years; antimalarial agents are also administered (188,219). Although Dpenicillamine has been used, some reports have shown connective tissue defects
similar to Ehlers-Danlos syndrome in children with antenatal exposure (221).
Scleroderma
Scleroderma is an autoimmune connective-tissue disorder that occurs at least five

times more frequently in women of childbearing age than in men, affecting the
skin, gastrointestinal tract, especially the esophagus, lungs, and kidneys (222).
The typical remissions and exacerbations of the disease make it difficult to assess
the effect of pregnancy. According to a 1989 report on 94 cases of scleroderma in
pregnancy and a review of the literature, pregnancy had either no effect on the
disease or was associated with exacerbations (223). A retrospective case control
study of 48 women revealed an association with first trimester pregnancy loss in
those patients with long-standing diffuse scleroderma and a 29% prematurity rate
(222). Fetal mortality may be as high as 20% and the development of neonatal
scleroderma has been described.
Sjgren Syndrome
This rare autoimmune disorder, also known as keratoconjunctivitis sicca or sicca
syndrome, involves lymphocytic infiltration of the salivary and lacrimal glands
resulting in loss of saliva and tears. Sjgren syndrome is both clinically and
immunologically related to SLE. Many circulating autoantibodies are present, as
well as anti-Ro (SSA) and anti-La (SSB) antibodies. Approximately 50% of
patients with Sjgren syndrome have rheumatoid arthritis (224). Pregnancy
complications of fetal loss and congenital heart block have been reported (225).
Antiphospholipid Antibodies and the

Antiphospholipid Syndrome
Circulating antiphospholipid antibodies and the antiphospholipid syndrome (APS)
are associated with clinical complications such as adverse pregnancy outcomes,
autoimmune thrombocytopenia, and thrombosis (226). These antibodies are
directed against negatively charged phospholipids present on cell membranes,
notably platelets and endothelial cells (227). The most common antiphospholipid
antibodies are the lupus anticoagulant and anticardiolipin antibodies. The lupus
anticoagulant was originally described as a circulating anticoagulant identified in
two patients with SLE (228). It is paradoxically named, however, for the
anticoagulant activity is seen in vitro, whereas it acts is as a potent thrombotic
agent in vivo.
Antiphospholipid antibodies may be found in the normal population, but they are
more frequently detected in patients with SLE (229,230). Diagnosis of the APS
involves both clinical and laboratory parameters: moderately high levels of the
antiphospholipid antibodies (lupus anticoagulant and/or anticardiolipin antibodies)
plus one clinical parameter (thrombosis, autoimmune thrombocytopenia, or
pregnancy loss) (226,231,232,233). Potential pregnancy complications include
those listed in Table 14-5 (226,227,234). The most striking adverse fetal
outcomes associated with APS are fetal death and IUGR. Pregnancy loss may be
the only clinical marker associated with the presence of antiphospholipid

antibodies, however. A 1994 study demonstrated a very high rate of
development of thrombotic complications after the identification of these
antibodies, even
P.211
in asymptomatic individuals (235). Other unusual reported complications include
fetal or neonatal thrombosis resulting from transplacental passage of maternal
antibodies (236,237) and fetal effects of therapy.
TABLE 14-5 PREGNANCY COMPLICATIONS SEEN IN ASSOCIATION WITH
ANTIPHOSPHOLIPID ANTIBODIESa
Fetal loss, including recurrent embryonic loss or fetal death after 10 weeks
Intrauterine growth restriction
Placental infarction
Preterm birth
Early onset severe preeclampsia
Nonreassuring fetal heart rate patterns
Unusual postpartum syndrome (e.g., cardiopulmonary disease, fever,
hemolytic uremic syndrome)
From American College of Obstetricians and Gynecologists.

Antiphospholipid syndrome. ACOG Educational Bulletin 244. Washington,
DC: ACOG, 1998; Branch DW. Antiphospholipid syndrome: laboratory
concerns, fetal loss, and pregnancy management. Semin Perinatol
1991;15:230237; and Kniaz D, Eisenberg GM, Elrad H, et al. Postpartum
hemolytic uremic syndrome associated with antiphospholipid antibodies. A
case report and review of the literature. Am J Nephrol 1992;12:126133.
Initial treatment of the APS involved use of low-dose aspirin (75 mg) and
corticosteroids (238,239); successful pregnancy outcomes were reported by many
investigators worldwide using similar regimens (227). Oropharyngeal candidiasis,
gestational diabetes, osteoporosis, and PROM are some of the complications seen
with corticosteroid use (227,240,241,242). An alternative regimen consisting of
low-dose aspirin and heparin currently is the preferred therapy
(226,240,243,244). Potential problems associated with heparin include
thrombocytopenia and heparin-induced osteoporosis, although these
complications are fewer with administration of low-molecular-weight heparin
(243,245). Intravenous immunoglobulin also has been used (246,247). In spite of
aggressive treatment, however, successful pregnancy outcomes are not
guaranteed.
Thyroid Disorders
The second most common endocrine disorder in women of reproductive age
consists of disorders of thyroid metabolism (248). Pregnancy significantly affects
thyroid physiology. Normal pregnancy is characterized by hypermetabolic effects
that resemble the clinical findings of hyperthyroidism. Although some thyroid
function tests may be altered during pregnancy, largely by hyperestrogenemia
and the resulting increase in thyroid-binding globulin, levels of free-circulating
hormone (T4 and T3) and thyroid-stimulating hormone (TSH) are unchanged.
Moreover, several obstetric conditions, notably hyperemesis gravidarum or
gestational trophoblastic disease, may cause abnormalities in thyroid function.
These principles must be understood when making the diagnosis of thyroid
disease during pregnancy.
Hyperthyroidism
Approximately 1 in 2,000 pregnancies will be complicated by thyrotoxicosis; the
condition that results from excessive production of thyroid hormone. Most cases of
hyperthyroidism are due to Graves disease, an autoimmune disorder
characterized by production of thyroid-stimulating immunoglobulin (TSI) and
thyroid-stimulating hormone-binding inhibitory immunoglobulin (TBII), both of
which act on TSH production (248,249,250). Graves disease accounts for 95% of
hyperthyroidism during pregnancy, affecting 1% of American women (249,250).
Other causes of hyperthyroidism include excess TSH production, gestational
trophoblast tumors, toxic multinodular goiter, subacute thyroiditis, overactive
thyroid adenoma, and extrathyroid source of thyroid hormone (248). Poorly
controlled or untreated hyperthyroidism during pregnancy is associated with
greater risks of severe preeclampsia, congestive heart failure, and preterm
delivery (248). The most serious consequence of uncontrolled thyrotoxicosis,
thyroid storm, presents with exaggerated features of thyrotoxicosis including
fever, altered mental status, vomiting, diarrhea, and cardiac dysfunction
manifested by arrhythmia and heart failure (248). Usually, this rare condition is
triggered by a precipitating event such as infection, trauma, or delivery; though
the diagnosis can be difficult to make, treatment of the underlying condition is
critical (248).
Medical treatment of hyperthyroidism involves blocking thyroid hormone
production and controlling the peripheral clinical symptoms. Propylthiouracil (PTU)
and methimazole, which block production of thyroid hormone, are safe in
pregnancy (251). Both agents cross the placenta, but methimazole has been
associated with a scalp disorder known as aplasia cutis (252). Peripheral
manifestations such as tachycardia are controlled with beta-blockers, notably
propranolol, widely used during pregnancy. In difficult cases, thyroidectomy may
be performed during pregnancy after medical control of thyrotoxicosis has been
achieved. Iodides may be used for short periods in preparation for thyroidectomy
or for management of thyroid storm.
Most fetal morbidity and mortality develop from uncontrolled maternal
hyperthyroidism. Prolonged iodide exposure after 10 to 12 weeks of gestation may
result in fetal hypothyroidism and goiter (248,250). Fetal thyrotoxicosis, which
results from transplacental passage of TSI (253), has been reported in
approximately 1% of infants of mothers with Graves disease and is associated
with fetal death (248,254,255,256). Fetal blood sampling has been helpful in
measuring fetal thyroid status (255,256).
Hypothyroidism
Overt hypothyroidism rarely complicates pregnancy, although retrospective
studies show high rates of stillbirths, low-birth-weight infants, medically indicated
preterm delivery, preeclampsia, and abruption in women with inadequately
treated hypothyroidism (248). In many cases, hypothyroidism develops after
thyroidectomy or
P.212
radioiodine therapy; other causes of hypothyroidism include Hashimoto thyroiditis,
carcinoma, or insufficient thyroid replacement (248,250). Replacement therapy
with thyroxine is recommended for these patients and TSH levels are used to
guide therapeutic dosages. Replacement doses of thyroxine may need to be
increased during pregnancy (257). Neonatal hypothyroidism needs to be excluded
if maternal antenatal treatment involved radioactive iodine; however, because
congenital hypothyroidism is a difficult diagnosis, screening for all infants is
routine (248,258).
Perinatal Infections
Any infection occurring during pregnancy has the potential for causing infectious
or teratogenic complications in the fetus, some with devastating effects. The two
important routes for fetal infection are hematogenous via the placenta and
ascending via the vagina and cervix, the latter usually occurring intrapartum. The
effect of an infectious agent on fetal growth and development depends on, among
other things, the type of organism, the infectious load, timing in gestation, and
potential organ systems affected. Many different organisms have been implicated
in causing fetal infection; Table 14-6 lists some of the important perinatal viral
infections (259,260,261,262,263,264,265), and Table 14-7 lists some of the
important perinatal nonviral infections (259,260,261,266,267,268).
Treatment of many perinatal infections either does not exist, as in cases with
viruses, or may not prevent congenital infection, as with syphilis. Consequently,
efforts to minimize the effect on the neonate focus primarily on prevention.
Examples include cesarean delivery in cases of active maternal genital herpes

infection, thereby decreasing the risk of intrapartum ascending infection, neonatal
administration of the hepatitis B vaccine, or administration of prophylactic
eyedrops to newborns to prevent gonococcal ophthalmia neonatorum. Attempts at
preventing early onset neonatal disease from group B streptococcus have resulted
in guidelines from the Centers for Disease Control and Prevention, adopted by
ACOG, for universal third-trimester screening and administration of intrapartum
antibiotics to carriers (266). There is much work to be done, however, in devising
strategies to prevent other perinatally transmitted illnesses such as congenital
cytomegalovirus infection or human immunodeficiency virus that may have
damaging effects on a significant number of children annually.
Thromboembolic Disorders
The risk of venous thromboembolism is five times greater during pregnancy than
in the nonpregnant population, with an absolute risk of 0.5 to 3.0 per 1,000
women (269). The most constant predisposing factor for thromboembolic disease
during pregnancy is venous stasis, although other factors include prolonged bed
rest, operative vaginal or cesarean delivery, sepsis, hemorrhage, multiparity, and
advanced maternal age (269,270). Although older reports demonstrated the
greatest risk for development of thromboembolism during the third trimester and
the immediate postpartum period, newer reports raise concerns that older studies
may have been skewed by the practices of common operative deliveries, delaying
postpartum ambulation, and suppression of lactation with oral estrogen
(269,270,271). Newer data suggest not only equal frequencies in all trimesters
but that antepartum events may occur at least as often as those postpartum
(269,271).
Maternal thromboembolism may result from an underlying inherited coagulopathy
(also known as thrombophilia). Most of these autosomally dominant disorders
result in deficiencies of antithrombin III, protein C and protein S, or activated
protein C resistance because of the factor V Leiden mutation
(269,271,272,273,274,275,276). Other inherited thrombophilias include the
prothrombin G20210A mutation, 4G/4G mutation in plasminogen activator
inhibitor 1 (PAI-1) gene, and the thermolabile variant of
methylenetetrahydrofolate reductase (C677T MTHFR), the latter resulting in
hyperhomocysteinemia (269,272,273,274). Individuals with inherited
coagulopathies have significantly increased risks of thromboembolism as well as
pregnancy complications such as stillbirth, IUGR, and severe preeclampsia
(269,271,272,273,274,275,276). The APS may also cause thromboembolic events
(226,269,271).
The clinical diagnosis of deep vein thrombosis during gestation is imprecise.
Noninvasive tests, such as impedance plethysmography or real-time Doppler
sonography, accurate in nonpregnant patients, may be difficult to interpret during
pregnancy; the data they provide as initial tests, however, may be useful
(269,270). Duplex ultrasound venography has been the standard in diagnosing
deep venous thromboembolism (277). For both pregnant and nonpregnant
patients suspected of having a pulmonary embolism, the ventilation-perfusion (V/
Q) scan is the recommended study (278). In recent years the utility of spiral
computed tomography (CT) has been demonstrated (279). With a 90% sensitivity
and 90% specificity, some authors recommend replacing the V/Q scan with the
spiral CT scan, although studies in pregnancy are lacking (269).
Treatment for acute deep vein thrombosis or pulmonary embolism during
pregnancy involves 5 to 10 days of intravenous heparin followed by subcutaneous
heparin in doses to achieve full-dose anticoagulation for a minimum of 3 months
(269,270,271). Antepartum prophylactic heparin therapy is indicated in patients
with a previous thrombotic event (269,271). The preferred anticoagulant during
pregnancy is heparin because it does not cross the placenta (280); the newer lowmolecular-weight heparins are safe in pregnancy (269,281). Warfarin is avoided
because of associated fetal malformations: first-trimester exposure may produce
an embryopathy involving stippled epiphyses, nasal and limb hypoplasia, and
hypertelorism (269,282,283), and second- or third-trimester exposure is
associated with central nervous system abnormalities (284). Warfarin may be
used postpartum in women who are breast-feeding (271).
Renal Disorders
Mild renal dysfunction typically has little, if any, effect on pregnancy outcome; however,
adverse pregnancy events are well described in women with moderate to severe renal
insufficiency (e.g., serum creatinine >1.4 mg/dL) (285,286,287). These pregnancies are
especially risky: maternal complications include anemia, vascular accidents, placental
abruption, chronic hypertension, pregnancy-induced hypertension, preeclampsia,
proteinuria, and worsening renal function; perinatal complications such as IUGR, stillbirth,
prematurity, polyhydramnios, and midtrimester pregnancy loss are not unusual (285,287,288).
Dialysis may be used during pregnancies complicated by renal insufficiency. More literature
is available for hemodialysis than for continuous ambulatory peritoneal dialysis (289,290,
291,292,293). Pregnancy success rates in dialysis patients are at most 52% (290), and
outcomes generally are more promising for women in whom dialysis has been initiated
during pregnancy compared to patients already on dialysis before
conception (289,291). Complications associated with peritoneal dialysis in pregnancy include

preterm labor (289) and acute peritonitis (291,294). Hemodialysis involves intermittent and
often significant fluid shifts, which may be accompanied by hypotension, electrolyte imbalances,
and preterm labor (288). In one reported case, in spite of efforts to eliminate major fluid shifts
and changes in arterial pressure in a patient at 32 weeks gestation, uterine artery Doppler
studies showed a significant increase after hemodialysis; redistribution of flow away from the
uteroplacental vascular bed during hemodialysis was postulated (288).
TABLE 14-6 SOME PERINATAL VIRAL INFECTIONSa

Time of Critical
Exposure
Potential Severe Fetal or

Neonatal Effects(s)
Cytomegalovirus (CMV) Doublestranded DNA

herpesvirus
Transplacental more Gestational age

than ascending; also not critical;
breast-feeding
primary infection
more severe
than recurrent
Rubella
RNA virus
Transplacental
Herpes simplex virus

(HSV)
Double-stranded Ascending
DNA virus
intrapartum
Fetus: IUGR, nonimmune

hydrops,
hepatosplenomegaly,
chorioretinitis,
microcephaly, cerebral
calcifications,
microophthalmia,
hydrocephaly. Newborn:
jaundice, petechiae, LBW,
hepatosplenomegaly,
thrombocytopenia
Congenital heart disease,
purpura, cataracts,
retinopathy, IUGR,
microcephaly
Microcephaly, seizures,
mental retardation,
mircoophthalmia, retinal
dysplasia, meningitis,
chorioretinitis (HSV-2 more
important than HSV-1)
Virus;
Type of Virus
Route of
Transmission
First trimester
Intrapartum
Varicella zoster
DNA herpesvirus Transplacental
Congenital
varicella If <20
weeks neonatal
varicella if 5 days
before or 2 days
after delivery
Parvovirus B-19
Single-stranded Transplacental
DNA virus
Severe effects
seen with
exposure before
20 weeks
Human immunodeficiency virus (HIV)
Retrovirus
Congenital varicella: limb

hypoplasia, cutaneous
scars, chorioretinitis,
corticalatrophy,
microcephaly; Neonatal
varicella if contracted
peripartum
Low risk for fetal morbidity;
possibly spontaneous
abortion, hydrops (from
aplastic anemia, myocarditis
or chronic hepatitis) and/or
stillbirth
Neonatal acquired
immunodeficiency syndrome
(AIDS)
Transplacental;
Any
ascending
intrapartum; breastfeeding
Hepatitis B
DNA virus
85%95% ascending
Postnatal chronic hepatitis,
intrapartum; rest
cirrhosis, hepatocellular
transplacental during
carcinoma
acute hepatitis or
postnatal via breastfeeding and close
contact
DNA, deoxyribonucleic acid; IUGR, intrauterine growth restriction; LBW, low birth weight; RNA, ribonucleic
acid.
From Bale JF. Congenital infections. Neurol Clin 2002;20:10391060; Duff P. Maternal and Perinatal
infection. In: Gabbe SG, Niebyl JR, Simpson JL, eds. Obstetricsnormal and problem pregnancies, 4th ed.
New York: Churchill-Livingstone 2002:2931348; American College of Obstetricians and Gynecologists.
Perinatal viral and parasitic infections. ACOG Practice Bulletin 20. Washington, DC: ACOG, 2000; American
College of Obstetricians and Gynecologists. Viral hepatitis in pregnancy. ACOG Educational Bulletin 248.
Washington, DC: ACOG, 1998; American College of Obstetricians and Gynecologists. Perinatal herpes simplex
virus infections. ACOG Technical Bulletin 122. Washington, DC: ACOG, 1988; American College of
Obstetricians and Gynecologists. Human immunodeficiency virus infections in pregnancy. ACOG Educational
Bulletin 232. Washington, DC: ACOG, 1997; and Kotler DP. Human immunodeficiency virus and pregnancy.
Gastroenterol Clin North Am 2003;32:437448, ix.
TABLE 14-7 SOME NONVIRAL PERINATAL INFECTIONSa

Organism (Disease) Type of Organism
Neisseria gonorrhoeae Gram-negative

(gonorrhea)
diplococcus
Group B streptococcus Gram-positive
bacteria
Route of
Time of Critical
Potential Severe Fetal or
Transmission
Exposure
Neonatal Effect(s)
Ascending
Intrapartum
Ascending
Intrapartum
Treponema pallidum
(syphilis)
Spirochete
Transplacental
Toxoplasma gondii
(toxoplasmosis)
Intracellular
parasite
Transplacental
Ophthalmia neonatorum
Early onset neonatal

infection: sepsis, meningitis,
pneumonia
Any trimester but Hepatosplenomegaly,
worse risk in first osteitis, hemolytic anemia,
and second
thrombocytopenia,
trimesters
pneumonia, hydrops,
cutaneous or mucous
membrane lesions, hepatitis,
stillbirth
More likely
Stillbirth, hydrocephalus,
transmitted
microcephaly, periventricular
during the third calcifications,
trimester but
hepatosplenomegaly,
more severe
ascites; neonate with
infection with
seizures, chorioretinitis
earlier infection leading to blindness, fever,
rash, hearing loss, mental
retardation
From Bale JF. Congenital infections. Neurol Clin 2002;20:10391060; Duff P. Maternal and Perinatal
Infection. In: Gabbe SG, Niebyl JR, Simpson JL, eds. Obstetricsnormal and problem pregnancies, 4th ed.
New York: Churchill-Livingstone 2002:2931348; American College of Obstetricians and Gynecologists.
Perinatal viral and parasitic infections. ACOG Practice Bulletin 20. Washington, DC: ACOG, 2000; American
College of Obstetricians and Gynecologists. Viral hepatitis in pregnancy. ACOG Educational Bulletin 248.
Washington, DC: ACOG, 1998; American College of Obstetricians and Gynecologists. Prevention of early-onset
group B streptococcal disease in newborns. ACOG Committee Opinion 279. Washington, DC: ACOG, 2002;
Ray JG. Lues-lues: maternal and fetal considerations of syphilis. Obstet Gynecol Surv 1995;50:845850; Ricci
JM, Fojaco RM, O'Sullivan MJ. Congenital syphilis: the University of Miami/Jackson Memorial Medical Center
experience, 19861988. Obstet Gynecol 1989;74:687693.
Pregnancy is no longer an unusual event among women of childbearing age with

functioning renal transplants, occurring in approximately 1 of 20 to 50 women
(193,295). More than 7,000 pregnancies have been reported in renal transplant
recipients (296). Overall, these pregnancies have low success rates, although
repeated successful pregnancies have been reported (297,298). Hypertension
before or during early pregnancy is associated with adverse perinatal events
(295). Typically, pregnancy results in increased renal function, which transiently
declines in late pregnancy. As many as 15% of patients with kidney disease may
experience permanent renal impairment (298). High rates of preterm delivery,
IUGR, hypertension and/or preeclampsia, and stillbirth are reported
(193,295,296,298). The risk of graft rejection, however, is not influenced by
pregnancy (193). Cyclosporine A, a potent anti-T-cell immunosuppressive agent,
has replaced older regimens of azathioprine and steroids in preventing graft
rejection (299). Cyclosporine has been used successfully during pregnancy
without evidence of teratogenicity (299,300,301).
Heart Disease
Maternal cardiac disease may be accompanied by significant maternal and
perinatal morbidity and mortality. Although the etiology of cardiac disease has
changed in the past 30 years, with congenital heart disease now more common
than rheumatic heart disease, the underlying pathophysiology remains the same.
Functional status before or early in pregnancy is an important prognostic indicator
of maternal and fetal outcome. A helpful and commonly used system for assessing
cardiac function is the New York Heart Association classification (Table 14-8)
(302,303). Better prognoses are expected during pregnancy for women with
functional classes I and II than for those with classes III or IV. Gravidae with
severe functional limitations account for 75% of the maternal deaths (304).
Preconception counseling is critical in this group of patients. The added
cardiovascular demands of pregnancy may be associated with cardiac
deterioration: more than 40% of women with heart disease will develop
pulmonary edema for the first time during the third trimester (302,305). Maternal
risks vary with the individual cardiac
P.215
lesion, several specific defects being associated with especially high risks of
maternal mortality (Table 14-9) (302,303,306).
TABLE 14-8 THE NEW YORK HEART ASSOCIATION (NYHA) FUNCTIONAL
CLASSIFICATION OF HEART DISEASEa
Class
Symptoms
Class I
Asymptomatic
Class II
Class III
Symptoms with greater than normal activity

Symptoms with normal activity
Class IV
Symptoms at rest
From American College of Obstetricians and Gynecologists. Cardiac
disease in pregnancy. ACOG Technical Bulletin 168. Washington, DC:

ACOG, 1992; Criteria Committee of the New York Heart Association.
Nomenclature and criteria for diagnosis of diseases of the heart and great
vessels, 8th ed. Boston: Little Brown, 1979.
Fetal risks include premature delivery, IUGR, and stillbirth, especially with
maternal cyanotic heart disease (302,307). A 2% to 5% incidence of fetal cardiac
anomalies has been suggested in women with congenital heart disease, although
with specific lesions, the risk may be as high as 26% (302,306,308). Most of the
drugs used in the treatment of cardiac disease are well tolerated and rarely
associated with significant fetal problems (e.g., beta-blockers, calcium channel
blockers, digitalis, and heparin) (304,306). Newer agents have not been well
studied during pregnancy.
TABLE 14-9 THE EFFECTS OF SOME MATERNAL CONGENITAL HEART
LESIONS DURING PREGNANCYa
Cardiac Lesion
Aortic insufficiency
Aortic stenosis
Ebstein anomaly
Specific Information
Well tolerated
Valve diameter must decrease to 1/3 for
hemodynamic significance; increased risk
of angina, MI, syncope, or sudden death
with severe disease because CO is fixed
and may not be able to compensate;
frequently associated with ischemic heart
disease; most critical time is at pregnancy
termination or delivery; PA catheterization
may be most useful in labor
Right-to-left shunt pulmonary HTN;
association with thromboembolism, CHF,
arrhythmias; 25% of patients have WolfParkinson-White syndrome and increased
risk for tachyarrhythmias
Eisenmenger syndrome
Pulmonary HTN with left-to-right shunt;

grave prognosis; maternal mortality 30%
to 50%; high association with
thromboembolism; therapeutic termination
Marfan syndrome
is recommended
Autosomal dominant; aortic dissection or
aortic or splenic artery aneurysm or
rupture associated with worse outcomes;
50% mortality seen with aortic root
Mitral insufficiency
diameter >40 mm
Well tolerated; CHF rare; pulmonary
edema more likely with preeclampsia
because of increase in afterload; because
of increased risk of atrial enlargement and
fibrillation, prophylactic digitalis
recommended in severe disease
Mitral stenosis
Tachycardia may result in fall in CO and
Mitral valve prolapse
BP; fluid balance crucial; most critical time

is immediately postpartum; hemodynamic
monitoring is recommended
Common; mostly asymptomatic; chest
Tetralogy of Fallot
pain or palpitations possible; responds to

beta-blockers
VSD, overriding aorta, RVH, and
pulmonary stenosis; patients are cyanotic;
maternal mortality 4% to 15% in
uncorrected patients; corrected cases fairly
good outcome
Well tolerated
Tricuspid or pulmonic valve

disease
BP, blood pressure; CHF, congestive heart failure; CO, cardiac output;
HTN, hypertension; MI, myocardial infarction; PA, pulmonary artery; RVH,
right ventricular hypertrophy; VSD, ventricular septal defect.
a
From American College of Obstetricians and Gynecologists. Cardiac

disease in pregnancy. ACOG Technical Bulletin 168. Washington, DC:
ACOG, 1992; Criteria Committee of the New York Heart Association.
Nomenclature and criteria for diagnosis of diseases of the heart and great
vessels, 8thed. Boston: Little Brown, 1979.
Cardiac surgery should be delayed until after completion of pregnancy if possible.
If it is required, however, surgery during the second trimester is preferable (306).
Maternal and perinatal mortality varies with the type of procedure. For example,
in mitral valve commissurotomy, the cardiovascular procedure that has been
performed the most during pregnancy, maternal mortality is under 3% and
perinatal mortality is less than 10%; in contrast, with open-heart surgery,
although maternal mortality is not significantly higher, fetal loss may be as high
as 20% (306).
Myocardial infarction (MI) rarely occurs during pregnancy; the incidence is
estimated at 1 in 10,000 pregnancies (309). Cases show a preponderance during
the third trimester and in multiparous women older than age 33 years (310), with
the majority of MIs located in the anterior wall (310,311). Maternal mortality
ranges from 19% to 37%
P.216
(310,311,312). Fetal mortality is similarly high, most resulting from maternal
death (310,311,312). Given that electrocardiography can be nondiagnostic in 50%
of pregnant women with an acute MI, the use of cardiac troponin I levels has been
found to be reliable in making an accurate diagnosis (309). Acute coronary artery
angioplasty recently was successful during pregnancy (313,314). Treatment with
thrombolytic agents for maternal myocardial infarction has been reported during
the second trimester, although the fetal risk of such therapy has not been
established (315). Some patients have had successful subsequent pregnancies
(316,317).
Successful pregnancies have been reported in women who have undergone heart
transplantation (304,318,319,320,321,322). Reported cases include four patients
who initially received their allograft because of peripartum cardiomyopathy from
an earlier pregnancy (322). Maternal hypertension, pre-eclampsia, and jaundice,
as well as fetal IUGR, are common (322). Immunosuppression issues are the
same as those discussed for patients with renal transplants.
Cancer
Cancer develops in approximately 1 of 1,000 pregnancies. The most common
invasive carcinoma originates in the cervix, affecting one of 2,200 pregnancies
(323,324). Almost 3% of all cervical cancers are diagnosed during pregnancy
(325). The second most common site for malignancy during pregnancy is the
breast, with cancer estimated to occur once in every 1,360 to 3,200 pregnancies
(326). Other frequently seen neoplasias include vulvar, ovarian, and colorectal
carcinoma, as well as leukemia, Hodgkin disease, and melanoma. Stage for stage,
comparing diagnoses in nonpregnant women with those in pregnant women,
carcinoma identified during pregnancy may be more advanced. Currently, this is
thought to reflect a delay in diagnosis, possibly because of the physiologic
changes of pregnancy, and not necessarily because the cancer is more aggressive
during the pregnant state (327). It has never been substantiated that pregnancy
termination alters cancer progression (327). Metastasis to the fetus or placenta

has been reported in fewer than 70 cases, mostly involving the breast, cervix,
leukemia, lymphoma, melanoma, and thyroid (328).
Decisions regarding the management of malignancies during pregnancy are
difficult and involve risk assessment for both mother and fetus. The difficulty of
these decisions is compounded by the timing in gestation. In general, firsttrimester treatment risks spontaneous abortion or fetal malformation, and early
third-trimester therapy places the pregnancy at risk for preterm labor. Surgery
can be performed at any time, although it is preferred during the second
trimester. Radiation or chemotherapy exposure in the first trimester may affect
organogenesis, resulting in either pregnancy loss or fetal structural defects
(329,330,331).
Radiation exposure just prior to and immediately after implantation poses the
greatest risk of spontaneous abortion (330). Animal studies demonstrate an allor-nothing effect because at this early point in development, the few embryonic
cells are not yet differentiated; therefore, the embryo either dies or develops
normally (330). During the first 3 weeks of gestation, radiation exposure of at
least 250 rads (2.5 cGy) poses the greatest risk for miscarriage (332). Exposure
to more than 50 rads (0.5 cGy) is more commonly associated with development of
severe malformations (333); however, congenital anomalies are unusual in
embryos exposed to more than 5-rads (0.05 cGy) (332). Some malformations
associated with radiation exposure during weeks 3 through 10 include
microcephaly, low birth weight, mental retardation, cataracts, retinal
degeneration, and skeletal abnormalities (332,333). Radiation exposure between
11 and 20 weeks of gestation is not commonly associated with fetal malformations
(327,330). Pregnancy termination has been suggested when fetal exposure is
greater than 10 rads (0.1 cGy) (334). In cases in which local irradiation is used to
treat malignancies of the pelvis, spontaneous abortion or fetal demise usually
occurs (325).
Use of cytotoxic chemotherapy is generally avoided in the first trimester, although
data vary for different agents (325,329). Exposure beyond the first trimester has
not been found to be associated with increased risks of congenital anomalies
(325,331). Long-term developmental data are lacking, however. In treating
patients with chemotherapeutic agents, recognition of the physiologic changes of
pregnancy may alter the efficacy, toxicity, and/or dosage regimens.
SUMMARY
Many maternal conditions have relevance for a developing pregnancy. Counseling
pregnant women with underlying medical disorders must encompass the effects of
the illness on the pregnancy, the effects of the pregnancy on the condition, as well
as potential complications of therapeutic interventions, and the risks of possible
premature delivery. Statistics indicating reductions in neonatal mortality rates are

encouraging; however, there is much yet to be learned in perinatology in order to
impact on the unacceptably high rates of premature deliveries and to improve the
health of women before and during pregnancy.
REFERENCES
1. Guyer B, Freedman MA, Strobino DM, et al. Annual summary of vital
statistics: trends in the health of Americans during the 20th century. Pediatrics
2000;106:1307-1317.
2. Arias E, MacDorman MF, Strobino DM, et al. Annual summary of vital statistics
2002. Pediatrics 2003;112:1215-1230.
3. McCormick MC, Shapiro S, Starfield BH. The regionalization of perinatal

services: summary of the evaluation of a national demonstration program.
JAMA 1985;253:799-804.
4. Yeast JD, Poskin M, Stockbauer JW, et al. Changing patterns in

regionalization of perinatal care and the impact on neonatal mortality. Am J
Obstet Gynecol 1998;178:131-135.
5. Hein HA, Lofgren MA. The changing pattern of neonatal mortality in a

regionalized system of perinatal care: a current update. Pediatrics
1999;104:1064-1069.
P.217
6. Wilcox A, Skjrven R, Buekens P, et al. Birth weight and perinatal mortality.
A comparison of the United States and Norway. JAMA 1995;273:709-711.
7. American College of Obstetricians and Gynecologists. Premature rupture of

membranes. ACOG practice bulletin 1. Washington, DC: ACOG, 1998.
8. Mercer BM, Lewis R. Preterm labor and preterm premature rupture of the
membranes. Diagnosis and management. Clin Infect Dis 1997;11:177-201.
9. Stevenson DK, Wright LL, Lemons JA, et al. Very low birth weight outcomes
of the National Institute of Child Health and Human Development Neonatal
Research Network, January 1993 through December 1994. Am J Obstet Gynecol

1998;179:1632-1639.
10. Singh GK, Yu SM. Infant mortality in the United States: trends, differentials,
and projections, 1950 through 2010. Am J Public Health 1995;85:957-964.
11. Amon E, Anderson GD, Sibai BM, et al. Factors responsible for a preterm
delivery of the immature newborn infant (less than or equal to 1000 gm). Am J
Obstet Gynecol 1987;156:1143-1148.
12. Meis PJ, Ernest JM, Moore ML. Causes of low birth weight births in public
and private patients. Am J Obstet Gynecol 1987;156: 1165-1168.
13. Tucker JM, Goldenberg RL, Davis RO, et al. Etiologies of preterm birth in an
indigent population: is prevention a logical expectation? Obstet Gynecol
1991;77:343-347.
14. American College of Obstetricians and Gynecologists. Assessment of risk

factors for preterm birth. ACOG practice bulletin 31. Washington, DC: ACOG,
2001.
15. Centers for Disease Control and Prevention. State-specific trends among
women who did not receive prenatal careUnited States 1980-1992. MMWR
Morb Mortal Wkly Rep 1994;43:939-942.
16. Harger JH, Hsing AW, Tuomala RE, et al. Risk factors for preterm premature
rupture of fetal membranes: a multicenter case-control study. Am J Obstet
Gynecol 1990;163:130-137.
17. Kristensen J, Langhoff-Roos J, Kristensen FB. Implications of idiopathic

preterm delivery for previous and subsequent pregnancies. Obstet Gynecol
1995;86:800-804.
18. Mercer BM, Goldenberg RL, Das A, et al. The preterm prediction study: a
clinical risk assessment system. Am J Obstet Gynecol 1996;174:1885-1895.
19. Romero R, Brody DT, Oyarzun E, et al. Infection and labor. III. Interleukin1: a signal for the onset of parturition. Am J Obstet Gynecol 1989;160:1117-
1123.
20. Mercer BM. Preterm premature rupture of the membranes. Obstet Gynecol
2003:101:178-193.
21. Lee T, Carpenter MW, Heber WW, et al. Preterm premature rupture of
membranes: risks of recurrent complications in the next pregnancy among a
population-based sample of gravid women. Am J Obstet Gynecol 2003;188:209213.
22. Kaltreider DF, Kohl S. Epidemiology of preterm delivery. Clin Obstet Gynecol
1980;23:17-31.
23. Bendon RW, Faye-Peterson O, Pavlova Z, et al. Fetal membrane histology in

preterm premature rupture of membranes: comparison to controls, and
between antibiotic and placebo treatment. The National Institute of Child Health
and Human Development Maternal Fetal Medicine Units Network. Pediatr Dev
Pathol 1999;2:552-558.
24. Cauci S, Hitti J, Noonan C, et al. Vaginal hydrolytic enzymes,

immunoglobulin A against Gardnerella vaginalis toxin, and risk of early preterm
birth among women in preterm labor with bacterial vaginosis or intermediate
flora. Am J Obstet Gynecol 2002;187:877-881.
25. Hillier SL, Nugent RP, Eschenbach DA, et al. Association between bacterial
vaginosis and preterm delivery of a low-birth-weight infant. N Engl J Med
1995;333:1737-1742.
26. Goldenberg RL, Iams JD, Mercer BM, et al. The preterm prediction study:
the value of new vs. standard risk factors in predicting early and all
spontaneous preterm births. NICHD MFMU Network. Am J Public Health
1998;88:233-238.
27. Carey JC, Klebanoff MA, Hauth JC, et al. Metronidazole to prevent preterm
delivery in pregnant women with asymptomatic bacterial vaginosis. National
Institute of Child Health and Human Development Network of Maternal-Fetal
Medicine Units. N Engl J Med 2000;342:534-540.
28. Hauth JC, Gilstrap LC, Hankins GD, et al. Term maternal and neonatal
complications of acute chorioamnionitis. Obstet Gynecol 1985;66:59-62.
29. Wilson JC, Levy DL, Wilds PL. Premature rupture of membranes prior to
term: consequences of nonintervention. Obstet Gynecol 1982;60:601-606.
30. McIntosh N, Harrison A. Prolonged premature rupture of membranes in the

preterm infant: a 7 year study. Eur J Obstet Gynecol Reprod Biol 1994;57:1-6.
31. Goldenberg RL. The management of preterm labor. Obstet Gynecol

2002:100:1020-1037.
32. Wright LL. National Institutes of Health (NIH) Consensus Development

Conference. Effect of corticosteroids for fetal maturation on perinatal outcomes.
Consensus Development Conference, February 28-March 2, 1994, Bethesda,
MD. Am J Obstet Gynecol 1995:173:246-252.
33. National Institutes of Health Consensus Development Panel. Antenatal

corticosteroids revisited: repeat courses-national institutes of health consensus
development conference statement, August 17-18, 2000. Obstet Gynecol
2001;98:144-150.
34. Crowley P. Prophylactic corticosteroids for preterm birth. Cochrane

Database Syst Rev 2000;2:CD000065.
35. Svare J, Langhoff-Roos J, Anderson LF, et al. Ampicillin-metronidazole

treatment in idiopathic preterm labour: a randomised controlled multicentre
trial. Br J Obstet Gynaecol 1997;104: 892-897.
36. Norman K, Pattinson RC, de Souza J, et al. Ampicillin and metronidazole

treatment in preterm labour: a multicentre, randomised controlled trial. Br J
Obstet Gynaecol 1994;101:404-408.
37. King J, Flenady V. Antibiotics for preterm labour with intact membranes.
Cochrane Database Syst Rev 2000;2:CD000246.
38. Mercer B, Miodovnik M, Thurnau G, et al. Antibiotic therapy for reduction of

infant morbidity after preterm premature rupture of the membranes: a
randomized controlled trial. JAMA 1997;278: 989-995.
39. Kenyon SL, Taylor DJ, Tarnow-Mordi W, et al. Broad spectrum antibiotics for
preterm, prelabor rupture of fetal membranes: the ORACLE I randomized trial.
Lancet 2001;357:979-988.
40. Creasy RK, Gummer BA, Liggins GC. System for predicting spontaneous
preterm birth. Obstet Gynecol 1980;55:692-695.
41. Main DM, Richardson DK, Hadley CB, et al. Controlled trial of a preterm
labor detection program: efficacy and costs. Obstet Gynecol 1989;74:873-877.
42. Sakai M, Sasaki Y, Yamagishi N, et al. The preterm labor index and fetal
fibronectin for prediction of preterm delivery with intact membranes. Obstet
Gynecol 2003;101:123-128.
43. Baumgarten K, Gruber W. Tocolyseindex. In: Dudenhausen JW, Saling E,

eds. Perinatale medizin. Stuttgart: Georg Thieme Verlag, 1974:197-199.
44. Katz M, Newman RB, Gill PJ. Assessment of uterine activity in ambulatory
patients at high risk of preterm labor and delivery. Am J Obstet Gynecol
1986;154:44-47.
45. US Preventive Services Task Force. Home uterine activity monitoring for
preterm labor. Policy statement. JAMA 1993;270: 369-370.
46. US Preventive Services Task Force. Home uterine activity monitoring for
preterm labor. Review article. JAMA 1993;270:371-376.
47. Iams JD, Newman RB, Thom EA, et al. The National Institute of Child Health
and Human Development Network of Maternal-Fetal Medicine Units. Frequency
of uterine contractions and the risk of spontaneous preterm delivery. N Engl J
Med 2002;346:250-255.
48. Goldenberg RL, Mercer BM, Meis PJ, et al. The preterm prediction study:
fetal fibronectin testing and spontaneous preterm birth. National Institute of
Child Health and Human Development Maternal-Fetal Medicine Units Network.
49. Peaceman AM, Andrews WW, Thorp JM, et al. Fetal fibronectin as a predictor
of preterm birth in patients with symptoms: a multicenter trial. Am J Obstet
Gynecol 1997;177:13-18.
50. Goldenberg RL, Mercer BM, Iams JD, et al. The preterm prediction study:
patterns of cervicovaginal fetal fibronectin as predictors of spontaneous preterm
delivery. National Institute of Child Health and Human Development MaternalFetal Medicine Units Network. Am J Obstet Gynecol 1997;177:8-12.
51. Goldenberg RL, Thom E, Moawad AH, et al. The preterm prediction study:
fetal fibronectin, bacterial vaginosis, and peripartum infection. National Institute
of Child Health and Human Development Maternal-Fetal Medicine Units
Network. Obstet Gynecol 1996;87:656-660.
P.218
52. American College of Obstetricians and Gynecologists. Fetal fibronectin
preterm labor risk test. ACOG committee opinion 187. Washington, DC: ACOG,
1997.
53. Amon E. Premature labor. In: Reece EA, Hobbins JC, Mahoney MJ, Petrie
RH, eds. Medicine of the fetus and mother. Philadelphia: JB Lippincott,
1992:1404-1429.
54. Papiernik E, Bouyer J, Collin D. Precocious cervical ripening and preterm

labor. Obstet Gynecol 1986;67:238-242.
55. Buekens P, Alexander S, Boutsen M, et al. Randomised controlled trial of

routine cervical examinations in pregnancy. European Community Collaborative
Study Group on Prenatal Screening. Lancet 1994;344:841-844.
56. Gomez R, Galasso M, Romero R, et al. Ultrasonographic examination of the

uterine cervix is better than cervical digital examination as a predictor of the
likelihood of premature delivery in patients with preterm labor and intact
membranes. Am J Obstet Gynecol 1994;171:956-964.
57. Berghella V, Tolosa jE, Kuhlman K, et al. Cervical ultrasonography

compared with manual examination as a predictor of preterm delivery. Am J
58. Iams JD, Goldenberg RL, Meis PJ, et al. The length of the cervix and the risk
of spontaneous premature delivery. National Institute of Child Health and
Human Development Maternal-Fetal Medicine Units Network. N Engl J Med
1996;334:567-572.
59. Iams JD, Paraskos J, Landon MB, et al. Cervical sonography in preterm
labor. Obstet Gynecol 1994;84:40-46.
60. Okitsu O, Mimura T, Nakayama T, et al. Early prediction of preterm delivery

by transvaginal ultrasonography. Ultrasound Obstet Gynecol 1992;2:402-409.
61. Quinn MJ. Vaginal ultrasound and cervical cerclage: a prospective study.
Ultrasound Obstet Gynecol 1992;2:410-416.
62. Iams JD, Goldenberg RL, Mercer BN, et al. The Preterm Prediction Study:
recurrence risk of spontaneous preterm birth. National Institute of Child Health
and Human Development Maternal-Fetal Medicine Units Network. Am J Obstet
Gynecol 1998;178:1035-1040.
63. Guidelines for perinatal care, 4th ed. Elk Grove Village, IL: American
Academy of Pediatrics; and Washington, DC: American College of Obstetricians
and Gynecologists, 1997:75-76.
64. Gravett MG, Nelson HP, DeRouen T, et al. Independent association of

bacterial vaginosis and Chlamydia trachomatis infection with adverse pregnancy
outcome. JAMA 1986;256: 1899-1903.
65. McGregor JA, French JI, Parker R, et al. Prevention of premature birth by
screening and treatment for common genital tract infections: results of a
prospective controlled evaluation. Am J Obstet Gynecol 1995;173:157-167.
66. Regan JA, Klebanoff MA, Nugent RP, et al. Colonization with group B
streptococci in pregnancy and adverse outcome. VIP Study Group. Am J Obstet
Gynecol 1996;174:1354-1360.
67. Centers for Disease Control and Prevention. Prevention of perinatal group B
streptococcal disease: a public health perspective. MMWR Morb Mortal Wkly Rep
1996;45:1-24.
68. Abrams B, Pickett KE. Maternal nutrition. In: Creasy RK, Resnik R, eds.
Maternal-fetal medicine, 4th ed. Philadelphia: WB Saunders, 1999:122-131.
69. American College of Obstetricians and Gynecologists. Nutrition during

pregnancy. ACOG technical bulletin 179. Washington, DC: ACOG, 1993.
70. Johnson JWC, Longmate JA, Frentzen B. Excessive maternal weight and
pregnancy outcome. Am J Obstet Gynecol 1992;167:353-372.
71. Institute of Medicine, Food and Nutrition Board, Committee on Nutritional

Status During Pregnancy and Lactation, Subcommittee on Dietary Intake and
Nutrient Supplements During Pregnancy and Subcommittee on Nutritional
Status and Weight Gain During Pregnancy. Nutrition during pregnancy. Part I.
Weight gain. Washington, DC: National Academies Press, 1990.
72. Castro LC, Avina RL. Maternal obesity and pregnancy outcomes. Curr Opin
73. Lu GC, Rouse DJ, DuBard M, et al. The effect of the increasing prevalence of
maternal obesity on perinatal morbidity. Am J Obstet Gynecol 2001;185:845849.
74. Institute of Medicine, Food and Nutrition Board, Committee on Nutritional

Status During Pregnancy and Lactation, Subcommittee on Dietary Intake and
Nutrient Supplements During Pregnancy and Subcommittee on Nutritional
Status and Weight Gain During Pregnancy. Nutrition during pregnancy. Part II.
Nutrient supplements. Washington, DC: National Academies Press, 1990.
75. American College of Obstetricians and Gynecologists. Vitamin A

supplementation during pregnancy. ACOG committee opinion 112. Washington,
DC: ACOG, 1992.
76. Medical Research Council Vitamin Study Research Group. Prevention of

neural tube defects: results of the Medical Research Council vitamin study.
Lancet 1991;338:131-137.
77. Centers for Disease Control and Prevention. Use of folic acid for prevention
of spina bifida and other neural tube defects1983-1991. MMWR Morb Mortal
Wkly Rep 1991;40:513-516.
78. Centers for Disease Control and Prevention. Recommendations for the use
of folic acid to reduce the number of cases of spina bifida and other neural tube
defects. MMWR Morb Mortal Wkly Rep 1992;41:1-7.
79. Survey alert. Fortified pasta coming. Obstet Gynecol Surv 1997;52:191.
80. National High Blood Pressure Education Program Working Group on high
blood pressure in pregnancy. Am J Obstet Gynecol 2000;183(1):S1-S22.
81. American College of Obstetricians and Gynecologists. Chronic hypertension

in pregnancy. ACOG practice bulletin 29. Washington, DC: ACOG, 2001.
82. Sibai BM. Chronic hypertension in pregnancy. Obstet Gynecol

2002;100:369-377.
83. Livingston JC, Sibai BM. Chronic hypertension in pregnancy. Obstet Gynecol
Clin North Am 2001;28:447-464.
84. Sibai BM, Lindheimer M, Hauth J, et al. Risk factors for preeclampsia,
abruptio placentae, and adverse neonatal outcomes among women with chronic
hypertension. N Engl J Med 1998;339:667-671.
85. American College of Obstetricians and Gynecologists. Diagnosis and

management of preeclampsia and eclampsia. ACOG practice bulletin 33.
Washington, DC: ACOG, 2002.
86. Martin JN, Blake PG, Perry KG, et al. The natural history of HELLP
syndrome: patterns of disease progression and regression. Am J Obstet Gynecol
1991;164:1500-1513.
87. Leduc L, Wheeler JM, Kirshon B, et al. Coagulation profile in severe

preeclampsia. Obstet Gynecol 1992;79:14-18.
88. Madazli R, Budak E, Calay Z, et al. Correlation between placental bed biopsy
findings, vascular cell adhesion molecule and fibronectin levels in preeclampsia. BJOG 2000;107:514-518.
89. Talledo OE, Chesley LC, Zuspan FP. Renin-angiotensin in normal and
toxemic pregnancies. III. Differential sensitivity to angiotensin II and
norepinephrine in toxemia of pregnancy. Am J Obstet Gynecol 1968;100:218222.
90. Gant NF, Daley GL, Chand S, et al. A study of angiotensin II pressor
response throughout primigravid pregnancy. J Clin Invest 1973;52:2682-2689.
91. Burrows RF, Andrew M. Neonatal thrombocytopenia in the hypertensive

disorders of pregnancy. Obstet Gynecol 1990;76: 234-238.
92. Schiff E, Friedman SA, Sibai BM. Conservative management of severe

preeclampsia remote from term. Obstet Gynecol 1994;84: 626-630.
93. Sibai BM, Mercer BM, Schiff E, et al. Aggressive versus expectant
management of severe preeclampsia at 28 to 32 weeks gestation: a randomized
controlled trial. Am J Obstet Gynecol 1994;171: 818-822.
94. The Eclampsia Trial Collaborative Group. Which anticonvulsant for women
with eclampsia? Evidence from the Collaborative Eclampsia Trial. Lancet
1995;345:1455-1463.
95. Lucas MJ, Leveno KJ, Cunningham FG. A comparison of magnesium sulfate
with phenytoin for the prevention of eclampsia. N Engl J Med 1995;333:201205.
96. Magann EF, Bass D, Chauhan SP, et al. Antepartum corticosteroids: disease
stabilization in patients with the syndrome of hemolysis, elevated liver
enzymes, and low platelets (HELLP). Am J Obstet Gynecol 1994;171:1148-1153.
97. O'Brien JM, Milligan DA, Barton JR. Impact of high-dose corticosteroid
therapy for patients with HELLP (hemolysis, elevated liver enzymes, and low
platelet count) syndrome. Am J Obstet Gynecol 2000;183:921-924.
98. Barr M, Cohen MM. ACE inhibitor fetopathy and hypocalvaria: the kidney-
skull connection. Teratology 1991;44:485-495.
99. Hanssens M, Keirse MJ, Vankelecom F, et al. Fetal and neonatal effects of
treatment with angiotensin-converting enzyme inhibitors in pregnancy. Obstet
Gynecol 1991;78:128-135.
P.219
100. Piper JM, Ray WA, Rosa FW. Pregnancy outcome following exposure to
angiotensin-converting enzyme inhibitors. Obstet Gynecol 1992;80:429-432.
101. CLASP (Collaborative Low-Dose Aspirin Study in Pregnancy) Collaborative

Group. CLASP: a randomised trial of low-dose aspirin for the prevention and
treatment of pre-eclampsia among 9364 pregnant women. Lancet
1994;343:619-629.
102. Heyborne KD. Preeclampsia prevention: lessons from the low-dose aspirin
therapy trials. Am J Obstet Gynecol 2000;183: 523-528.
103. Sibai BM. Prevention of preeclampsia: a big disappointment. Am J Obstet

Gynecol 1998;179:1275-1278.
104. Caritis S, Sibai B, Hauth J, et al. Low-dose aspirin to prevent preeclampsia

in women at high risk. National Institute of Child Health and Human
Development Network of Maternal-Fetal Medicine Units. N Engl J Med
1998;338:701-705.
105. Harrington K, Carpenter RG, Goldfrad C, et al. Transvaginal Doppler

ultrasound of the uteroplacental circulation in the early prediction of preeclampsia and intrauterine growth retardation. Br J Obstet Gynaecol
1997;104:674-681.
106. Harrington K, Goldfrad C, Carpenter RG, et al. Transvaginal uterine and

umbilical artery Doppler examination of 12-16 weeks and the subsequent
development of pre-eclampsia and intrauterine growth retardation. Ultrasound
107. Bower SJ, Harrington KF, Schuchter K, et al. Prediction of pre-eclampsia

by abnormal uterine Doppler ultrasound and modification by aspirin. Br J Obstet
Gynaecol 1996;103:625-629.
108. American College of Obstetricians and Gynecologists. Gestational diabetes.

ACOG practice bulletin 30. Washington, DC: ACOG, 2001.
109. Report of the Expert Committee on the Diagnosis and Classification of

Diabetes Mellitus. American Diabetes Association. Expert Committee on the
Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 2001;4[Suppl
1]:S5-S20.
110. White P. Pregnancy complicating diabetes. Am J Med 1949;7: 609-616.
111. Lucas MJ. Diabetes complicating pregnancy. Obstet Gynecol Clin North Am
2001;28:513-536.
112. American College of Obstetricians and Gynecologists. Fetal macrosomia.

ACOG practice bulletin 22. Washington, DC: ACOG, 2000.
113. Alexander GR, Himes JH, Kaufman RB, et al. A United States national
reference for fetal growth. Obstet Gynecol 1996;87: 163-168.
114. Menticoglou SM, Manning FA, Morrison I, et al. Must macrosomic fetuses
be delivered by a cesarean section? A review of outcome for 786 babies $4,500
g. Aust N Z J Obstet Gynaecol 1992; 32:100-103.
115. Lipscomb KR, Gregory K, Shaw K. The outcome of macrosomic infants

weighing at least 4500 grams: Los Angeles county + University of Southern
California experience. Obstet Gynecol 1995;85:558-564.
116. Brard J, Dufour P, Vinatier D, et al. Fetal macrosomia: risk factors and
outcome. A study of the outcome concerning 100 cases 4500 g. Eur J Obstet
Gynecol Reprod Biol 1998;77:51-59.
117. Nesbitt TS, Gilbert WM, Herrchen B. Shoulder dystocia and associated risk
factors with macrosomic infants born in California. Am J Obstet Gynecol
1998;179:476-480.
118. Johnstone FD, Prescott RJ, Steel JM, et al. Clinical and ultrasound
prediction of macrosomia in diabetic pregnancy. Br J Obstet Gynaecol
1996;103:747-754.
119. Sandmire HF. Whither ultrasonic prediction of fetal macrosomia? Obstet

Gynecol 1993;82:860-862.
120. Acker DB, Sachs BP, Friedman EA. Risk factors for shoulder dystocia.
121. Naylor CD, Sermer M, Chen E, et al. Cesarean delivery in relation to birth
weight and gestational glucose tolerance: pathophysiology or practice style?
Toronto Trihospital Gestational Diabetes Investigators. JAMA 1996;275:11651170.
122. Verma A, Mitchell BF, Demianczuk N, et al. Relationship between plasma

glucose levels in glucose-intolerant women and newborn macrosomia. J Matern
Fetal Med 1997;6:187-193.
123. Sermer M, Naylor CD, Gare DJ, et al. Impact of increasing carbohydrate
intolerance on maternal-fetal outcomes in 3637 women without gestational
diabetes. The Toronto Tri-Hospital Gestational Diabetes Project. Am J Obstet
Gynecol 1995;173: 146-156.
124. Adams KM, Li H, Nelson RL, et al. Sequelae of unrecognized gestational

diabetes. Am J Obstet Gynecol 1998;178:1321-1332.
125. Gross TL, Sokol RJ, Williams T, et al. Shoulder dystocia: a fetal-physician
risk. Am J Obstet Gynecol 1987;156:1408-1418.
126. Keller JD, Lopez-Zeno JA, Dooley SL, et al. Shoulder dystocia and birth
trauma in gestational diabetes: a five-year experience. Am J Obstet Gynecol
1991;165:928-930.
127. Meshari AA, DeSilva S, Rahman I. Fetal macrosomia-maternal risks and

fetal outcome. Int J Gynecol Obstet 1990;32:215-222.
128. Sacks DA. Fetal macrosomia and gestational diabetes: what's the problem?
129. Bahar AM. Risk factors and fetal outcome in cases of shoulder dystocia
compared with normal deliveries of a similar birthweight. Br J Obstet Gynaecol
1996;103:868-872.
130. Nocon JJ, McKenzie DK, Thomas LJ, et al. Shoulder dystocia: an analysis of
risks and obstetric maneuvers. Am J Obstet Gynecol 1993;168:1732-1739.
131. Langer O, Berkus MD, Huff RW, et al. Shoulder dystocia: should the fetus
weighing greater than or equal to 4000 grams be delivered by cesarean
section? Am J Obstet Gynecol 1991;165:831-837.
132. McFarland LV, Raskin M, Daling JR, et al. Erb/Duchenne's palsy: a

consequence of fetal macrosomia and method of delivery. Obstet Gynecol
1986;68:784-788.
133. Ecker JL, Greenberg JA, Norwitz ER, et al. Birth weight as a predictor of
brachial plexus injury. Obstet Gynecol 1997;89:643-647.
134. Perlow JH, Wigton T, Hart J, et al. Birth trauma. A five-year review of
incidence and associated perinatal factors. J Reprod Med 1996;41:754-760.
135. Bryant DR, Leonardi MR, Landwehr JB, et al. Limited usefulness of fetal
weight in predicting neonatal brachial plexus injury. Am J Obstet Gynecol
1998;179:686-689.
136. Gordon M, Rich H, Deutschberger J, et al. The immediate and long-term

outcome of obstetric birth trauma. I. Brachial plexus paralysis. Am J Obstet
Gynecol 1973;117:51-56.
137. Hardy AE. Birth injuries of the brachial plexus: incidence and prognosis. J
Bone Joint Surg [Br] 1981;63-B:98-101.
138. Kolderup LB, Laros RK, Musci TJ. Incidence of persistent birth injury in
macrosomic infants: association with mode of delivery. Am J Obstet Gynecol
1997;177:37-41.
139. Oppenheim WL, Davis A, Growdon WA, et al. Clavicle fractures in the
newborn. Clin Orthop 1990;250:176-180.
140. Chez RA, Carlan S, Greenberg SL, et al. Fractured clavicle is an

unavoidable event. Am J Obstet Gynecol 1994;171:797-798.
141. Moore TK, Cayle JE. The amniotic fluid index in normal pregnancy. Am J
Obstet Gynecol 1990;162:1168-1173.
142. Desmedt EJ, Henry OA, Beischer NA. Polyhydramnios and associated
maternal and fetal complications in singleton pregnancies. Br J Obstet Gynaecol
1990;97:1115-1122.
143. Ventura SJ, Martin JA, Curtin SC, et al. Births: final data for 1999. National
vital statistics reports, vol. 49, no. 1. Hyattsville, MD: National Center for Health
Statistics, 2001.
144. Pregnancy outcomes in the Diabetes Control and Complications Trial. Am J

Obstet Gynecol 1996;174:1343-1353.
145. Mills JL, Simpson JL, Driscoll SG, et al. Incidence of spontaneous abortion
among normal women and insulin-dependent diabetic women whose
pregnancies were identified within 21 days of conception. N Engl J Med
1988;319:1617-1623.
146. Dicker D, Feldberg D, Samuel N, et al. Spontaneous abortion in patients

with insulin-dependent diabetes mellitus: the effect of preconceptional diabetic
control. Am J Obstet Gynecol 1988;158: 1161-1164.
147. Miller E, Hare JW, Cloherty JP, et al. Elevated maternal hemoglobin A1c in
early pregnancy and major congenital anomalies in infants of diabetic mothers.
N Engl J Med 1981;304:1331-1334.
148. Miodovnik M, Mimouni F, Siddiqi TA, et al. Periconceptional metabolic

status and risk for spontaneous abortion in insulin-dependent diabetic
pregnancies. Am J Perinatol 1988;4:368-373.
149. Cousins L. Congenital anomalies among infants of diabetic mothers.

Etiology, prevention, diagnosis. Am J Obstet Gynecol 1983;147:333-338.
150. Gabbe SG. Congenital malformations in infants of diabetic mothers. Obstet

Gynecol Surv 1977;32:125-132.
151. Mills JL. Malformations in infants of diabetic mothers. Teratology

1982;25:385-394.
152. Mills J, Baker L, Goldman AS. Malformations in infants of diabetic mothers

occur before the seventh gestational week. Implications for treatment. Diabetes
1979;28:292-293.
P.220
153. Leslie RDG, Pyke DA, John PN, et al. Hemoglobin A1 in diabetic pregnancy.
Lancet 1978;2:958-959.
154. Reller MD, Kaplan S. Hypertrophic cardiomyopathy in infants of diabetic

mothers: an update. Am J Perinatol 1988;5:353-358.
155. Veille J-C, Sivakoff M, Hanson R, et al. Interventricular septal thickness in

fetuses of diabetic mothers. Obstet Gynecol 1992; 79:51-54.
156. Kitzmiller JL, Gavin LA, Gin GD, et al. Preconception care of diabetes.
Glycemic control prevents congenital anomalies. JAMA 1991;265:731-736.
157. Lucas MJ, Leveno KJ, Williams ML, et al. Early pregnancy glycosylated
hemoglobin, severity of diabetes, and fetal malformations. Am J Obstet Gynecol
1989;161:426-431.
158. Coustan DR. Perinatal mortality and morbidity. In: Reece EA, Coustan DR,
eds. Diabetes mellitus in pregnancy, 2nd ed. New York: Churchill Livingstone,
1995:361-367.
159. Richey SD, Sandstad JS, Leveno KJ. Observations concerning unexplained
fetal demise in pregnancy complicated by diabetes mellitus. J Matern Fetal Med
1995;4:169-172.
160. Salvesan DR, Brudenell MJ, Nicolaides KH. Fetal polycythemia and
thrombocytopenia in pregnancies complicated by maternal diabetes mellitus.
Am J Obstet Gynecol 1992;166:1287-1293.
161. Landon MB, Gabbe SG. Fetal surveillance and timing of delivery in
pregnancy complicated by diabetes mellitus. Obstet Gynecol Clin North Am
1996;23:109-123.
162. Bourbon JR, Farrell PM. Fetal lung development in the diabetic pregnancy.
Pediatr Res 1985;19:253-267.
163. Ojomo EO, Coustan DR. Absence of evidence of pulmonary maturity at

amniocentesis in term infants of diabetic mothers. Am J Obstet Gynecol
1990;163:954-957.
164. Cruz AC, Buhi WC, Birk SA, et al. Respiratory distress syndrome with
mature lecithin/sphingomyelin ratios: diabetes mellitus and low Apgar scores.
165. Curet LB, Olson RW, Schneider JM, et al. Effect of diabetes mellitus on
amniotic fluid lecithin/sphingomyelin ratio and respiratory distress syndrome.
166. Fadel HE, Saad SA, Davis H, et al. Fetal lung maturity in diabetic
pregnancies: relation among amniotic fluid insulin, prolactin, and lecithin. Am J
167. Piper JM, Langer O. Does maternal diabetes delay fetal pulmonary
maturity? Am J Obstet Gynecol 1993;168:783-786.
168. Parker CR Jr, Hauth JC, Hankins GD, et al. Endocrine maturation and lung
function in premature neonates of women with diabetes. Am J Obstet Gynecol
1989;160:657-662.
169. Landon MB, Catalano PM, Gabbe SG. Diabetes Mellitus. In: Gabbe SG, ed.
Obstetricsnormal and problem pregnancies, 4th ed. Orlando, FL: Churchill
Livingstone, 2002:1105-1106.
170. Peevy KJ, Landaw SA, Gross SJ. Hyperbilirubinemia in infants of diabetic
mothers. Pediatrics 1980;66:417-419.
171. Widness JA, Cowett RM, Coustan DR, et al. Neonatal morbidities in infants
of mothers with glucose intolerance in pregnancy. Diabetes 1985;34[Suppl
2]:61-65.
172. Report of the Expert Committee on the Diagnosis and Classification of

Diabetes Mellitus. Diabetes Care 1997;20: 1183-1197.
173. Metzger BE, Coustan DR. The Organizing Committee. Summary and
recommendations of the Fourth International Workshop-Conference on
Gestational Diabetes Mellitus. Diabetes Care 1998;21:B161-B167.
174. O'Sullivan JB. Body weight and subsequent diabetes mellitus. JAMA
1982;248:949-952.
175. Langer O, Conway DL, Berkus MD, et al. A comparison of glyburide and
insulin in women with gestational diabetes mellitus. N Engl J Med
2000;343:1134-1138.
176. Petri M. Hopkins Lupus Pregnancy Center: 1987 to 1996. Rheum Dis Clin
North Am 1997;23:1-13.
177. Mascola MA, Repke JT. Obstetric management of the high-risk lupus
pregnancy. Rheum Dis Clin North Am 1997;23:119-132.
178. Le Huong D, Wechsler B, Vauthier-Brouzes D, et al. Outcome of planned

pregnancies in systemic lupus erythematosus: a prospective study on 62
pregnancies. Br J Rheumatol 1997;36:772-777.
179. Lima F, Buchanan NM, Khamashta MA, et al. Obstetric outcome in

systemic lupus erythematosus. Semin Arthritis Rheum 1995;25: 184-192.
180. Mintz G, Niz J, Gutierrez G, et al. Prospective study of pregnancy in

systemic lupus erythematosus. Results of a multidisciplinary approach. J
Rheumatol 1986;13:732-739.
181. Out HJ, Derksen RH, Christiaens GC. Systemic lupus erythematosus and
pregnancy. Obstet Gynecol Surv 1989;44:585-591.
182. Faussett MF, Branch DW. Autoimmunity and pregnancy loss. Semin Reprod
Med 2000;18:379-392.
183. Johnson MJ, Petri M, Witter FR, et al. Evaluation of preterm delivery in a
systemic lupus erythematosus pregnancy clinic. Obstet Gynecol 1995;86:396399.
184. Font J, Lopez-Soto A, Cervera R, et al. Antibodies to thromboplastin in

systemic lupus erythematosus: isotype distribution and clinical significance in a
series of 92 patients. Thromb Res 1997;86:37-48.
185. Ogasawara M, Aoki K, Hayashi Y. A prospective study on pregnancy risk of

antiphospholipid antibodies in association with systemic lupus erythematosus. J
Reprod Immunol 1995;28: 159-164.
186. Lockshin MD. Pregnancy does not cause systemic lupus erythematosus to
worsen. Arthritis Rheum 1989;32:665-670.
187. Lockshin MD, Reinitz E, Druzin ML, et al. Lupus pregnancy. Case-control
prospective study demonstrating absence of lupus exacerbation during or after
pregnancy. Am J Med 1984;77:893-898.
188. Ramsey-Goldman R, Schilling E. Immunosuppressive drug use during

pregnancy. Rheum Dis Clin North Am 1997;23:149-167.
189. Meehan RT, Dorsey JK. Pregnancy among patients with systemic lupus
erythematosus receiving immunosuppressive therapy. J Rheumatol
1987;14:252-258.
190. Lockshin MD. Pregnancy associated with systemic lupus erythematosus.

Semin Perinatol 1990;14:130-138.
191. Stuart MJ, Gross SJ, Elrad H, et al. Effects of acetylsalicylic-acid ingestion
on maternal and neonatal hemostasis. N Engl J Med 1982;307:909-912.
192. Alstead EM, Ritchie JK, Lennard-Jones JE, et al. Safety of azathioprine in
pregnancy in inflammatory bowel disease. Gastroenterology 1990;99:443-446.
193. Davison JM. Dialysis, transplantation, and pregnancy. Am J Kidney Dis

1991;17:127-134.
194. Pilarski LM, Yacyshyn BR, Lazarovits AI. Analysis of peripheral blood
lymphocyte populations and immune function from children exposed to
cyclosporine or to azathioprine in utero. Transplantation 1994;57:133-144.
195. Khamashta MA, Buchanan NM, Hughes GR. The use of hydroxychloroquine
in lupus pregnancy: the British experience. Lupus 1996;5[Suppl 1]:S65-S66.
196. Parke A, West B. Hydroxychloroquine in pregnant patients with systemic

lupus erythematosus. J Rheumatol 1996;23:1715-1718.
197. Parke AL, Rothfield NF. Antimalarial drugs in pregnancythe North

American experience. Lupus 1996;5[Suppl 1]:S67-S69.
198. Phillips-Howard PA, Wood D. The safety of antimalarial drugs in pregnancy.

Drug Saf 1996;14:131-145.
199. Shumak KH, Rock GA. Therapeutic plasma exchange. N Engl J Med
1984;310:762-771.
200. Wei N, Klippel JH, Huston DP, et al. Randomised trial of plasma exchange
in mild systemic lupus erythematosus. Lancet 1983;1:17-22.
201. Tseng C-E, Buyon JP. Neonatal lupus syndromes. Rheum Dis Clin North Am
1997;23:31-54.
202. Watson RM, Lane AT, Barnett NK, et al. Neonatal lupus erythematosus. A
clinical, serological and immunogenetic study with review of the literature.
Medicine (Baltimore) 1984;63: 362-378.
203. Olah KS, Gee H. Fetal heart block associated with maternal anti-Ro (SSA)
antibody-current management. A review. Br J Obstet Gynaecol 1991;98:751755.
204. Reichlin M, Friday K, Harley JB. Complete congenital heart block followed
by anti-Ro-SS-A in adult life. Studies of an informative family. Am J Med
1988;84:339-344.
205. Buyon JP, Waltuck J, Kleinman C, et al. In utero identification and therapy
of congenital heart block. Lupus 1995;4:116-121.
206. Richards DS, Wagman AJ, Cabaniss ML. Ascites not due to congestive
heart failure in a fetus with lupus-induced heart block. Obstet Gynecol
1990;76:957-959.
207. Copel JA, Buyon JP, Kleinman CS. Successful in utero therapy of fetal heart
block. Am J Obstet Gynecol 1995;173:1384-1390.
208. Copel JA. Management of fetal cardiac arrhythmias. Obstet Gynecol Clin
North Am 1997;24:201-211.
209. Watson WJ, Katz VL. Steroid therapy for hydrops associated with antibodymediated congenital heart block. Am J Obstet Gynecol 1991;165:553-554.
P.221
210. Herreman G, Galezewski N. Maternal connective tissue disease and
congenital heart block [Letter]. N Engl J Med 1985;312: 1328-1329.
211. Walkinshaw SA, Welch CR, McCormack J, et al. In utero pacing for fetal
congenital heart block. Fetal Diagn Ther 1994; 9:183-185.
212. McCarron DP, Hellmann DB, Traill TA, et al. Neonatal lupus erythematosus
syndrome: late detection of isolated heart block. J Rheumatol 1993;103:12121214.
213. Reed BR, Lee LA, Harmon C, et al. Autoantibodies to SS-A/Ro in infants
with congenital heart block. J Pediatr 1983;103: 889-891.
214. Eronen M, Sirn MK, Ekblad H, et al. Short- and long-term outcome of
children with congenital complete heart block diagnosed in utero or as a
newborn. Pediatrics 2000;106:86-91.
215. Rider LG, Buyon JP, Rutledge J, et al. Treatment of neonatal lupus: case
report and review of the literature. J Rheumatol 1993; 20:1208-1211.
216. Davison MB, Radford DJ. Fetal and neonatal congenital complete heart
block. Med J Aust 1989;150:192-198.
217. Dedhia HV, DiBartolomeo A. Rheumatoid arthritis. Crit Care Clin

2002;18:841-854, ix.
218. Arnett FC, Edworthy SM, Bloch DA, et al. The American Rheumatism
Association 1987 revised criteria for the classification of rheumatoid arthritis.
Arthritis Rheum 1988;31:315-324.
219. Nelson JL, Ostensen M. Pregnancy and rheumatoid arthritis. Rheum Dis
Clin North Am 1997;23:195-212.
220. Johnson MJ Obstetric complications and rheumatic disease. Rheum Dis Clin
North Am 1997;23:169-182.
221. Buchanan WW, Needs CJ, Brooks PM. Rheumatic diseases: the
arthropathies. In: Gleicher N, ed. Principles and practice of medical therapy in
pregnancy, 3rd ed. Norwalk: Appleton & Lange, 1998:538-545.
222. Steen VD. Pregnancy in women with systemic sclerosis. Obstet Gynecol
1999;94:15-20.
223. Maymon R, Fejgin M. Scleroderma in pregnancy. Obstet Gynecol Surv

1989;44:530-534.
224. Belilos E, Carsons S. Rheumatologic disorders in women. Med Clin North

Am 1998;82:77-101.
225. Julkunen H, Kaaja R, Kurki P, et al. Fetal outcome in women with primary
Sjgren's syndrome. A retrospective case-control study. Clin Exp Rheumatol

1995;13:65-71.
226. American College of Obstetricians and Gynecologists. Antiphospholipid

syndrome. ACOG educational bulletin 244. Washington, DC: ACOG, 1998.
227. Branch DW. Antiphospholipid syndrome: laboratory concerns, fetal loss,

and pregnancy management. Semin Perinatol 1991;15:230-237.
228. Conley CL, Hartmann RD. A hemorrhagic disorder caused by circulating

anticoagulant in patients with disseminated lupus erythematosus. J Clin Invest
1952;31:621-622.
229. Love PE, Santoro SA. Antiphospholipid antibodies: anticardiolipin and the
lupus anticoagulant in systemic lupus erythematosus (SLE) and in non-SLE
disorders. Prevalence and clinical significance. Ann Intern Med 1990;112:682698.
230. Ninomiya C, Taniguchi O, Kato T, et al. Distribution and clinical significance

of lupus anticoagulant and anticardiolipin antibody in 349 patients with systemic
lupus erythematosus. Intern Med 1992;31:194-199.
231. Harris EN. Syndrome of the black swan. Br J Rheumatol 1987;26: 324-326.
232. Hughes GRV, Harris EN, Gharavi AE. The anticardiolipin syndrome. J
Rheumatol 1986;13:486-489.
233. Lockshin MD. Antiphospholipid antibody. Babies, blood clots, biology. JAMA
1997;277:1549-1551.
234. Kniaz D, Eisenberg GM, Elrad H, et al. Postpartum hemolytic uremic

syndrome associated with antiphospholipid antibodies. A case report and review
of the literature. Am J Nephrol 1992; 12:126-133.
235. Silver RM, Draper ML, Scott JR, et al. Clinical consequences of
antiphospholipid antibodies: an historic cohort study. Obstet Gynecol
1994;83:372-377.
236. Sheridan-Pereira M, Porreco RP, Hays T, et al. Neonatal aortic thrombosis

associated with the lupus anticoagulant. Obstet Gynecol 1988;71:1016-1018.
237. Silver RK, MacGregor SN, Pasternak JF, et al. Fetal stroke associated with
elevated maternal anticardiolipin antibodies. Obstet Gynecol 1992;80:497-499.
238. Lubbe WF, Butler WS, Palmer SJ, et al. Fetal survival after prednisone
suppression of maternal lupus-anticoagulant. Lancet 1983;1(8338):1361-1363.
239. Lubbe WF, Liggins GC. Lupus anticoagulant and pregnancy. Am J Obstet
Gynecol 1985;153:322-327.
240. Cowchock FS, Reece EA, Balaban D, et al. Repeated fetal losses associated
with antiphospholipid antibodies: a collaborative randomized trial comparing
prednisone with low-dose heparin treatment. Am J Obstet Gynecol
1992;166:1318-1323.
241. Landy HJ, Isada NB, McGinnis J, et al. The effect of chronic steroid therapy
on glucose tolerance in pregnancy. Am J Obstet Gynecol 1988;159:612-615.
242. Landy HJ, Kessler C, Kelly WK, et al. Obstetric performance in patients
with the lupus anticoagulant and/or anticardiolipin antibodies. Am J Perinatol
1992;9:146-151.
243. Kutteh WH. Antiphospholipid antibody-associated recurrent pregnancy

loss: treatment with heparin and low-dose aspirin is superior to low-dose aspirin
alone. Am J Obstet Gynecol 1996; 174:1584-1589.
244. Rosove MH, Tabsh K, Wasserstrum N, et al. Heparin therapy for pregnant
women with lupus anticoagulant or anticardiolipin antibodies. Obstet Gynecol
1990;75:630-634.
245. Nelson-Piercy C, Letsky EA, de Swiet M. Low-molecular-weight heparin for

obstetric thromboprophylaxis: experience of sixty-nine pregnancies in sixty-one
women at high risk. Am J Obstet Gynecol 1997;176:1062-1068.
246. Branch DW, Peaceman AM, Druzin M, et al. A multicenter, placebocontrolled pilot study of intravenous immune globulin treatment of
antiphospholipid syndrome during pregnancy. The Pregnancy Loss Study Group.

Am J Obstet Gynecol 2000;18: 122-127.
247. Branch DW, Porter TF, Paidas MJ, et al. Obstetric uses of intravenous
immunoglobulin: successes, failures, and promises. J Allergy Clin Immunol
2001;108:S133-S138.
248. American College of Obstetricians and Gynecologists. Thyroid disease in

pregnancy. ACOG practice bulletin 244. Washington, DC: ACOG, 2002.
249. Mestman JH, Goodwin TM, Montoro MM. Thyroid disorders of pregnancy.
Endocrinol Metab Clin North Am 1995;24:41-71.
250. Mazzaferri EL. Evaluation and management of common thyroid disorders in

women. Am J Obstet Gynecol 1997;176: 507-514.
251. Wing DA, Millar LK, Koonings PP, et al. A comparison of propylthiouracil
versus methimazole in the treatment of hyperthyroidism in pregnancy. Am J
Obstet Gynecol 1994;170: 90-95.
252. Van Dijke CP, Heydendael RJ, De Kleine MJ. Methimazole, carbimazole and
congenital skin defects. Ann Intern Med 1987; 106:60-61.
253. Matsuura N, Konishi J, Fujieda K, et al. TSH-receptor antibodies in mothers

with Graves' disease and outcome in their offspring. Lancet 1988;1:14-17.
254. Houck JA, Davis RE, Sharma HM. Thyroid-stimulating im-munoglobulin as

a cause of recurrent intrauterine fetal death. Obstet Gynecol 1988;71:10181019.
255. Wenstrom KD, Weiner CP, Williamson RA, et al. Prenatal diagnosis of fetal
hyperthyroidism using funipuncture. Obstet Gynecol 1990;76:513-517.
256. Porreco RP, Bloch CA. Fetal blood sampling in the management of
intrauterine thyrotoxicosis. Obstet Gynecol 1990;76:509-512.
257. Toft AD. Drug therapy: thyroxine therapy. N Engl J Med 1994; 154:785-
787.
258. American Academy of Pediatrics AAP Section on Endocrinology and

Committee on Genetics, and American Thyroid Association Committee on Public
Health: newborn screening for congenital hypothyroidism: recommended
guidelines. Pediatrics 1993;91: 1203-1209.
259. Bale JF. Congenital infections. Neurol Clin 2002;20:1039-1060.
260. Duff P. Maternal and Perinatal Infection. In: Gabbe SG, Niebyl JR, Simpson
JL, eds. Obstetricsnormal and problem pregnancies, 4th ed. New York:
Churchill-Livingstone, 2002:293-1348.
261. American College of Obstetricians and Gynecologists. Perinatal viral and

parasitic infections. ACOG practice bulletin 20. Washington, DC: ACOG, 2000.
262. American College of Obstetricians and Gynecologists. Viral hepatitis in

pregnancy. ACOG educational bulletin 248. Washington, DC: ACOG, 1998.
P.222
263. American College of Obstetricians and Gynecologists. Perinatal herpes
simplex virus infections. ACOG technical bulletin 122. Washington, DC: ACOG,
1988.
264. American College of Obstetricians and Gynecologists. Human

immunodeficiency virus infections in pregnancy. ACOG educational bulletin 232.
265. Kotler DP. Human immunodeficiency virus and pregnancy. Gastroenterol

Clin North Am 2003;32:437-448, ix.
266. American College of Obstetricians and Gynecologists. Prevention of earlyonset Group B streptococcal disease in newborns. ACOG committee opinion
279. Washington, DC: ACOG, 2002.
267. Ray JG. Lues-lues: maternal and fetal considerations of syphilis. Obstet
Gynecol Surv 1995;50:845-850.
268. Ricci JM, Fojaco RM, O'Sullivan MJ. Congenital syphilis: the University of
Miami/Jackson Memorial Medical Center experience, 1986-1988. Obstet Gynecol
1989;74:687-693.
269. American College of Obstetricians and Gynecologists. Thromboembolism in

pregnancy. ACOG practice bulletin 19. Washington, DC: ACOG, 2000.
270. Toglia MR, Weg JG. Venous thromboembolism during pregnancy. N Engl J
Med 1996;335:108-114.
271. Barbour LA, Pickard J. Controversies in thromboembolic disease during

pregnancy: a critical review. Obstet Gynecol 1995;86: 621-633.
272. Lockwood CJ. Inherited thrombophilias in pregnant patients: detection and

treatment paradigm. Obstet Gynecol 2002;99: 333-341.
273. Lockwood CJ. Heritable coagulopathies in pregnancy. Obstet Gynecol Surv

1999;54:754-765.
274. den Heijer M, Koster T, Blom HJ, et al. Hyperhomocysteinemia as a risk

factor for deep-vein thrombosis. N Engl J Med 1996;334: 759-762.
275. Dizon-Townson DS, Nelson LM, Jang H, et al. The incidence of factor V
Leiden mutation in an obstetric population and its relationship to deep vein
thrombosis. Am J Obstet Gynecol 1997;176: 883-886.
276. Rouse DJ, Goldenberg RL, Wenstrom KD. Antenatal screening for factor V
Leiden mutation: a critical appraisal. Obstet Gynecol 1997;90:848-851.
277. Weinmann EE, Salzman EW. Deep-vein thrombosis. N Engl J Med

1994;331:1630-1641.
278. The PIOPED Investigators. Value of the ventilation/perfusion scan in acute

pulmonary embolism: results of the Prospective Investigation of Pulmonary
Embolism Diagnosis (PIOPED). JAMA 1990;263:2753-2759.
279. Powell T, Mller NL. Imaging of acute pulmonary thromboembolism:

should spiral computed tomography replace the ventilation-perfusion scan? Clin
Chest Med 2003;24:29-38.
280. Ginsberg JS, Hirsh J. Use of antithrombotic agents during pregnancy.

Chest 1995;108[Suppl 4]:305S-311S.
281. American College of Obstetricians and Gynecologists. Prevention of deep

vein thrombosis and pulmonary embolism. ACOG practice bulletin 21.
282. Hall JG, Pauli RM, Wilson KM. Maternal and fetal sequelae of
anticoagulation during pregnancy. Am J Med 1980;68:122-140.
283. Wong V, Cheng CH, Chan KC. Fetal and neonatal outcome of exposure to
anticoagulants during pregnancy. Am J Med Genet 1993;45:17-21.
284. Stevenson RE, Burton OM, Ferlauto GJ, et al. Hazards of oral
anticoagulants during pregnancy. JAMA 1980;243:1549-1551.
285. Cunningham FG, Cox SM, Harstad TW, et al. Chronic renal disease and
pregnancy outcome. Am J Obstet Gynecol 1990;163:453-459.
286. Hou SH, Grossman SD, Madias NE. Pregnancy in women with renal disease
and moderate renal insufficiency. Am J Med 1985; 78:185-194.
287. Jones DC, Hayslett JP. Outcome of pregnancy in women with moderate or
severe renal insufficiency. N Engl J Med 1996;335: 226-232.
288. Krakow D, Castro LC, Schwieger J. Effect of hemodialysis on uterine and

umbilical artery Doppler flow velocity waveforms. Am J Obstet Gynecol
1994;170:1386-1388.
289. Elliott JP, O'Keeffe DF, Schon DA, et al. Dialysis in pregnancy: a critical
review. Obstet Gynecol Surv 1992;46:319-324.
290. Hou SH. Pregnancy in women on haemodialysis and peritoneal dialysis.
Ballieres Clin Obstet Gynaecol 1994;8:481-500.
291. Jakobi P, Ohel G, Szylman P, et al. Continuous ambulatory peritoneal

dialysis as the primary approach in the management of severe renal
insufficiency in pregnancy. Obstet Gynecol 1992;79: 808-810.
292. Nageotte MP, Grundy HO. Pregnancy outcome in women requiring chronic
hemodialysis. Obstet Gynecol 1988;72:456-459.
293. Yasin SY, Beydoun SN. Hemodialysis in pregnancy. Obstet Gynecol Surv
1988;43:655-668.
294. Tison A, Lozowy C, Benjamin A, et al. Successful pregnancy complicated

by peritonitis in a 35-year old CAPD patient. Perit Dial Int 1996;16[Suppl 1]:
S489-S491.
295. Sturgiss SN, Davison JM. Perinatal outcome in renal allograft recipients:
prognostic significance of hypertension and renal function before and during
pregnancy. Obstet Gynecol 1991;78:573- 577.
296. Cohen D, Galbraith C. General health management and long-term care of

the renal transplant recipient. Am J Kidney Dis 2001;38[Suppl 6]:S10-S24.
297. Ehrich JH, Loirat C, Davison JM, et al. Repeated successful pregnancies
after kidney transplantation in 102 women (report by the EDTA registry).
Nephrol Dial Transplant 1996;11:1314-1317.
298. Davison JM. Renal transplantation and pregnancy. Am J Kidney Dis

1987;9:374-380.
299. Burrows DA, O'Neil TJ, Sorrells TL. Successful twin pregnancy after renal
transplant maintained on cyclosporine A immunosuppression. Obstet Gynecol
1988;72:459-461.
300. Gaughan WJ, Moritz MJ, Radomski JS, et al. National Transplantation
Pregnancy Registry: report on outcomes in cyclosporine-treated female kidney
transplant recipients with an interval from transplant to pregnancy of greater
than five years. Am J Kidney Dis 1996;28:266-269.
301. Olshan AF, Mattison DR, Zwanenburg TS. International Commission for
Protection Against Environmental Mutagens and Carcinogens. Cyclosporine A:
review of genotoxicity and potential for adverse human reproductive and
developmental effects. Report of a Working Group on the genotoxicityof
cyclosporine A, August 18, 1993. Mutat Res 1994;317:163-173.
302. American College of Obstetricians and Gynecologists. Cardiac disease in

pregnancy. ACOG technical bulletin 168. Washington, DC: ACOG, 1992.
303. Criteria Committee of the New York Heart Association. Nomenclature and
criteria for diagnosis of diseases of the heart and great vessels, 8th ed. Boston:
Little Brown, 1979.
304. Gei AF, Hankins GDV. Cardiac disease and pregnancy. Obstet Gynecol Clin
North Am 2001;28:465-512.
305. Szekely P, Turner R, Snaith L. Pregnancy and the changing pattern of

rheumatic heart disease. Br Heart J 1973;35: 1293-1303.
306. McAnulty JH, Morton MJ, Ueland K. The heart and pregnancy. Curr Probl
Cardiol 1988;9:589-660.
307. Patton DE, Lee W, Cotton DB, et al. Cyanotic maternal heart disease in
pregnancy. Obstet Gynecol Surv 1990;45:594-600.
308. Whittemore R, Hobbins JC, Engle MA. Pregnancy and its outcome in
women with and without surgical treatment of congenital heart disease. Am J
Cardiol 1982;50:641-651.
309. Shade GH, Ross G, Bever FN, et al. Troponin I in the diagnosis of acute
myocardial infarction in pregnancy, labor, and post partum. Am J Obstet
Gynecol 2002;187:1719-1720.
310. Roth A, Elkayam U. Acute myocardial infarction associated with pregnancy.

Ann Intern Med 1996;125:751-762.
311. Badui E, Enciso R. Acute myocardial infarction during pregnancy and
puerperium: a review. Angiology 1996;47:739-756.
312. Hankins GDV, Wendel GD, Leveno KJ, et al. Myocardial infarction during
pregnancy: a review. Obstet Gynecol 1985;65:139-146.
313. Ascarelli MH, Grider AR, Hsu HW. Acute myocardial infarction during
pregnancy managed with immediate percutaneous transluminal coronary
angioplasty. Obstet Gynecol 1996;88: 655-657.
314. Eikman FM. Acute coronary artery angioplasty during pregnancy. Cathet
Cardiovasc Diagn 1996;38:369-372.
315. Schumacher B, Belfort MA, Card RJ. Successful treatment of acute

myocardial infarction during pregnancy with tissue plasminogen activator. Am J
316. Dufour P, Berard J, Vinatier D, et al. Pregnancy after myocardial infarction

and a coronary artery bypass graft. Arch Gynecol Obstet 1997;259:209-213.
317. Frenkel Y, Barkai G, Reisin L, et al. Pregnancy after myocardial infarction:

are we playing safe? Obstet Gynecol 1991;77:822-825.
P.223
318. Baxi LV, Rho RB. Pregnancy after cardiac transplantation. Am J Obstet
Gynecol 1993;169:33-34.
319. Key TC, Resnik R, Dittrich HC, et al. Successful pregnancy after cardiac
transplantation. Am J Obstet Gynecol 1989;160:367-371.
320. Kirk EP. Organ transplantation and pregnancy: a case report and review.
321. Lwenstein BR, Vain NW, Perrone SV, et al. Successful pregnancy and
vaginal delivery after heart transplantation. Am J Obstet Gynecol 1988;158:589590.
322. Scott JR, Wagoner LE, Olsen SL, et al. Pregnancy in heart transplant
recipients: management and outcome. Obstet Gynecol 1993;82:324-327.
323. Antonelli NM, Dotters DJ, Katz VL, et al. Cancer in pregnancy: a review of
the literature. Part I. Obstet Gynecol Surv 1996;125-134.
324. Hacker NF, Berek JS, Lagasse LD, et al. Carcinoma of the cervix associated
with pregnancy. Obstet Gynecol 1982;59:735-746.
325. Berman ML, DiSaia PJ, Brewster WR. Pelvic malignancies, gestational
trophoblastic neoplasia, and nonpelvic malignancies. In: Creasy RK, Resnik R,
eds. Maternal-fetal medicine, 4th ed. Philadelphia: WB Saunders, 1999:11281150.
326. Donegan WL. Breast cancer and pregnancy. Obstet Gynecol 1977;50:244252.
327. Schwartz PE. Cancer in pregnancy. In: Reece EA, Hobbins JC, Mahoney MJ,
Petrie RH, eds. Medicine of the fetus and mother. Philadelphia: JB Lippincott,
1992:1257-1281.
328. Dildy GA, Moise KJ, Carpenter RJ, et al. Maternal malignancy metastatic to
the products of conception: a review. Obstet Gynecol Surv 1989;44:535-540.
329. Buekers TE, Lallas TA. Chemotherapy in pregnancy. Obstet Gynecol Clin
North Am 1998;25:323-329.
330. Mayr NA, Wen BC, Saw CB. Radiation therapy during pregnancy. Obstet
Gynecol Clin North Am 1998;25:301-321.
331. Sweet DL, Kinzie J. Consequences of radiotherapy and antineoplastic

therapy for the fetus. J Reprod Med 1976;17:241-246.
332. Debakan A. Abnormalities in children exposed to x-irradiation during

various stages of gestation: tentative timetable of radiation injury to the human
fetus. Part I. J Nucl Med 1968;9: 471-477.
333. Brent RC. The effect of embryonic and fetal exposure to x-ray,
microwaves, and ultrasound: counseling the pregnant and non-pregnant patient

about these risks. Semin Oncol 1989;16: 347-368.
334. Orr JW Jr, Shingleton HM. Cancer in pregnancy. Curr Prob Cancer
1983;8:1-50.
Chapter 15
The Effects of Maternal Drugs on the
Developing Fetus
David A. Beckman
Lynda B. Fawcett
Robert L. Brent
Every conception has a risk of ending in abortion or serious congenital anomaly
(Tables 15-1,15-2). Furthermore it is axiomatic that every drug administered or
taken by a pregnant woman presents the mother and fetus with both risks and
benefits. The controversies in this field are primarily related to the nature and
magnitude of the risks for these drugs.
Abortion and birth defects have some common etiologies, but in many instances
the causes of these two areas of adverse reproductive outcome are divergent.
Most human teratogens affect the embryo during a relatively narrow period of
early embryonic development (18-40 days for major malformations excluding
genital malformations and cleft palate which have longer periods of sensitivity).
However, there are a few teratogens and many fetotoxic agents that have
deleterious effects during the second and even the third trimester.
In this chapter, we evaluate the data concerning the potential risks of selected
prescribed and self-administered drugs in human pregnancy. The evaluations were
made after a review of the available clinical, epidemiological and experimental
data and an analysis based on reproducibility, consistency and biological
plausibility. Only key references or reviews are cited which will guide the reader to
additional relevant literature.
CHARACTERIZATION OF ADVERSE
REPRODUCTIVE OUTCOMES
Spontaneous Abortion
The definition of spontaneous abortion is based on the stage of embryonic
development when viability was not possible outside the uterus. This stage is
presently considered to be 20 weeks or less of gestation and a fetal weight of less
than 500 grams, although these criteria are not universally accepted.
The frequency of spontaneous abortion varies with the stage of gestation Tables
15-2: more than 80% of abortions occur in the first trimester and there is a
steady decline in the risk of abortion as pregnancy progresses. Therefore it is

essential that epidemiological studies into the cause of abortion compare control
and exposed populations with the same mean stage and range of abortion. Two
pregnant populations with a 2-week difference in mean stage of pregnancy will
have a different background incidence of abortion. Abortion in human populations
include the following causes Tables 15-3:
Chromosomal Abnormalities
The earlier the abortion, the higher the proportion of chromosomal abnormalities
(1,2). Approximately 53% of spontaneous abortions in the first trimester are as a
result of chromosomal abnormalities, 36% are as a result of chromosomal
abnormalities in the second trimester and only 5% of stillbirths in the third
trimester are as a result of chromosomal abnormalities. Over 95% of abortuses
with chromosomal abnormalities were as a result of autosomal trisomy, double
trisomy, monosomy, triploidy or tetraploidy (3,4). Most chromosomal
abnormalities are not the cause of repetitive abortion, although in about 4% of
couples with two or more spontaneous abortions, a normal-appearing parent could
be a carrier for a balanced translocation or may be a mosaic with abnormal cells in
the germ cell line. Environmental exposures during pregnancy cannot account for
any of these abortions because most aneuploidies result from meiotic
nondisjunction during gametogenesis before conception.
TABLE 15-1 FREQUENCY OF REPRODUCTIVE RISKS IN THE HUMAN
Reproductive Risk
Immunologically and clinically diagnosed spontaneous
Frequency
350,000
abortions per 106 conceptions

Clinically recognized spontaneous abortions per 106
pregnancies
150,000
Genetic diseases per 106 births

Multifactorial or polygenic (genetic-environmental
interactions)
Dominantly inherited disease
Autosomal and sex-linked genetic disease
Cytogenetic (chromosomal abnormalities)
New mutations
110,000
90,000
10,000
1,200
5,000
3,000
Major congenital malformations per 106 births
30,000
Prematurity per 106 births
40,000
Fetal growth retardation per 106 births
30,000
Stillbirths per 106 pregnancies (>20 wks)
20,900
Modified from Brent RL. Environmental factors: miscellaneous. In: BrentRL,

Harris ML, eds. Prevention of embryonic fetal and perinatal disease.
Bethesda: DHEW (NIH), 1976:211-218, with permission.
P.225
Abortions With Normal Chromosomes (Euploidy)

Hertig (1) and many other investigators reported the occurrence of malformed or
blighted embryos as a cause of abortion. These embryonic losses may occur later
in the first trimester and have been shown to have normal karyotypes (3). The
etiology of these abortions are manifold and include the following:
TABLE 15-2 ESTIMATED OUTCOME OF 100 PREGNANCIES VERSUS TIME
FROM CONCEPTION
Time from
Conception
Percent Survival
to Term
Preimplantation
0-6 days
25
Postimplantation
7-13 days
14-20 days
3-5 wk
55
73
79.5
6-9 wk
90
10-13 wk
92
14-17
18-21
22-25
26-29
30-33
34-37
96.26
97.56
98.39
98.69
98.98
99.26
wk
wk
wk
wk
wk
wk
Last Time for Induction of

Selected Malformations
22-23 days; cyclopia; sirenomelia,

microtia
26 days; anencephaly
28 days; meningomyelocele
34 days; transposition of great
vessels
36 days; ccenter lip,
6 wk; diaphragmatic hernia, rectal
atresia,
ventricular septal defect, syndactyly
9 wk; ccenter palate
10 wk; omphalocele
12 wk; hypospadias
38 + wk
99.32
38 + wk; central nervous system

(CNS) cell depletion
Modified from Schardein JL, ed. Chemically induced birth defects. New
York: Marcel Dekker, 1993.
Genetic Abnormalities
Dominant mutations (lethals), polygenic genetic abnormalities, and recessive
disease may rarely account for repetitive abortion but in most instances they will
occur sporadically. A review of gene knockouts and mutations in mice suggests
that embryonic death resulted from disrupting basic cellular functions, vascular
circulation, hematopoiesis or nutritional supply from the mother rather than
affecting embryonic organ systems (5).
Maternal Diabetes
Type I (insulin dependent) diabetes mellitus with poor metabolic control increases
the risk of abortion and still births but there is no increased risk with good
metabolic control.
Maternal Hyper- and Hypothyroidism

Abnormal thyroid function is rare in patients with recurrent abortion.
Corpus Luteum or Placental Progesterone

Deficiency (Luteal Phase Deficiency)
It is controversial whether low hormone levels after implantation result from
impending abortion or are the cause of the abortion.
Maternal Infection
Infections of the genital tract could be responsible for abortion but it is not easy to
document causality. The data
P.226
suggesting that infection with Chlamydia trachomatis, Borrelia burgdorferi,
Mycoplasma hominis, Listeria monocytogenes, and Ureaplasma urealyticum result
in abortion are not conclusive. In contrast, maternal disease resulting in fetal
infection with Treponema pallidum, Plasmodium falciparum, Toxoplasma gondii,
herpes simplex virus, parvovirus B19, or cytomegalovirus has the potential to
cause stillbirth or spontaneous abortion.
TABLE 15-3 ETIOLOGY OF SPONTANEOUS ABORTION IN THE HUMAN
Chromosomal abnormalities from either the maternal or paternal
gonadocytes account for 50%-70% of abortions
Abortions with Normal Chromosomes (Euploidy)
Genetic abnormalities: dominant mutations (lethal), polygenic genetic
abnormalities, recessive disease from either the maternal, paternal, or
both parents gonadocytes.
Severe maternal disease states: diabetes, hypothyroidism, hepatitis,
collagen diseases, untreated hyperthyroidism, severe malnutrition
Corpus luteum or placental progesterone deficiency (luteal phase
deficiency)
Maternal infection which results in fetal infection: Treponema pallidum,
Plasmodium falciparum, Toxoplasma gondii, herpes simplex virus,
parvovirus B19, or cytomegalovirus
Antiphospholipid antibodies: lupus anticoagulant, anticardiolipin
antibodies
Maternal-fetal histocompatibility
Overmature gametes
Mechanical or physical problems: uterine abnormalities, multiple
pregnancies, very rarely trauma
Cervical incompetence
Abnormal placentation: hypoplastic trophoblast, circumvallate
implantation
Embryos and fetuses with severe malformations or growth retardation
Severe, Debilitating Maternal Diseases

Hepatitis, collagen diseases, untreated hyperthyroidism, Wilson's disease or
severe malnutrition can lead to abortion.
Antiphospholipid Antibodies
Lupus anticoagulant and anticardiolipin antibodies predispose women to recurrent
abortion in both first and second trimesters as a result of vascular disruption in
the placenta.
Maternal-Fetal Histocompatibility
It is suggested that embryonic loss increases if the mother and fetus are more
histocompatible at the human leukocyte antigen (HLA) locus, resulting in the
failure to develop maternal blocking antibodies against paternal antigens.
Overmature Gametes
Either the ovum or sperm could age because insemination occurred a few days
prior to ovulation or ovulation occurred prior to insemination. The magnitude of
this risk factor as a cause of spontaneous abortion is not known and some
investigators are skeptical that this phenomenon is clinically significant.
Mechanical or Physical Problems Related to

Uterine Abnormalities, Multiple Pregnancies or
Trauma
A hostile intrauterine environment can result from submucosal or intramural
myomas, adhesions (Asherman's syndrome), multiple embryos or abnormalities of
the uterus (bifid uterus, infantile uterus). Uterine trauma from a direct blow or
penetrating injury may rarely be responsible for an abortion and if this type of
injury occurs it would be more likely to result in a still birth at midgestation or
later.
Cervical Incompetence
Cervical incompetence is more likely to result in second trimester than first
trimester abortions.
Abnormal Placentation
Hypoplastic trophoblast and circumvallate implantation increase the risk of fetal
loss.
Some Environmental Teratogens and

Reproductive Toxins
Severe malformations or growth retardation implantation increase the risk of fetal
loss.
Congenital Malformations
The etiology of congenital malformations can be divided into three categories:
unknown, genetic, and environmental Tables 15-4. The etiology of 65%-75% of
human malformations is unknown. A significant proportion of congenital
malformations of unknown etiology is likely to have an important genetic
component. Malformations with an increased recurrence risk, such as cleft lip and
palate, anencephaly, spina bifida, certain congenital heart diseases, pyloric

stenosis, hypospadias, inguinal hernia, talipes equinovarus, and congenital
dislocation of the hip, fit in the category of multifactorial disease and in the
category of polygenic inherited disease (6). The multifactorial/threshold
hypothesis postulates the modulation of a continuum of genetic characteristics by
intrinsic and extrinsic (environmental) factors (6). Although the modulating
factors are not known, they probably include: placental blood flow, placental
transport, site of implantation, maternal disease states, maternal malnutrition,
infections, drugs, chemicals, and spontaneous errors of development.
Spontaneous errors of development may account for some of the malformations
that occur without apparent abnormalities of the genome or environmental
influence. We postulate that there is some probability for error during embryonic
development based on the fact that embryonic development is such a complicated
process. It is estimated that 75% of all conceptions are lost before term; 50%
within the first 3 weeks of development (1,2). The World Health Organization
estimated that 15% of all clinically recognizable pregnancies end in a spontaneous
abortion, 50%-60% of which are as a result of chromosomal abnormalities
P.227
(4,7,8). Finally, 3%-6% of offspring are malformed, which represents the
background risk for human maldevelopment. This means that, as a conservative
estimate, 1,176 clinically recognized pregnancies will result in approximately 176
miscarriages and 30 to 60 of the infants will have congenital anomalies in the
remaining 1,000 live births. The true incidence of pregnancy loss is much higher
because undocumented pregnancies are not included in this risk estimate.
TABLE 15-4 ETIOLOGY OF HUMAN CONGENITAL MALFORMATIONS
OBSERVED DURING THE FIRST YEAR OF LIFE
Suspected Cause
Unknown
Polygenic
Multifactorial (gene-environment interactions)
Spontaneous errors of development
Synergistic interactions of teratogens
Genetic
Autosomal and sex-linked inherited genetic
disease
Cytogenetic (chromosomal abnormalities)
New mutations
Environmental
Percent of Total
65-75
15-25
10
Maternal conditions: alcoholism; diabetes;
endocrinopathies; phenylketonuria; smoking and

nicotine; starvation; nutritional deficits
Infectious agents: rubella, toxoplasmosis,
syphilis, herpes simplex, cytomegalovirus,

varicella-zoster, Venezuelan equine encephalitis,
parvovirus B19
Mechanical problems (deformations):
Amniotic band constrictions; umbilical cord
1-2
constraint; disparity in uterine size and uterine

contents
Chemicals, drugs, high dose ionizing radiation,
hyperthermia
<1
Modified from Brent RL. Environmental factors: miscellaneous. In: Brent

RL, Harris ML, eds. Prevention of embryonic fetal and perinatal disease.
Bethesda: DHEW (NIH), 1976:211218; Brent RL. Definition of a teratogen and the relationship of teratogenicity to carcinogenicity [Editorial].
Teratology 1986;34:359360.
Based on his review of the literature, Wilson (9) provided a format of theoretical
teratogenic mechanisms: mutation; chromosomal aberrations; mitotic
interference; altered nucleic acid synthesis and function; lack of precursors,
substrates, or coenzymes for biosynthesis; altered energy sources; enzyme
inhibition; osmolar imbalance, alterations in fluid pressures, viscosities, and
osmotic pressures; and altered membrane characteristics. We suggest a revised
list of mechanisms for teratogenesis Tables 15-5.
Even though an agent can produce one or more of these pathological processes,
exposure to such an agent does not guarantee that maldevelopment will occur.
Furthermore, it is likely that a drug, chemical or other agent can have more than
one effect on the pregnant woman and the developing conceptus and therefore
the nature of the drug or its biochemical or pharmacological effects will not in
themselves predict a teratogenic effect in the human. In fact, the discovery of
human teratogens has come primarily from human epidemiological studies.
Animal studies and in vitro studies can be very helpful in determining the
mechanism of teratogenesis and the pharmacokinetics related to teratogenesis
(10). However, even if one understands the pathological effects of an agent, one
cannot predict the teratogenic risk of an exposure without taking into
consideration the developmental stage, the magnitude of the exposure and the
reparability of the embryo.
TABLE 15-5 MECHANISMS OF TERATOGENESIS
1. Cell death or mitotic delay beyond the recuperative capacity of the embryo
or fetus.
2. Inhibition of cell migration, differentiation and cell communication.
3. Interference with histogenesis by processes such as cell depletion,
necrosis, calcification, or scarring.
4. Biologic and pharmacological receptor-mediated developmental effects.
5. Metabolic inhibition or nutritional deficiencies.
6. Physical constraint, vascular disruption, inflammatory lesions, amniotic
band syndrome.
Various maternal viral, bacterial, and parasitic infections are known to cause
maldevelopment in humans including cytomegalovirus, fetal herpes virus
infections (type 1 or 2), parvovirus B19 (erythema infectiosum), rubella virus,
congenital syphilis (Treponema pallidum), T. gondii infection, varicella-zoster
virus, and venezuelan equine encephalitis (11). The incidence of serum antibody
to human immunodeficiency virus (HIV) in pregnant women is increasing from the
1991 estimate of 1.5 per 1,000 women delivering in the United States (12); the
incidence is as high as 31% in pregnant women in some African cities (13).
Several studies support the conclusion that asymptomatic HIV pregnancies are not
associated with an increased risk of congenital malformations, low birth weight or
abortion (14,15,16,17). It is likely that sexually transmitted diseases,
opportunistic maternal infections, and symptomatic HIV pregnancies may increase
the risk of low birth weight and morbidity in noninfected offspring.
The lethal or developmental effects of infectious agents are the result of mitotic
inhibition, direct cytotoxicity or necrosis. Repair processes may result in
metaplasia, scarring or calcification, which causes further damage by interfering
with histogenesis. Infectious agents appear to be exceptions to some of the
principles of teratogenesis because the relevance of dose and time of exposure
cannot be demonstrated as readily for replicating teratogenic agents.
Transplacental transmission of an infectious agent does not necessarily result in
congenital malformations, growth retardation, or lethality.
Vascular disruption is a rare event associated with intrauterine death and a wide
range of structural anomalies, including cerebral infarctions, certain types of
visceral and urinary tract malformations, congenital limb
P.228
amputations of the nonsymmetrical type; and orofacial malformations such as
mandibular hypoplasia, cleft palate, and Moebius syndrome, which vary too widely
to constitute a recognized syndrome. Some anomalies associated with twin
pregnancies can be explained by vascular disruption resulting from placental
anastomoses in shared placenta of monozygotic twins, anastomoses in a small

percentage of dichorionic placentas in the case of dizygotic twins, or death of one
twin resulting in emboli, intravascular coagulation and altered fetal hemodynamics
in the co-twin (18,19). Vascular disruption may also result from physical trauma
causing chorion bleeding, such as chorionic villous sampling, and exposure to
some developmental toxicants, such as cocaine and misoprostol. Although uterine
bleeding during the first trimester may result in fetal anomalies, the
malformations associated with vascular disruption can also occur later in
gestation. This topic is discussed in greater detail below with specific drugs.
Adverse Effects Produced Later in Pregnancy

The fetal period is characterized by histogenesis involving cell growth,
differentiation, and migration. Drugs that produce permanent cell depletion,
vascular disruption, necrosis, specific tissue or organ pathology, physiological
decompensation, or severe growth retardation have the potential to cause
deleterious effects throughout gestation. Sensitivity of the fetus for induction of
mental retardation and microcephaly is greatest at the end of the first and the
beginning of the second trimester. Other permanent neurological effects can be
induced in the second and third trimesters.
The classic example of a drug that presents little risk to the developing embryo
during organogenesis but can affect the near-term fetus if high doses are used is
aspirin. It is possible that other antiinflammatory drugs present a similar risk.
FACTORS THAT AFFECT SUSCEPTIBILITY TO THE

DELETERIOUS EFFECTS OF DRUGS
A basic tenet of environmentally produced embryo- and fetotoxicity is the effects
of teratogenic or abortigenic milieu which have certain characteristics in common
and follow certain basic principles. These principles determine the quantitative and
qualitative aspects of developmental toxicity Tables 15-6.
Stage of Development
The induction of developmental toxicity by environmental agents usually results in
a spectrum of morphological anomalies or intrauterine death, which varies in
incidence depending on stage of exposure and dose. The developmental period at
which an exposure occurs will determine which structures are most susceptible to
the deleterious effects of the drug or chemical and to what extent the embryo can
repair the damage. The period of sensitivity may be narrow or broad, depending
on the environmental agent and the malformation in question. Limb defects,
produced by thalidomide, have a very short period of susceptibility Tables 15-7
although microcephaly produced by radiation has a long period of susceptibility.
Our knowledge of the susceptible stage of the embryo to various environmental

influences is continually expanding and is vital to evaluating the significance of
individual exposures or epidemiological studies.
TABLE 15-6 FACTORS THAT INFLUENCE SUSCEPTIBILITY TO
DEVELOPMENTAL TOXICANTS
Stage of development: The developmental period at which an exposure

occurs will determine which structures are most susceptible to the adverse
effects of chemicals and drugs and to what extent the embryo can repair
the damage.
Magnitude of the exposure: Both the severity and incidence of toxic effects
increase with dose.
Threshold phenomena: The threshold dose is the dose below which the
incidence of death, malformation, growth retardation, or functional deficit
is not statistically greater than that of nonexposed subjects.
Pharmacokinetics and metabolism: The physiologic changes in the
pregnant woman and during fetal development and the bioconversion of
compounds can significantly influence the developmental toxicity of drugs
and chemicals by affecting absorption, body distribution, active
metabolites, and excretion.
Maternal diseases: A maternal disease may increase the risk of fetal
anomalies or abortion with or without exposure to a chemical or drug.
Placental transport: Most drugs and chemicals cross the placenta. The rate
and extent to which a drug or chemical crosses the placenta are influenced
by molecular weight, lipid solubility, polarity or degree of ionization,
plasma protein binding, receptor mediation, placental blood flow, pH
gradient between the maternal and fetal serum and tissues, and placental
metabolism of the chemical or drug.
Genotype: The maternal and fetal genotypes may result in differences in
cell sensitivity, placental transport, absorption, metabolism, receptor
binding and distribution of an agent, and account for some variations in
toxic effects among individual subjects and species.
During the first period of embryonic development, from fertilization through the
early postimplantation stage, the embryo is most sensitive to the embryolethal
effects of drugs and chemicals. Surviving embryos have malformation rates
similar to the controls not because malformations cannot be produced at this
stage but because significant cell loss or chromosome abnormalities at theses
stages have a high likelihood of killing the embryo. Because of the
omnipotentiality of early embryonic cells, surviving embryos have a much greater
ability to have normal developmental potential. Wilson and Brent (20)
demonstrated that the all-or-none phenomenon, or marked resistance to
teratogens, disappears over a period of a few hours in the rat during early
organogenesis utilizing ionizing X-irradiation as the experimental teratogen. The
term all-or-none phenomenon has been misinterpreted by some investigators to
indicate that malformations cannot be produced at this stage. On the contrary, it
is likely that certain drugs, chemicals, or other insults during this stage of
development can result in malformed offspring, but the nature of embryonic
development
P.229
at this stage will still reflect the basic characteristic of the all-or-none
phenomenon which is a propensity for embryo lethality rather than surviving
malformed embryos.
TABLE 15-7 DEVELOPMENTAL STAGE SENSITIVITY TO THALIDOMIDEINDUCED LIMB REDUCTION DEFECTS IN THE HUMAN
Days from Conception for
Induction of Defects
Limb Reduction Defects
21-26
22-23
23-34
Thumb aplasia
Microtia
Hip dislocation
24-29
24-33
25-31
Amelia, upper limbs

Phocomelia, upper limbs
Preaxial aplasia, upper limbs
27-31
28-33
Amelia, lower limbs

Preaxial aplasia, lower limbs;
phocomelia, lower limbs; femoral
hypoplasia; girdle hypoplasia

30-36
Triphalangeal thumb
Modified from Brent RL, Holmes LB. Clinical and basic science lessons from
the thalidomide tragedy: what have we learned about the causes of limb
defects? Teratology 1988;38:241-251, with permission.
The period of organogenesis (from day 18 through about day 40 of postconception
in the human), is the period of greatest sensitivity to teratogenic insults and the
period when most gross anatomic malformations can be induced. Most
environmentally produced major malformations occur before the 36th day of
gestation in the human. The exceptions are malformations of the genito-urinary
system, the palate, the brain or deformations as a result of problems of
constraint, disruption or destruction. Severe growth retardation in the whole
embryo or fetus may also result in permanent deleterious effects in many organs
or tissues.
The fetal period is characterized by histogenesis involving cell growth,
differentiation, and migration. Teratogenic agents that produce permanent cell
depletion, vascular disruption, necrosis, specific tissue or organ pathology,

physiological decompensation, and/or severe growth retardation have the
potential to cause deleterious effects throughout gestation. Additionally,
sensitivity of the fetus for induction of mental retardation and microcephaly is
greatest at the end of the first and the beginning of the second trimester. Other
permanent neurological effects can be induced in the second and third trimesters.
Effects such as cell depletion or functional abnormalities, not readily apparent at
birth, may give rise to changes in behavior or fertility which are apparent only
later in life. The approximate last gestational day on which certain malformations
may be induced in the human is presented in Table 15-2.
Magnitude of the Exposure

The dose-response relationship is extremely important when comparing effects
among different species because usage of mg/kg doses are, at most, rough
approximations. Dose equivalence among species can only be accomplished by
performing pharmacokinetic studies, metabolic studies, and dose-response
investigations in the human and the species being studied. Furthermore, the
response should be interpreted in a biologically sound manner. One example is
that a substance given in large enough amounts to cause maternal toxicity is
likely to also have deleterious effects on the embryo such as death, growth
retardation, or retarded development. Another example is that because the
steroid receptors that are necessary for naturally occurring and synthetic
progestin action are absent from nonreproductive tissues early in development,
the evidence is against the involvement of progesterone or its synthetic analogues
in nongenital teratogenesis (21,22).
An especially anxiety-provoking concept is that the interaction of two or more
drugs or chemicals may potentiate their developmental effects. Although this is an
extremely difficult hypothesis to test in the human, it is an especially important
consideration because multichemical or multitherapeutic exposures are common.
Fraser (23) warns that the actual existence of a threshold phenomenon when
nonteratogenic doses of two teratogens are combined could easily be
misinterpreted as potentiation or synergism. Potentiation or synergism should be
invoked only when exposure to two or more drugs is just below their thresholds
for toxicity.
Several considerations affect the interpretation of dose-response relationships:
Active metabolites: Metabolites may be the proximate teratogen rather than the
administered chemical, i.e., the metabolites phosphoramide mustard and acrolein
may produce maldevelopment resulting from exposure to cyclophosphamide (24).
Duration of exposure: A chronic exposure to a prescribed drug can contribute to
an increased teratogenic risk, e.g., anticonvulsant therapy; in contrast an acute
exposure to the same drug may present little or no teratogenic risk.
Fat-solubility: Fat-soluble substances such as polychlorinated biphenyls (25) can

produce fetal maldevelopment for an extended period after the last ingestion or
exposure in a woman because they have an unusually long half life.
Threshold Phenomena
The threshold dose is the dosage below which the incidence of death,
malformation, growth retardation, or functional deficit is not statistically greater
than that of controls. The threshold level of exposure is usually from less than one
to three orders of magnitude below the teratogenic or embryopathic dose for
drugs and chemicals that kill or malform half the embryos. A teratogenic agent
therefore has a no-effect dose as compared to mutagens or carcinogens, which
have a stochastic dose response curve. Threshold phenomena are compared to
stochastic phenomena in Table 15-8. The severity and incidence of malformations
produced by every exogenous teratogenic agent that has been appropriately
tested have exhibited threshold phenomena during organogenesis (9).
TABLE 15-8 STOCHASTIC AND THRESHOLD DOSE-RESPONSE RELATIONSHIPS OF DISEASES PRODUCED

BY ENVIRONMENTAL AGENTS
Relationship
Stochastic
phenomena
Pathology
Site
Damage to a
deoxyribonucleic
single cell may acid DNA
result in
disease
Diseases
Risk
Cancer, mutation Some risk exists

at all dosages;
at low
exposures the
Definition
Incidence of
disease
increases but
severity and
risk is below the nature of the

spontaneous risk disease
remain the
Threshold
phenomena
Multicelluar
injury
High variation in
etiology, affecting
many cell and organ
processes
Malformation,
growth
retardation,
death, chemical
toxicity, etc.
No increased
risk below the
threshold dose
same
Both severity
and incidence
of the
disease
increase with
dose
Modified from Brent RL. Definition of a teratogen and the relationship of teratogenicity to carcinogenicity
[Editorial]. Teratology 1986;34:359360, with permission.
P.230
Pharmacokinetics and Metabolism

The physiological alterations in pregnancy and the bioconversion of compounds
can significantly influence the teratogenic effects of drugs and chemicals by
affecting absorption, body distribution, active form(s), and excretion of the
compound. Physiological alterations in the mother during pregnancy which affect
the pharmacokinetics of drugs include the following: decreased gastrointestinal
motility and increased intestinal transit time may delay absorption of drugs
absorbed in the small intestine as a result of increased stomach retention but
enhance absorption of slowly absorbed drugs; decreased plasma albumin
concentration, which alters the kinetics of compound normally bound to albumin;
increased plasma and extracellular fluid volumes that affect concentrationdependent transfer of compounds; renal elimination, which is generally increased
but is influenced by body position during late pregnancy; inhibition of metabolic
inactivation in the liver late in pregnancy; and variations in uterine blood flow,
although little is known about how this affects transfer across the placenta
(26,27,28).
The fetus also undergoes physiological alterations, which affect the
pharmacokinetics of drugs (28): the amount and distribution of fat varies with
development and affects the distribution of lipid-soluble drugs; the fetal circulation
contains a higher concentration of unbound drug largely because the plasma fetal
proteins are lower in concentration than in the adult and may lower drug affinity;
the functional development of pharmacological receptors is likely to proceed at
different rates in the various tissues; and i.v. drugs excreted by the fetal kidneys
may be recycled by swallowing of amniotic fluid.
The role the placenta plays in drug pharmacokinetics, reviewed by Juchau and
Rettie (29) and Miller (25), involves transport [discussed in detail in the section
Placental Transport; the presence of receptors sites for a number of endogenous
and xenobiotic compounds (-adrenergic, glucocorticoid, epidermal growth factor,
IgG-Fc, insulin, low-density lipoproteins, opiates, somatomedin, testosterone,
transcobalamin II, transferrin, folate, retinoid)] (25); and the bioconversion of
xenobiotics. Bioconversion of xenobiotics has been shown to be important in the
teratogenic activity of several xenobiotics. There is strong evidence that reactive
metabolites of cyclophosphamide, 2-acetylaminofluorene, and nitroheterocycles
(niridazole) are the proximal teratogens (30). There is also experimental evidence
that suggests that other chemicals undergo conversion to intermediates that have
deleterious effects on embryonic development including phenytoin, procarbazine,
rifampicin, diethylstilbestrol, some benzhydrylpiperazine antihistamines,
adriamycin, testosterone, benzo(a)pyrene, methoxyethanol, caffeine, and
paraquat (29,30).
The major site of bioconversion of chemicals in vivo is likely to be the maternal
liver. Placental P450-dependent monooxygenation of xenobiotics will occur at low
rates unless induced by such compounds as those found in tobacco smoke (29).
However, the fetus also develops functional P450 oxidative isozymes capable of
converting proteratogens to active metabolites.
Maternal Disease
Maternal disease states such as diabetes mellitus, epilepsy, phenylketonuria, and
endocrinopathies are associated with adverse effects on the fetus. In some cases,
it may be difficult to determine whether a maternal disease or the treatment for
the disease plays a role in the etiology of malformations associated with the
treatment for that disease during pregnancy. For example, the genetic and
environmental milieu, which cause epilepsy may also contribute to the
maldevelopment associated with exposure to diphenylhydantoin (31).
The role of maternal malnutrition is an important area for investigation because it
may be a contributing factor to many teratogenic milieu. A series of investigations
provided evidence suggesting that folic acid supplementation could reduce the
incidence of recurrence of neural tube defects in the human (32,33,34,35). It was
later shown convincingly that periconceptional supplementation with folic acid, 4
mg/day, reduces the risk of recurrence of neural tube defects in subsequent
siblings of children with neural tube defects (36). Furthermore, low-dose folic acid
supplementation, 0.8 mg/day, was reported to decrease the
P.231
incidence of neural tube defects in a population not at increased risk for these
defects (37). Although folate supplementation reduces the incidence of neural
tube defects, folate supplementation will not prevent all neural tube defects and it
is not known whether folic acid supplementation corrects an undefined metabolic
defect or a nutritional deficiency.
Placental Transport
The exchange between the mammalian embryo and the maternal organism is
controlled by the placenta which includes the chorioplacenta, the yolk sac placenta
and the paraplacental chorion. The placenta varies in structure and function
among species and for each stage of gestation. As an example, the rodent yolk
sac placenta continues to function as an organ of transport for a much greater
part of gestation than in the human. Thus differences in placental function and
structure may affect our ability to apply teratogenic data developed in one species
directly to other species, including the human (38). As pharmacokinetic
techniques and the actual measurement of metabolic products in the embryo
become more sophisticated, the appropriateness of utilizing animal data to project

human effects may improve.
Historically a placental barrier was thought to exist, which prevented harmful
substances from reaching the embryo. It is now clear that there is no placental
barrier per se. The fact is that most drugs and chemicals cross the placenta. It
will be a rare substance that will cross the placental barrier in one species and be
unable to reach the fetus in another (39). No such chemical exists except for
selected proteins whose actions are species-specific.
Even before there were chemical techniques to demonstrate the presence of drugs
or chemicals in the embryo there was clear evidence that they had reached the
fetus because of clinical manifestations of the drugs: anticoagulants such as
warfarin can affect the clotting of fetal blood; many drugs can affect the fetal
cardiac rate; changes in the fetal electroencephalogram (EEG) can be
demonstrated as a result of the many drugs that affect the central nervous
system (CNS); and newborns may exhibit withdrawal symptoms from drugs taken
by their mothers, either medications or substances of abuse such as alcohol or
opiates.
These observations demonstrate clinically significant placental transport of drugs
only in the latter portion of gestation and may not be a means of evaluating
embryonic exposure during early organogenesis.
Those factors which determine the ability of a drug or chemical to cross the
placenta and reach the embryo include: molecular weight, lipid affinity or
solubility, polarity or degree of ionization, protein binding, and receptor mediation.
Compounds with low molecular weight, lipid affinity, nonpolarity, and without
protein-binding properties will cross the placenta with ease and rapidity. As an
example, ethyl alcohol is a chemical which reaches the embryo rapidly and in
concentrations equal to or greater than the level in the mother.
High molecular weight compounds like heparin, 20,000 daltons, do not readily
cross the placenta and therefore heparin is used to replace warfarin-like
compounds during pregnancy for the treatment of hypercoagulation conditions.
Rose Bengal does not cross the placenta. In general, compounds with molecular
weights of 1,000 or greater do not readily cross the placenta, although 600 dalton
compounds usually do; most drugs are 250-400 daltons and cross the placenta
(40).
In addition to the particular properties of the drug or chemical, three other
conditions affect the quantitative aspect of placental transport: (a) placental blood
flow, (b) the pH gradient between the maternal and fetal serum and tissues, and
(c) placental metabolism of the chemical or drug. The biotransformation
properties of the placenta and/or maternal organism are important because a
number of chemicals or drugs are not teratogenic in their original form.
The most important concept regarding placental transport of teratogens must be

reemphasized. An agent is teratogenic because it affects the embryo directly or
indirectly by its ability to produce a toxic effect in the embryo or extraembryonic
membranes at exposures which are attained in the human being, not because it
crosses the placenta per se.
Genotype
The genetic constitution of an organism is an important factor in the susceptibility
of a species to a drug or chemical. More than 30 disorders of increased sensitivity
to drug toxicity or effects have been reported in the human as a result of an
inherited trait (41). The effect of a drug or chemical depends on both the maternal
and fetal genotypes and may result in differences in cell sensitivity, placental
transport, absorption, metabolism (activation, inactivation, active metabolites),
receptor binding, and distribution of an agent. This accounts for some variations in
teratogenic effects among species and in individual subjects.
Estimating the Developmental Risks of Drugs

During Human Pregnancy
Evaluation of Data Available for the Human
Although chemicals and drugs can be evaluated for fetotoxic potential by utilizing
in vivo animal studies and in vitro systems, it should be recognized that these
testing procedures are only one component in the process of evaluating the
potential teratogenic risk of drugs and chemicals in the human. The evaluation of
the teratogenicity of drugs and chemicals should include, when possible, (a) data
obtained from human epidemiological studies, (b) secular trend data in humans,
(c) animal developmental toxicity studies, (d) the dose-response relationship for
developmental toxicity and the relationship to the human pharmacokinetic
equivalent dose in the animal studies, and (e) considerations of biological
plausibility Tables 15-9 (42,43). This method is of greatest value when utilized for
P.232
the evaluation of chemicals and drugs that have been in use for some time or for
evaluating new drugs that have a similar mechanism of action, structure,
pharmacology and purpose similar to other, extensively studied agents. The ability
to establish a causal relationship between an environmental agent and abortigenic
effect is more difficult, for the following reasons:
Abortion is a very frequent reproductive event and therefore the incidence

can vary considerably between different populations of women. Differences
in the abortion incidence between two populations in a single study may be

as a result of chance alone.
There are multiple causes of abortion and most epidemiological studies

dealing with abortion make no attempt to determine the etiology of the
abortions. Since most abortions are as a result of preconceptual or
periconceptual events, it is extremely difficult to match patients in case
control studies and it would be necessary to have large increases in a
particular etiological category of environmentally induced abortion to
demonstrate a statistically significant increase in the incidence of
spontaneous abortion in an exposed population of pregnant women.
Confounding factors appear to be more significant in abortion studies than

in birth defect studies (cocaine, smoking, alcohol, syphilis, narcotics,
caffeine). This further decreases the possibility that the agent being studied
has a direct abortigenic effect.
The incidence of therapeutic abortions is difficult to estimate or control for in

most epidemiological studies (44,45).
TABLE 15-9 EVIDENCE FOR POTENTIAL DEVELOPMENTAL TOXICITY IN

THE HUMAN
Epidemiological studies: Epidemiological studies consistently demonstrate

an increased incidence of pregnancy loss or of a particular spectrum of
fetal effects in exposed human populations.
Secular trend data: Secular trends demonstrate a relationship between the
incidence of pregnancy loss or a particular fetal effect and the changing
exposures in human populations. The percent of the population exposed
must be large for this analysis.
Animal developmental toxicity studies: An animal model mimics the human
developmental effect at clinically comparable exposures. Since mimicry
may occur in only one animal species, if it occurs at all, it would not
necessarily be observed during an initial developmental toxicology study.
Developmental toxicity studies are therefore indicative of a potential
hazard in general rather than the potential for a specific adverse effect on
the fetus.
Dose-response relationship: Developmental toxicity in the human increases
with dose and the developmental toxicity in animals occurs at a dose that
is pharmacokinetically equivalent to the human dose.
Biological plausibility: The mechanisms of developmental toxicity are
understood or the results are biologically plausible.

Modified from Brent RL. Method of evaluating alleged human teratogens
[Editorial]. Teratology 1978;17:83; Brent RL. Definition of a teratogen and
the relationship of teratogenicity to carcinogenicity [Editorial]. Teratology
1986;34:359360, with permission.
One of the advantages of reproductive effects is that there is frequently, but not
always, concordance of effects involving more than one parameter (growth,
malformations, abortion, stillbirth, prematurity, etc.). Isolated abortion studies
that do not study the totality of reproductive effects are at a serious
disadvantage, because spurious or nonetiological results may be misinterpreted as
being causally related to a drug or environmental toxicant.
Some investigators and regulatory agencies divide drugs and chemicals into
developmentally toxic and nontoxic compounds. In reality, potential
developmental toxicity can be evaluated only if one considers, as a minimum, the
agent, the dose, the species and the stage of gestation. Working definitions for
developmental toxicity in the human are suggested in Table 15-10.
Potential human teratogens and abortifacients comprise a large group of drugs
because they include all drugs and chemicals that can produce embryotoxic and
fetotoxic effects at some exposure. Since these exposures are not utilized or
attained in the human, they represent no or minimal risks to the human embryo.
Misconceptions in Evaluating Developmental

Toxicity in the Human
Misconceptions have lead to confusion regarding the potential effects of even
proven teratogens. Examples of erroneous concepts include: if an agent can
produce one
P.233
type of malformation, it can produce any malformation; an agent presents a risk
at any dose, once it can be proven to be teratogenic; and an agent that is
teratogenic is likely to be abortigenic.
TABLE 15-10 DEFINITIONS OF POTENTIAL FOR DEVELOPMENTAL
TOXICITY IN THE HUMAN
Developmental toxicant: An agent or milieu that has been demonstrated to

produce permanent alterations or death in the embryo or fetus following
intrauterine exposures that usually occur or are attainable in the human.
Potential for developmental toxicity: An agent or milieu that has not been
demonstrated to produce permanent alterations or death in the embryo or
fetus following intrauterine exposures that usually occur or are attainable
in the human, but can affect the embryo or fetus if the exposure is raised
substantially above the usual exposure. Most chemicals and drugs
have thepotential for interrupting a pregnancy or inducing developmental
defects if the exposure is increased sufficiently.
Little or no potential for developmental toxicity: An agent or milieu that
has been demonstrated to produce no embryo- or fetotoxicity at any
attainable dose in the human. In contrast, an environmental agent may be
so toxic that it has no developmental toxicity in the human because it kills
the mother before or at the same dose that it begins to have adverse
effects on the embryo.
Modified from Brent RL. Method of evaluating alleged human teratogens
[Editorial]. Teratology 1978;17:83, with permission.
These concepts are incorrect. The data clearly indicate that proven teratogens do
not have the ability to produce every birth defect. Many teratogens can be
identified on the basis of the malformations that are produced. Thus, the concept
of the syndrome is probably more appropriate in clinical teratology than any other
area of clinical medicine. Some symptoms or signs appear in many teratogenic
syndromes, such as growth retardation or mental retardation, and therefore are
not very discriminating. On the other hand, rare or specific effects, such as
deafness, retinitis, or a pattern of cerebral calcifications, may point to a specific
teratogen. It is also true that there is substantial overlap in malformation
syndromes which may not always be separable. Environmentally produced birth
defects may be confused with genetically determined malformations. Using
thalidomide as an example, a patient with bilateral radial aplasia and a ventricular
septal defect may have the Holt Oram syndrome or the thalidomide syndrome. It
may or may not be possible to make a diagnosis with absolute certainty, even if
one has a history of thalidomide ingestion during pregnancy. It is possible,
however, to refute the suggestion that thalidomide was responsible for congenital
malformations in an individual by the nature of the limb malformation.
The specificity of some teratogens can sometimes point to the mechanism or site
of action. For instance, the predominant central nervous system effects of methyl
mercury are understood when one realizes the propensity for organic mercury to
be stored in lipid.
Epidemiologists sometimes use poor judgment when grouping malformations. As
an example, limb reduction defects are frequently studied regarding their
association with environmental teratogens but, in some studies, limb defects that
are clearly related to problems of organogenesis are lumped with congenital
amputations even though it is very unlikely that any agent will be responsible for
both types of malformations. It is clear that epidemiological studies could be
markedly improved if there was more input from clinical teratologists in planning
and performing the studies.
Case control studies concerning spontaneous abortion may contain serious errors
unless the populations being studied are similar regarding the stage of pregnancy
when abortion occurred. This study design diminishes the possibility that the
abortion rate will differ on the basis of the selection process and not the drug or
environmental agent being studied. Unfortunately most epidemiological studies
dealing with drug or environmentally induced abortion do not attempt to
determine the etiology of the abortion.
Potential Embryo- and Fetotoxicity of Selected

Prescribed and Self-administered Drugs
We evaluated the literature concerning selected drugs that cause or are suggested
to cause deleterious effects during pregnancy in the human. The data included
human epidemiological studies, secular trend data in humans where appropriate,
and animal developmental toxicity studies. In our analysis we considered the doseresponse relationship of teratogenicity, the relationship to the human
pharmacokinetic equivalent dose in the animal studies and biological plausibility
Tables 15-9 (42,43). Table 15-11 focuses on these drugs, listing their potential
adverse effects in the human. Although these drugs account for a small
percentage of all malformations and abortions, they are important because these
exposures may be preventable.
Alcohol
Adverse effects in offspring from excessive alcohol consumption during pregnancy
were recognized more than 200 years ago (46). It was Jones and associates (47)
however, who defined the fetal alcohol syndrome (FAS) in children with
intrauterine growth retardation, microcephaly, mental retardation, maxillary
hypoplasia, flat philtrum, thin upper lip, and reduction in the width of palpebral
fissures. Cardiac abnormalities were also seen. Many of the children of alcoholic
mothers had FAS and all of the affected children evidenced developmental delay
(47,48).
A period of greatest susceptibility is not clearly established but the risk for
adverse effects increases with increased consumption and binge drinking early in
pregnancy may be associated with an increased risk of alcohol-related effects
(49). The risk of decreased brain growth and differentiation that results from high
alcohol consumption is greater during the second and third trimester. Chronic
consumption of 6 oz of alcohol per day constitutes a high risk although the FAS is
not likely when the mother consumes fewer than two drinks (equivalent to 1 oz of
alcohol) per day (50). Reduction of alcohol consumption or cessation of drinking
early in pregnancy will reduce the incidence and severity of alcohol-related effects
(49,51,52,53) but may not entirely eliminate the risk of some degree of physical
or behavioral impairment. The human syndrome is likely to involve the direct
effects of alcohol and the indirect effects of genetic susceptibility and poor
nutrition. Alcoholism can have maternally deleterious effects on intermediary
metabolism and nutrition, especially if alcoholic cirrhosis is present, which can
contribute to an adverse milieu for the developing embryo.
Although alcoholic mothers frequently smoke and consume other drugs, there is
little doubt from the human and animal data that alcohol ingestion alone can have
a disastrous effect on the developing embryo or fetus. The reported incidence of
FAS varies widely in different studies but appears to be approximately 6% in
offspring of women who drink heavily during pregnancy (51). Fetal alcohol
syndrome may be the most commonly recognized cause of environmentally
induced mental deficiency; there are at least several hundred children born each
year with full FAS and probably many more with subtler fetal alcohol effects
(50,54).
TABLE 15-11 EFFECTS AND ESTIMATED RISKS OF SELECTED PRESCRIBED

AND SELF-ADMINISTERED DRUGS HUMAN PREGNANCY
Reported Effects or
Selected Drugs
Alcohol
Associations and
Estimated Risks
Commentsa
Fetal alcohol syndrome:
Quality of available
intrauterine growth
retardation, maxillary
hypoplasia, reduction in
information: good to
excellent. Direct
cytotoxic effects of
width of palpebral
fissures, characteristic
ethanol and indirect

effects of alcoholism.
but not diagnostic facial

features, microcephaly,
While a threshold
teratogenic dose is
mental retardation. An
increase in spontaneous
likely it will vary in

individuals because of a
abortion has been

reported but since
multiplicity of factors.
mothers who abuse

alcohol during pregnancy
have multiple other risk
factors, it is difficult to
determine whether this is
a direct effect on the
embryo. Consumption of
6 oz of alcohol or more
per day constitutes a
high risk but it is likely
that detrimental effects
can occur at lower
exposures.
Aminopterin,
methotrexate
Microcephaly,
hydrocephaly, cleft
palate,
information: good.
Anticancer,
meningomyelocele,
intrauterine growth
retardation, abnormal
cranial ossification,
antimetabolic agents;
folic acid antagonists
that inhibit
dihydrofolate reductase,
reduction in derivatives
of first branchial arch,
mental retardation,
resulting in cell death.
postnatal growth
retardation. Aminopterin
can induce abortion
within its therapeutic
range; it is used for this
purpose to eliminate
ectopic embryos. Risk
from therapeutic doses is
unknown but appears to
be moderate to high.
Androgens
Masculinization of female Quality of available

embryo: clitoromegaly
information: good.
with or without fusion of Effects are dose and
labia minora. Nongenital stage dependent;
malformations are not a
reported risk. Androgen
exposures which result in

masculinization have
little potential for
inducing abortion. Based
on animal studies,
behavioral
masculinization of the
female human will be
rare.
Angiotensin-converting The therapeutic use of
enzyme (ACE) inhibitors ACE inhibitors has
neither a teratogenic
effect nor an abortigenic
effect in the first
trimester. Since this
group of drugs does not
stimulates growth and

differentiation of sex
steroid receptorcontaining tissue.
information: good.
Antihypertensive
agents; adverse fetal
effects are related to
severe fetal
hypotension over a long
interfere with
period fo time during
organogenesis, they can the second or third
be used in a woman of
trimester.
reproductive age; if the
woman becomes
pregnant, therapy can be
changed during the first
trimester without an
increase in the risk of
teratogenesis. Later in
gestation these drugs
can result in fetal and
neonatal death,
oligohydramnios,
pulmonary hypoplasia,
neonatal anuria,
intrauterine growth
retardation, and skull
hypoplasia. Risk is
dependent on dose and
length of exposure.
Antibiotics
Streptomycin:
Streptomycin and a
information: fair to
group of ototoxic drugs

can affect the eighth
good. Long duration

maternal therapy during
nerve and interfere with

hearing; it is a relatively
low-risk phenomenon.
There are not enough
data to estimate the
abortigenic potential of
streptomycin. Because
the deleterious effect of
streptomycin is limited to
the eighth nerve, it is
unlikely to affect the
incidence of abortion.
pregnancy is associated
with hearing deficiency
in offspring.
information: good.
Effects seen only if
exposure is late in the
first or during second or
third trimester, since
tetracyclines have to
interact with calcified
tissue.
Tetracycline: Bone
staining and tooth
staining can occur with
therapeutic doses.
These antibiotics are

used in late pregnancy
for the treatment of
sexually transmitted
Persistent high doses can diseases.

cause hypoplastic tooth
enamel. No other
congenital malformations
are at increased risk. The
usual therapeutic doses
present no increased risk
of abortion to the
embryo or fetus.
Penicillin G benzathine
used for the treatment of
syphilis produces no
adverse fetal effects in
the usual therapeutic
regimens:
Ceftriaxone and
doxycycline used for the
treatment of gonorrhea
produces no adverse
fetal effects in the usual
therapeutic regimens.
Erythromycin base or
stearate used for the
treatment of Chlamydia
involves a possible
Antihypertensive
(excluding ACE
inhibitors)
increased risk of
cholestatic hepatitis in
regimens.
Clonidine: a direct alpha
adrenergic agonist that
appears to be relatively
safe during pregnancy
but there are few
available data.
Hydralazine: a
vasodilator often used in
combination with
methyldopa and is
considered to be safe.
Methyldopa: a centrally
acting adrenergic
antagonist and currently
the safest
antihypertensive drug
available for use during
pregnancy with no
reported adverse effects
on the fetus or on mental
and physical
development.
Nifedipine: a calcium
channel blocker whose
potential for adverse
effects with its long term
use in the treatment of
hypertension is unknown.
Propranolo: a -blocker
whose prolonged use
may increase the risk of
intrauterine growth
retardation.
Antituberculosis therapy Drugs prescribed for the
treatment of tuberculosis
include aminoglycosides,
ethambutol, isoniazid,
rifampin, and
ethionamide. The
ototoxic effects of
streptomycin (discussed
above) are the only
proven adverse effects of
these drugs on the fetus.
Therapeutic exposures to
other tuberculostatic
drugs appear to
represent a very small
risk of teratogenesis and
Aspirin
even less risk of abortion.

No increased risk for
Used for treatment of
malformations or
preeclamspia, idiopathic
abortion low dose
placental insufficiency,
regimen (6050;150 mg
per day). Aspirin should
be discontinued 1 week
before anticipated
delivery to reduce the
risk for maternal or
neonatal bleeding.
systemic lupus
erythematosus,
increased platelet
aggregation
Benzodiazepines
Benzodiazepines appear
to have minimal or no
The benzodiazepines
are widely used as
increased risk of
malformations at
tranquilizers during
pregnancy.
therapeutic ranges;
higher exposures may
increase the risk. The
risk for abortion is
unknown.
Chlordiazepoxide
(Librium), appears to
have a minimal risk for
congenital anomalies and
no increased risk for
abortion at therapeutic
doses. Higher exposures
are likely to increase the
risk of adverse effects on
the fetus but the
magnitude of the
increase is not known.
Diazepam (Valium): third
trimester exposure can
reversibly affect the fetus
and neonate there is
minimal increased risk of
congenital malformations
and no demonstrated
increased risk of
abortions from
therapeutic exposures.
Meprobamate: weakly
associated with a variety
of congenital
malformations but the
data are not sufficient to
confirm or rule out a
small increase risk of
malformations due to
exposures early in
pregnancy.
Caffeine
Caffeine is teratogenic in Quality of available

rodent species with
doses of 150 mg/kg.
There is no convincing
good. Behavioral effects
have been reported and
data that moderate or

appear to be transient
usual exposures (300 mg or temporary; more
per day or less) present
a measurable risk in the
information is needed
concerning the
human for any

population with higher
malformation or group of exposures.
malformations. On the
other hand, excessive
caffeine consumption
(exceeding 300 mg per
day) during pregnancy is
associated with growth
Carbamazepine
retardation and
embryonic loss.
Minor craniofacial defects
(upslanting palpebral
fissures, epicanthal folds,
short nose with long
philtrum), fingernail
hypoplasia, and
developmental delay.
Teratogenic risk is not
known but likely to be
significant for minor
defects. There are too
few data to determine
whether carbamazapine
presents an increased
risk for abortion. Since
embryos with multiple
malformations are more
likely to abort, it would
appear that
carbamazepine presents
little risk because an
increase in these types of
malformations has not
good. Anticonvulsant;
little is known
concerning mechanism.
Epilepsy may itself
contribute to an
increased risk for fetal
anomalies.
been reported.
Cocaine
Preterm delivery; fetal

loss; placental abruption; information: fair to
intrauterine growth
good. Cocaine causes a
retardation;
microcephaly;
complex pattern of
cardiovascular effects
neurobehavioral
abnormalities; vascular
dueto its local

anesthetic and
disruptive phenomena
resulting in
sympathomimetic
activities in the mother.
limbamputation, cerebral Fetopathology is likely

infarctions and certain
to be due to decreased
types of visceral and
urinary tract
uterine blood flow and

fetal vascular effects.
malformations. There are Because ofthe

few data to indicate that mechanism of cocaine
cocaine increases the
risk of first trimester
abortion. The low but
increased risk of vascular
disruptive phenomena
due to vascular
compromise of the
pregnant uteruswould
more likely result in
midgestation abortion or
stillbirth. It is possible
that higher doses could
result in early abortion.
Risk for deleterious
effects on fetal outcome
is significant; risk for
major disruptive effects
is low, but can occur in
the latter portion of the
first trimester as well as
the second and third
trimesters.
teratogenicity, a welldefined cocaine

syndrome is not likely.
Poor nutrition
accompanies drug
abuse and multiple drug
abuse is common.
Coumarin derivatives
Nasal hypoplasia;
stippling of secondary
epiphysis; intrauterine
growth retardation;
information: good.
Anticoagulant; bleeding
is an unlikely
anomalies of eyes,
hands, neck; variable
explanation for effects

produced in the first
central nervous system

anatomical defects
trimester. CNS defects

may occur anytime
(absent corpus callosum, during second and third

hydrocephalus,
trimester and may be
asymmetrical brain
hypoplasia). Risk from
related to bleeding.
exposure 10% to 25%

during 8th to 14th week
of gestation. There is
also an increased risk of
Cyclophosphamide
pregnancy loss. There is

a risk to the mother and
fetus from bleeding at
the time of labor and
delivery.
Growth retardation,
ectrodactyly, syndactyly,
cardiovascular
anomalies, and other
minor anomalies.
Teratogenic risk appears
to be increased but the
information: fair.
Anticancer, alkylating
agent; requires
cytochrome P450 monooxydase activation;
interacts with DNA,
magnitude of the risk is resulting in cell death.

uncertain. Almost all
chemotherapeutic agents
have the potential for
inducing abortion. This
risk is dose-related; at
the lowest therapeutic
doses the risk is small.
Diethylstilbestrol (DES)
Clear cell
adenocarcinoma of the
vagina occurs in about
1:1,000 to 10,000
good. Synthetic
estrogen; stimulates
females who were

estrogen receptorexposed in utero. Vaginal containing tissue, may
adenosis occurs in about cause misplaced genital
75% of females exposed tissue which has a
in utero before the 9th
week of pregnancy.
greater propensity to
develop cancer.
Anomalies of the uterus

and cervix may play a
role in decreased fertility
and an increased
incidence of prematurity
although the majority of
women exposed to DES
in utero can conceive and
deliver normal babies. In
utero exposure to DES
increased the incidence
of genitourinary lesions
and infertility in males.
DES can interfere with
zygote survival, but it
does not interfere with
embryonic survival when
given in its usual dosage
after implantation.
Offspring who were
exposed to DES in utero
have an increased risk
for delivering
permaturely, but do not
appear to be at increased
risk for first trimester
abortion.
Digoxin
No adverse fetal effects

reported with usual

fetal dysrhythmia.
Diphenylhydantoin
Hydantoin syndrome:
microcephaly, mental
retardation, cleft lip/
palate, hypoplastic nails
good. Anticonvulsant;
direct effect on cell
and distal phalanges;

characteristic, but not
membranes, folate, and

vitamin K metabolism.
diagnostic facial features. Metabolic intermediate

Associations documented (epoxide) has been
only with chronic
suggested as the
exposure. Wide variation teratogenic agent.
in reported risk of
malformations but
appears to be no greater
than 10%. The few
epidemiological data
indicate a small risk of
Glucocorticoids
abortion for therapeutic

exposures for the
treatment of epilepsy.
For short term
treatment, i.e.,
prophylactic therapy for
a head injury, there is no
appreciable risk.
Dexamethasone,
Betamethasone,
Dexamethasone,
Hydrocortisone,
Glucocorticoids are used

late in pregnancy to
reduce respiratory
distress in premature
Methylprednisone:
Glucocorticoids have not
been shown to be
teratogenic but chronic
infants and to treat

congenital adrenal
hyperplasia. They are
also used in the
glucocorticoid therapy
may result in prematurity
and intrauterine growth
retardation.
treatment of rheumatic
diseases, other acute
and chronic
inflammatory diseases,
and organ
transplantation.
Indomethacin
Can prolong labor and

Used for the prevention
may predispose neonate or reduction of
to necrotizing
enterocolitis when used
intraventricular
hemorrhage in
as a tocolytic.
premature infant and

for treatment of
polyhydramnious.
Lithium carbonate
Although animal studies

have demonstrated a
clear teratogenic risk,
good. Antidepressant;
the effect in humans is

uncertain. Early reports
mechanism has not

been defined.
indicated an increased
incidence of Ebstein's
anomaly, other heart and
Methylene blue
great vessel defects, but

as more studies are
reported the strength of
this association has
diminished. Lithium
levels within the
therapeutic range (1.2
mg%) do not increase
the risk of abortion.
Hemolytic anemia and
jaundice in neonatal
information: poor to
period after exposure
fair. Used to mark
late in pregnancy. There amniotic cavity during
may be a small risk for
amniocentesis.
intestinal atresia but this
is not yet clear. No
indication of increased
risk of abortion.
Misoprostol
Misoprostol is a synthetic Quality of available

prostaglandin analog that information: Fair
has been used by
Classical animal
millions of women for
teratology studies
illegal abortion. A low
incidence of vascular
would not be helpful in

discovering these
disruptive phenomenon,
such as limb reduction
effects, because
vascular disruptive
defects and Mobius

syndrome, has been
effects occur after the

period of early
Oxazolidine-2,4-diones
reported.
Fetal trimethadione
organogenesis.
(trimethadione,
paramethadione)
syndrome: V-shaped eye information: good to

brows, low-set ears with excellent.
D-Penicillamine
Phenobarbitol
anteriorly folded helix,

high-arched palate,
Anticonvulsants; affects
cell membrane
irregular teeth, CNS

anomalies, severe
developmental delay.
Wide variation in
reported risk.
Characteristic facial
features are documented
only with chronic
exposure. The
abortifacient potential
has not been adequately
studied, but appears to
permeability. Actual
mechanism of action
has not been
determined.
be minimal.
Cutis laxa,
hyperflexibility of joints.
Condition appears to be
good. Copper chelating
reversible and the risk is

low. There are no human
data on the risk of
abortion.
agent; produces copper

deficiency inhibiting
collagen synthesis and
maturation.
May be used for the
reported for usual

prevention or reduction
of intraventricular
hemorrhage in
premature infant.
Progestins
Masculinization of female Quality of available

embryo exposed to high information: good.
doses of
sometestosterone-
Stimulates or interferes
with sex steroid
derived progestins and

may interact with
receptor-containing
tissue.
progesterone receptors
in the liver and brain
later in gestation. The
dose of progestins
present in modern oral
contraceptives presents
no masculinization or
feminization risks. All
progestins present no
risk for nongenital
malformations. Many
synthetic progestins and
natural progesterone
have been used to treat
luteal phase deficiency,
embryos implanted via invitro fertilization (IVF)
threatened abortion or
bleeding in pregnancy
with variable results.
Conversely, synthetic
progestins that interfere
with progesterone
function may cause early
pregnancyloss; RU-486 is
presently used
specifically for
thispurpose.
Retinoids, systemic
Increased risk of CNS,
(isotretinoin, etrentinate) cardio-aortic, ear, and

information: fair. Used
clefting defects. Microtia, in treatment of chronic
anotia, thymic aplasia
dermatoses. Retinoids
and other branchial arch, can cause direct
aortic arch abnormalities cytotoxicity and alter
and certain congenital
heart malformations.
programmed cell death;

affect many cell types
Exposed embryos are at but neural crest cells

greater risk for abortion. are particularly
This is plausible since
many of the
sensitive.
malformations, such as
neural tube defects, are
associated with an
increased risk of abortion.
Retinoids, topical
(tretinoin)
Epidemiological studies,
animal studies and
absorption studies in
humans do not suggest a
teratogenic risk.
Regardless of the risks
associated with
systemically
administered retinoids,
topical retinoids present
little or no risk for
intrauterine growth
information: poor.
Topical administration
of tretinoin inanimals in
therapeutic doses
isnotteratogenic,
although
massiveexposures can
produce maternal
toxicity and
reproductive effects.
More importantly,
retardation,
topical administration in
teratogenesis or abortion humans results innonbecause they are
measurable blood levels.
minimally absorbed and
Rh immune globulin
only a small percentage

of skin is exposed.
have been associated
with Rh-Ig prophylaxis
against Rh immunization.
Smoking and nicotine
Placental lesions;
intrauterine growth
retardation; increased
postnatal morbidity and
excellent. While tobacco

smoke contains many
mortality. While there

have been some studies
components, nicotine
can result in vascular
reporting increases in
spasm vasculitis which
anatomical
malformations, most
studies do not report an
has resulted in a higher

incidence of placental
pathology.
association. There is no
syndrome associated
with maternal smoking.
Maternal or placental
complications can result
Thalidomide
in fetal death. Exposures

to nicotine and tobacco
smoke are a significant
risk for pregnancy loss in
the first and second
trimester.
Limb reduction defects
(preaxial preferential
effects, phocomelia),
facial hemangioma,
esophageal or duodenal
atresia, anomalies of
external ears, eyes,
kidneys, and heart,
increased incidence of
neonatal and infant
mortality. The
thalidomide syndrome,
while characteristic and
recognizable, can be
mimicked by some
genetic diseases.
Although there are fewer
data pertaining to its
abortigenic potential,
there appears to be an
excellent. Sedative
hypnotic agent. The
etiology of thalidomide
teratogenesis has not
been definitively
determined.
increased risk of abortion.

Thyroid: iodides,
antithyroid drugs
(thioamides)
Fetal hypothyroidism or
goiter with variable
neurologic and aural
information: good.
Fetopathic effect of
damage. Maternal
hypothyroidism is
iodides and antithyroid

drugs involves
associated with an
metabolic block,
increase in infertility and decreased thyroid
abortion. Maternal intake hormone synthesis and
of 12 mg of iodide per
gland development.
day or more increases
the risk of fetal goiter.
Thioamides may cause
fetal goiter but dose can
be adjusted to minimize
this effect.
Tocolytics
Toluene
There are no reports of

adverse fetal outcome
resulting from exposure
to therapeutic doses of
terbutaline, ritodrine, or
magnesium sulfate.
Intrauterine growth
retardation; craniofacial
anomalies; microcephaly.
It is likely that high
exposures from abuse or
intoxication increase the
information: poor to
fair. Neurotoxicity is
produced in adults who
abuse toluene; a similar
effect may occur in the
risk of teratogenesis and fetus.

abortion. Occupational
exposures should present
no increase in the
teratogenic or
abortigenic risk. The
magnitude of the
increased risk for
teratogenesis and
abortion in abusers is not
known because the
exposure in abusers is
too variable.
Valproic acid
Malformations are
primarily neural tube

defects and facial
dysmorphology. The
information: good.
Anticonvulsant; little is
known about the
facial characteristics
teratogenic action of
associated with this drug valproic acid.
are not diagnostic. Small
head size and
developmental delay
have been reported with
high doses. The risk for
spina bifida is about 1%
but the risk for facial
dysmorphology may be
greater. Because
therapeutic exposures
Vitamins
increase the incidence of

neural tube defects, one
would expect a slight
increase in the incidence
of abortion.
Biotin: No adverse fetal
effects for the usual
therapeutic regimen
Cyanocobalamin: No
adverse fetal effects for
regimen.

multiple carboxylase
deficiency
vitamin B12- responsive
methylmalonic acidemia
Used for reduction in
Folic acid: The efficacy of recurrence of neural
folic acid
tube defects
supplementation for
reducing the risk of
information: good. High
neural tube defect
recurrence may be
limited to a select portion
of the population. There
are no adverse fetal
concentrations of
retinoic acid are
cytotoxic; it may
interact with DNA to
delay differentiation
and/or sinhibit protein
effects for the usual

therapeutic regimen.
synthesis. Quality of
Vitamin A: The same
available information:
malformations that have poor. Mechanism is
been reported with the
likely to involve a
retinoids have been

reported with very high
doses of vitamin A
disruption of cell
calcium regulation with
excessive doses.
(retinol). Exposures
below 10,000 IU. present
no risk to the fetus.
Vitamin A in its
recommended dose
presents no increased
risk for abortion.
Vitamin D: Large doses
given in vitamin D
prophylaxis are possibly
involved in the etiology
of supravalvular aortic
stenosis, elfin faces, and
mental retardation.
There is no data on the
abortigenic effect of
vitamin D.
Modified from Friedman JM, Prolifka JE. Teratogenic effects of drugs
(TERIS) 2nd ed. Baltimore: Johns Hopkins University Press, 2000.
P.234
P.235
P.236
P.237
P.238
Aminopterin and Methotrexate

Aminopterin and methotrexate (methylaminopterin) are folic acid antagonists that
inhibit dihydrofolate reductase, resulting in cell death during the S phase of the
cell cycle (55). Aminopterin-induced therapeutic abortions have resulted in
malformations (hydrocephalus, cleft palate, meningomyelocele and growth
retardation) in some of the abortuses (56,57,58). Three case reports of children
exposed to aminopterin in utero included observations of growth retardation,
abnormal cranial ossification, high-arched palate, and reduction in derivatives of
the first branchial arch (59). The pattern of malformations associated with
exposure to either compound has been referred to as the fetal aminopterin/
methotrexate syndrome (60). Key
P.239
features of this pattern of malformations include prenatal growth deficiency,
abnormal cranial ossification, micrognathia, small low set ears, and limb
abnormalities. There have also been three case reports to date of severe
developmental delay in children with methotrexate syndrome (61,62,63).
Methotrexate is used therapeutically as an abortifacient, treatment of rheumatoid
arthritis and other autoimmune disorders, and as an antineoplastic agent. Skalko
and Gold demonstrated a threshold effect and a dose-dependent increase in
malformations in mice exposed to methotrexate in utero (64). Although
malformations were induced in rats at doses exceeding those used in humans
(65), smaller doses than those used in humans have resulted in malformation in
rabbits (66). Analysis of human data indicate a critical period of exposure to
methotrexate from 6-8 weeks from conception at a dose above 10 mg per week
for the development of aminopterin/ methotrexate syndrome (67). The risk of
adverse effects as a result of aminopterin in the usual therapeutic range is not
know precisely but appears to be moderate to high (58).
Androgens
Masculinization of the external genitalia of the female has been reported following
in utero exposure to large doses of testosterone, methyltestosterone, and
testosterone enanthate (68,69,70). The masculinization is characterized by

clitoromegaly with or without fusion of the labia minora and no indication of
nongenital malformations. Affected females experience normal secondary sexual
development at puberty (71).
Many animal models show the masculinizing effects of androgens. Well-known
studies were performed by Greene and coworkers in the rat, Raynaud in the
mouse, Bruner and Witschi in the hamster, Jost in the rabbit, and Wells and Van
Wagenen in the monkey (72,73,74,75,76). These studies demonstrated the
masculinization of the urogenital sinus, its derivatives, and the external genitalia,
although there was little effect on the mullerian ducts, and ovarian inversion did
not occur. Based on experimental animal studies of altered sexually dimorphic
behavior in female guinea pigs, rats and monkeys (77,78,79,80,81,82),
behavioral masculinization of the female as a result of prenatal exposure to
androgens in the human will be rare. The available literature indicates that the
effects of androgens on the fetus are dependent on the dose and stage of
development during which exposure occurred.
Angiotensin Converting Enzyme Inhibitors

The first angiotensin converting enzyme (ACE) inhibitor, captopril, was introduced
in 1981 for the treatment of severe refractory hypertension. Since then the
number of ACE inhibitors has increased and now includes enalapril, lisinopril,
quinapril, perindopril, fosinopril, ramipril, and cilazapril (83). Their relative
effectiveness, combined with a paucity of side effects as compared to other
antihypertensives, have made these drugs extremely popular for the treatment of
all types of hypertension and congestive heart failure and diabetic nephropathy.
ACE inhibitors are competitive inhibitors of ACE, a carboxypeptidase that forms an
integral part of the renin angiotensin system (83). ACE catalyses the conversion of
angiotensin I to angiotensin II, one of the most potent vasoconstrictors known. It
is the same enzyme as kininase II, and also catalyses the breakdown of
bradykinin. A vasodilatory peptide itself, bradykinin stimulates the release of other
vasodilatory substances including prostaglandins and endothelium derived
relaxation factor (84). Both mechanisms of action contribute to the decrease in
blood pressure resulting from ACE inhibi-tion (85).
Although they are considered relatively safe for the treatment of hypertension, the
use of ACE inhibitors during pregnancy has been associated with adverse fetal
outcomes in both humans and experimental animals. The first case of adverse
fetal outcome in humans was reported in 1981 (86). In that report, treatment
with captopril began on the 26th week of gestation, oligohydramnios was detected
2 weeks later, and a cesarean section was performed the following week. The
child was anuric and hypotensive and died on day 7. The kidneys and bladder
were morphologically normal but hemorrhagic foci were found in the renal cortex
and medulla. Numerous cases of severe and often lethal adverse fetal effects
associated with ACE inhibitor use during pregnancy have since been reported
(87,88,89). The most consistent findings have been associated with a disruption
of fetal renal function resulting in oligohydramnios and neonatal anuria
accompanied by severe hypotension (87,89). Intrauterine growth retardation,
pulmonary hypoplasia, hypocalvaria, persistent patent ductus arteriosis, and renal
tubular dysgenesis have also been reported (89,90,91). Some of these effects
may also result from the condition for which ACE inhibitors were prescribed (87).
These effects have been associated with ACE inhibitor treatment only during the
second and third trimester. There are no reports of adverse fetal outcome
associated with ACE inhibitor use during the first trimester (89,91). As ACE
inhibitors do not appear to affect organogenesis in either humans or in animal
studies, it is not a classical teratogen. For this reason Pryde and associates (89)
have proposed the term ACE inhibitor fetopathy to describe the characteristic
syndrome that results from ACE inhibitor use during pregnancy.
The majority of adverse fetal effects associated with ACE inhibitor use during
pregnancy result from the direct therapeutic action of ACE inhibitors on the fetus.
ACE inhibitors readily cross the placenta in which they inhibit fetal ACE activity
(89,92). The decreased renal blood flow caused by vasodilation of renal efferent
arterioles results in a loss of glomerular filtration pressure leading to fetal anuria
and oligohydramnios (92,93). This in turn may result in other adverse fetal
outcomes, such as pulmonary hypoplasia. Fetal urine production and tubular
function does not begin until approximately 9-12 weeks of gestation
P.240
and probably explains the lack of adverse fetal effects when ACE inhibitor
treatment is discontinued in the first trimester. Renal dysplasia, in particular a
lack of renal proximal tubule differentiation, has also been noted in some effected
fetuses (90,93,94).
Exposure to ACE inhibitors during pregnancy has also resulted in several cases of
hypocalvaria, an ossification defect of the membranous bones of the skull that
leaves the fetal brain inadequately protected (89,90). Although the pathogenesis
is still unknown, inadequate perfusion of developing bone as a result of fetal
hypotension combined with pressure from uterine muscles as a result of
oligohydramnios may explain this defect (90,91). It has also been suggested that
ACE inhibitors may affect ossification by acting on osteoblast-derived growth
factors (90).
Despite consistent reports of adverse ace inhibitor fetopathy, there are no
controlled studies available to assess the risks associated with the use of ACE
inhibitors during pregnancy. Because there is no reported incidence of adverse
fetal effects as a result of ACE inhibitor use during the first trimester of
pregnancy, there is no contraindication to ACE inhibitors in women of reproductive
age. If the woman becomes pregnant therapy is then changed to an alternative
antihypertensive that represents less risk to the fetus.
Antibiotics
The incidence of intraamniotic infection is about 1% of all pregnancies but 3%40% of women with ruptured membranes for 24 hours or more (95).
Intraamniotic infection is associated with increased morbidity in the newborn
including pneumonia, and sepsis. There is also a significant increase in perinatal
mortality associated with intraamniotic infection although this is in part related to
prematurity.
Increased neonatal mortality and morbidity, especially from group B streptococcal
infection, can be largely prevented by intrapartum chemoprophylaxis. Neonatal
sepsis is also significantly reduced if mothers receive intrapartum antibiotics. With
few exceptions (such as streptomycin) acute exposure to antibiotics in usual
therapeutic doses poses little significant risk to the fetus, especially in comparison
to the potentially devastating effects of neonatal sepsis.
Aminoglycosides
This class of antimicrobials includes gentamicin, tobramycin, streptomycin and
kanamycin. The only drug in this category with confirmed developmental toxicity
is streptomycin. Based on case reports, there appears to be a small increased risk
of sensorineural deafness in offspring of women treated with streptomycin for
tuberculosis during pregnancy (96,97). Other congenital anomalies have not been
associated with in utero exposure to streptomycin in the human (98). Because of
the risk for ototoxicity, streptomycin should not be given during pregnancy.
However, other aminoglycosides, such as kanamycin, appears to have minimal
risk of causing similar adverse effects (99,100).
Anti-Tuberculosis Therapy
Drugs prescribed for the treatment of tuberculosis include aminoglycosides,
ethambutol, isoniazid, rifampin, and ethionamide. The ototoxic effects of
streptomycin are the only proven adverse effects of these drugs on the fetus.
Neither ethambutol nor rifampin have been associated with an increase in the
incidence of growth retardation, premature birth or malformations
(96,97,101,102).
Early reports did not associate therapeutic exposures to isoniazid with an
increased risk of malformations (96,97) but there is an unconfirmed association
with CNS dysfunction (103,104). There was one attempted suicide involving 50
tablets of isoniazid per day during the 12th week that resulted in a stillbirth with
arthrogryposis multiplex congenita syndrome (105). Isoniazid may have small
increased risk for adverse effects on the CNS but there is no apparent increase in
risk for congenital malformations or abortions with therapeutic exposures. Only

one report associated ethionamide with an increased risk of teratogenic effects
(106). However, this association is tenuous and not supported by other case
reports (107).
As a general observation, the antituberculosis drugs produce adverse effects on
the fetus of experimental animals in greater doses than those equivalent to
therapeutic exposures and not in all species. Embryotoxicity can be demonstrated
for streptomycin in the mouse, kanamycin in the guinea pig, isoniazid in the rat,
rifampin in the rat, and ethionamide in the rabbit (108,109,110). While this does
not eliminate the possibility of adverse effects on the fetus following exposure to
the other prescribed tuberculostatic medications discussed, therapeutic exposures
appear to represent a very small risk of teratogenesis and even less risk of
abortion.
Cephalosporins
Cephalosporins are frequently prescribed during pregnancy. Examples include
cephalexin (Keflex), cefixime (Suprax), and cefaclor (Ceclor). Although generally
considered safe for use in pregnancy, epidemiological studies examining the
effects of this class of antibiotics on the developing fetus are scarce. However,
available data indicate that, although a small risk cannot be completely excluded,
a high risk of congenital anomalies resulting from in utero exposure in the first
trimester is unlikely (111).
Macrolides
This category includes erythromycin, azithromycin, and clarithromycin (Biaxin).
Erythromycin does not readily cross the placenta and is generally considered safe
for use in pregnancy. There is less available data on azithromycin and
clarithromycin however existing data indicate that any potential risk is minimal. As
is the case with other sexually transmitted diseases, chlamydia infection is on the
increase. The infant most likely acquires chlamydial infection during parturition
P.241
at an incidence of approximately 50%. Erythromycin is an effective prenatal
treatment.
Metronidazole
Metronidazole (Flagyl) is used most frequently to treat trichomoniasis, various
other protozoan and anaerobic bacterial infections, and to help prevent preterm
labor. Administration is given orally, rectally and parenterally. Vaginal absorption
occurs readily but plasma levels achieved are much lower than those achieved by
other routes. There are a large number of studies examining the risk of exposure
during pregnancy to the developing fetus. Epidemiological studies, and the

majority of animal studies have not revealed an increased risk of congenital
anomalies with exposure during the first trimester (112,113,114,115,116,117), or
when used to prevent preterm labor in the third trimester (118,119). Based on
these studies therapeutic use of metronidazole during pregnancy is generally
considered safe. Although some bacterial test systems indicated mutagenicity,
there is no evidence for increased chromosomal abnormalities in human cells
(120,121,122).
Penicillin
The most frequently prescribed drugs in this category include penicillin, ampicillin,
and amoxicillin. Newer agents may also contain an agent to enhance effectiveness
against resistant microbials such as the use of amoxicillin with clavulanic acid
(Augmentin). There are no adverse fetal effects reported for any penicillins.
Among other uses penicillin is used in pregnancy to treat Treponema pallidum in
pregnant women.
Quinolones
Fluoroquinolones (ciprofloxacin, norfloxacin, ofloxacin), are used to treat urinary
infections. Reports of use in early pregnancy have not revealed any teratogenic or
developmental risk (123,124,125,126,127,128,129,130). Studies in immature
dogs and rodents revealed a potential for arthropathy and cartilage erosion (131).
Musculoskeletal dysfunction was not noted in a multicenter prospective controlled
study of 200 women exposed to fluoroquinolones during pregnancy (124). Effects
on cartilage and bone would be unlikely in the first trimester and there are few
studies addressing long term effects with late pregnancy exposure. Based on
available evidence quinolone antibiotics do not appear to pose a risk to the fetus
at therapeutic doses however more studies are needed to address the potential
for osteotoxicity in the neonate and juvenile.
Sulfonamides
Sulfonamides are usually combined with other antibiotics, such as in trimethoprimsulfamethoxazole (Bactrim). Due to the possibility of jaundice in the newborn,
sulfonamides should not be used during the last trimester of pregnancy, or during
nursing. Trimethoprim is a folate metabolism antagonist and thus has potential for
adverse embryonic and fetal effects (see Trimethoprim below).
Tetracyclines
This class of antibiotics includes tetracycline and doxycycline. Tetracycline crosses
the placenta but is not concentrated by the fetus. Tetracyclines complex with
calcium and the organic matrix of newly forming bone without altering the
crystalline structure of hydroxyapatite (132). Although tetracycline has been
shown to discolor teeth without affecting the likelihood of developing carries
(133,134), very high doses may depress skeletal bone growth. No congenital
malformations of any other organ system have been associated with antenatal
tetracycline exposures (98). Several case reports of limb reduction defects in
human embryos exposed to tetracycline are not supported by epidemiological
studies or animals studies. Therapeutic doses of tetracycline are associated with
no or minimal increased risk of congenital malformations but they are likely to
result in some degree of dental staining, which does not appear to have a
deleterious effect on the offspring.
Untreated Neisseria gonorrhoeae can lead to serious consequences for the
infected woman. Treatment with ceftriaxone plus doxycycline has no reported
adverse effects on the fetus.
Trimethoprim
Trimethoprim is an inhibitor of microbial dihydrofolate reductase, and is usually
combined with a sulfonamide for treatment of urinary tract infections. Early
studies did not report an association between exposure to trimethoprim during
pregnancy and congenital anomalies (135,136,137,138). However, more recent
studies using much larger study populations have reported increases in the
incidence of neural tube defects, orofacial clefts and cardiovascular defects among
infants exposed to trimethoprim during the first three months of pregnancy
(139,140,141,142). Use of a vitamin supplement containing folic acid during
treatment resulted in a reduced risk in these studies.
Trimethoprim was not teratogenic in rodents using less than 10 times the human
therapeutic dose for treatment of urinary tract infections, and caused
malformations and intrauterine death at high doses (>16 times therapeutic doses)
(143,144,145). The relevance of these findings to human risk at therapeutic doses
is unclear.
Due to the potential for adverse effects on folate metabolism in the pregnant
woman and fetus, trimethoprim should be avoided during pregnancy.
Antihypertensives (Excluding Angiotensin

Converting Enzyme Inhibitors)
Clonidine
Clonidine exerts its hypotensive effect by a direct alpha adrenergic agonistic
action in the CNS. It is used to treat
P.242
hypertension and may also be given epidurally with epidural anesthesia. It
appears to be relatively safe during pregnancy but there are few available data
regarding effects of exposure during the first trimester of pregnancy. When used
during labor with epidural anesthesia there are reports of a small but increased
frequency of fetal bradycardia (146,147) and, even less commonly, transient
neonatal hypotension (148,149).
Hydralazine
Hydralazine is a vasodilator often used in combination with methyldopa for the
treatment of preexisting hypertension in pregnancy and is considered to be safe.
Although there is one report of fetal thrombocytopenia (150), over 120 normal
pregnancies have been reported (151). Additionally there have been reports of an
increased frequency of fetal distress in neonates born to women treated with
hydralazine near term (152,153,154).
Methyldopa
Methyldopa is currently the safest antihypertensive drug available for use during
pregnancy (155). Methyldopa is a centrally acting adrenergic antagonist with no
reported adverse effects on the fetus or on mental and physical development.
Nifedipine
Nifedipine is a calcium channel blocker used for the treatment of preterm labor
with no reported adverse effects. The potential for adverse effects with its long
term use in the treatment of hypertension is unknown.
Propranolol
Propranolol is a -blocker useful in treating preexisting hypertension during
pregnancy. There have been no reports of associations with use in the first
trimester and congenital anomalies (156). However, prolonged use may cause
intrauterine growth retardation (157,158,159,160,161,162,163). Moreover, there
have been several reports of difficulties in perinatal adaptation associated with
maternal treatment late in pregnancy and shortly prior to delivery including apnea
and respiratory distress (159,160,164,165,166).
Aspirin
Aspirin acts principally by inhibiting prostaglandin synthesis by irreversibly
acetylating and inactivating fatty acid cyclooxygenase. Low-dose aspirin (60-150
mg per day) is used clinically in the prevention or treatment of a variety of

conditions that affect maternal and fetal health. For instance, aspirin is used for
the treatment of systemic lupus erythematosus, antiphospholipid syndrome, and
preeclampsia.
Of obvious concern is the potential for increased fetal and maternal bleeding
associated with low-dose aspirin use near the time of delivery or higher doses
used to treat preterm labor. Jankowski and associates reported a study involving
25 women threatened with premature delivery who were given 3.6 g of oral
aspirin per day for 4 successive days within 10 days prior to delivery (167). They
and others have found no adverse effect on maternal bleeding at delivery, fetal
hemorrhage, or circulatory disorders.
There is some evidence that aspirin combined with dipyridamole may prevent or
ameliorate intrauterine growth retardation associated with idiopathic
uteroplacental insufficiency. Fetal growth was also improved with aspirin (150 mg
per day) alone in fetuses with high, but not extreme, umbilical artery systolic/
diastolic ratio. Third trimester exposure to low daily doses of aspirin were not
associated with adverse fetal outcome in these studies (168,169).
High fetal losses are associated with lupus anticoagulant antibodies, anticardiolipin
antibodies and systemic lupus erythematosus. Surviving fetuses experience an
increased incidence of growth retardation, fetal distress, and preterm delivery.
Low-dose aspirin (60-80 mg per day) in combination with prednisone (20-80 mg
per day) improves pregnancy outcome and greatly reduces thrombosis (170).
Case reports suggest that 75 to 300 mg of aspirin per day in combination with
dipyridamole might reduce the risk of late fetal loss in patients with arterial
thromboembolism, thrombotic thrombocytopenia purpura, and idiopathic or
essential thrombocythemia (170).
Although doubts concerning the safety of aspirin during pregnancy have been
expressed, aspirin use during the first trimester is not associated with an
increased teratogenic risk (171). Third trimester exposure to 80 mg per day or
less are not associated with an early constriction of the ductus arteriosus (172).
Much larger doses of aspirin during the third trimester have been associated with
an increased length of gestation, duration of labor, frequency of postmaturity and
blood loss at delivery (173) and justify careful fetal surveillance.
Benzodiazepines
The benzodiazepines, such as chlordiazepoxide (Librium), diazepam (Valium),
xanax, and meprobamate, are widely used as tranquilizers during pregnancy and,
therefore, it is not surprising that they have been associated with congenital
malformations in some publications.
Chlordiazepoxide was associated with various anomalies after exposure during
early pregnancy but no syndrome was identified (98,174). Other studies were
inconclusive or found no association (175,176,177). Chlordiazepoxide appears to
have a minimal risk for congenital anomalies and no increased risk for abortion at
therapeutic doses. Higher exposures are likely to increase the risk of adverse
effects on the fetus but the magnitude of the increase is not known.
Some studies reported an association between diazepam and increased incidence
of congenital malformations
P.243
(177). However, a follow-up study found no associations (178). The majority of
studies of fetal outcome following in utero exposure to diazepam are negative
(112,175,176,177,178,179,180). Behavior alterations have been reported in
infants exposed to benzodiazepines, mostly diazepam (181), but this observation
must be confirmed and the long-term developmental outcome evaluated before it
can be appropriately interpreted. Although third trimester exposure to diazepam
can reversibly affect the fetus and neonate (182) there is minimal increased risk
of congenital malformations and no demonstrated increased risk of abortions from
therapeutic exposures.
Meprobamate has been weakly associated with a variety of congenital
malformations (183,184). Other studies found no associations (98,176). Because
of inconsistencies, the data are not sufficient to confirm or rule out a small
increase risk of malformations as a result of exposures early in pregnancy.
Benzodiazepines appear to have minimal increased risk of malformations at
therapeutic ranges; higher exposures may increase the risk. The risk for abortion
is unknown but given the widespread use of these drugs, it is unlikely that a
significant abortigenic effect would have gone unnoticed.
Caffeine
Caffeine is a methylated xanthine, which acts as a CNS stimulant. It is contained
in many beverages including coffee, tea, and colas, and chocolate. Caffeine is also
present in many over-the-counter medications, such as cold and allergy tablets,
analgesics, diuretics, and stimulants; the latter lead to relatively minimal
population intakes. Caffeine containing food and beverages are consumed in large
quantities by most of the human populations of the world. The per capita
consumption of caffeine from all sources is estimated to be about 200 mg/day, or
about 3 to 7 mg/kg per day (185). Consumption of caffeinated beverages during
pregnancy is quite common and is estimated to be approximately 144 mg per day
(186).
Current evidence, does not appear to implicate the usual exposure of caffeine as a
human teratogen, however, associations between maternal coffee drinking during
pregnancy and miscarriage or poor fetal growth have been reported in
epidemiological studies (187,188,189,190,191). In many instances these

associations are largely attributable to confounding effects of maternal cigarette
smoking or other factors. Some of these studies (192) have serious
methodological limitations. If maternal consumption of caffeine-containing
beverages in conventional amounts during pregnancy does have an association
with the rate of miscarriage or fetal growth retardation, the effect appears to be
relatively small.
In other studies no association has been found between caffeine consumption
during pregnancy and congenital defects (98,193,194). For instance, Rosenberg
and associates analyzed six selected birth defects in relation to maternal ingestion
of more than 8 mg/kg per day of tea, coffee, or cola (195). The defects were
inguinal hernia, cleft lip/cleft palate, cardiac defects, pyloric stenosis, cleft palate
(isolated) and neural tube fusion defects. None of the point estimates of relative
risk was significantly greater than unity, suggesting that caffeine was not a major
teratogen, at least for the defects evaluated.
Interpretation of the available information pertaining to the animal and human
studies regarding the teratogenicity of caffeine leads us to conclude that the usual
exposure of caffeine does not present a measurable risk in the human for any one
malformation or group of malformations. There is a clear indication that the
consumer must ingest a substantial amount of caffeine to have an effect on the
developing embryo or fetus; total consumption of 300 mg/day may be a safe
upper daily limit. Most reviewers and investigators concluded that there is a
threshold, below which caffeine does not exert a detrimental effect, and the usual
human consumption falls in this nontoxic range. The quantity of caffeine
consumed in an average cup of coffee, about 1.4 to 2.1 mg/kg (196), is believed
to be below the amount that induces congenital defects in animals. Quantities of
caffeine in tea and soft drinks would be even less.
Carbamazepine
Although epidemiological and case report studies have not yielded consistent
results, exposure to carbamazepine has been associated with minor craniofacial
defects, fingernail hypoplasia, developmental delay (197,198), reduced
birthweight, length and head circumference (199) and neural tube defects (200).
Confounding the issue is the possibility that epilepsy itself may increase the risk
for malformations (201). However, an attempted suicide involving carbamazepine
produced blood levels of 27-28 g/mL (the therapeutic range is 8-12 g/mL)
during what was estimated to be 3-4 weeks postconception (202). The fetus was
later determined to have myeloschisis with carba-mazepine the only known
exogenous risk factor. This suggests that carbamazepine has the potential to
produce neural tube defects at about two- to three-fold the therapeutic level. It
appears that the risk for minor defects is significant but the risk for all teratogenic
effects is not known. The risk for abortion is also not known but appears to be
small.
Cocaine
Cocaine (benzoylmethylecgonine) is one of the most commonly used illicit drugs
by women of reproductive age. Reported estimates for cocaine use during
pregnancy range from 3% to 17%, the highest rates occurring in inner city
populations (203). Because of its widespread use during pregnancy and the
growing cost of caring for cocaine exposed neonates, there has been increasing
concern over the risks associated with prenatal cocaine use to maternal and fetal
health. Yet despite numerous clinical studies linking prenatal cocaine use with a
variety of adverse maternal
P.244
and fetal effects, methodological limitations in these studies have made it difficult
to establish a causal relationship between these alleged effects and maternal
cocaine use. Not only are the timing, frequency and dose of cocaine use hard to
determine, but adverse effects as a result of low socioeconomic status, poor
nutrition, multiple drug use, infections, and a lack of prenatal care are difficult to
dissociate from effects as a result of cocaine use alone (203). As such the issue of
how much risk to the fetus is associated with cocaine use during pregnancy is
unresolved. Nonetheless a growing body of literature supports the concept that
cocaine is a developmental toxicant. Adverse effects attributed to prenatal cocaine
exposure include a higher incidence of spontaneous abortion, placental abruption,
still birth, prematurity, low birth weight, growth retardation, decreased head
circumference, intracerebral hemorrhage, congenital defects, neurobehavioral
abnormalities, and a possible association with increased risk of sudden infant
death syndrome (SIDS) (203,204). These effects are reduced but not eliminated
in mothers receiving appropriate prenatal care. Like other developmental toxins,
outcome is dependent on dose and time of use.
The majority of adverse effects associated with cocaine use during pregnancy
appear to be as a result of high levels of cocaine abuse in later stages of gestation
rather than in the first trimester or organogenesis (205). Moderate usage of
cocaine only in the first trimester does not appear to result in adverse fetal
outcome and may not pose an increased risk to the fetus (206).
Adverse fetal outcomes associated with maternal cocaine use are thought to
primarily result from the vasoconstrictive effects of cocaine on both the maternal
and fetal vasculature (205). Vasoconstriction of the uterine arteries, which are
normally fully dilated during pregnancy, may compromise fetal growth and
development. Studies in animals have confirmed that cocaine reduces uterine
artery and placental blood flow leading to reduced oxygen and nutrient supply to
the fetus (207). Fetal cardiovascular effects resulting from uterine
vasoconstriction include hypertension, tachycardia, hypoxia and an increase in
cerebral blood flow (207,208). Cocaine also crosses the placenta in which it has a
direct effect on the fetal vasculature flow (207). Fetal hypertension combined with
increased cerebral blood flow may result in intracerebral hemorrhage or infarction,
which has been reported to occur in cocaine exposed fetuses in both human and
animal studies (209,210,211).
Disruption of uterine and fetal vasculature may also lead to a variety of congenital
anomalies that have been associated with cocaine abuse. A significant association
between cocaine use and an increased incidence of genitourinary tract
malformations has been found (209,211,212). Other defects reported include limb
reduction defects, nonduodenal intestinal atresia, cardiac anomalies, hypospadias,
prune belly syndrome as a result of urethral obstruction, hydronephrosis and
crossed renal ectopia (208,209,211). Two cases of limb-body wall complex have
also been reported (213) With the exception of genitourinary tract malformations,
the sample size in these clinical studies has not been sufficient to determine a
statistically significant relationship between cocaine use and these congenital
anomalies (214).
Coumarin Derivatives
Nasal hypoplasia following exposure to several drugs, including warfarin, during
pregnancy was reported by DiSaia (215). Kerber and associates (216) were the
first to suggest warfarin as the teratogenic agent. Coumarin anticoagulants have
since been associated with nasal hypoplasia, calcific stippling of the secondary
epiphysis, and CNS abnormalities. Warfarin embryopathy has been described and
an overview of the difficulties in relating a congenital malformation to an
environmental cause and has been published (217,218). There is an estimated
10% risk for affected infants following exposure during the period from the eighth
through the fourteenth week of pregnancy, although this risk has been reported to
be much lower in some series, and other factors besides dose and gestational
stage seem to play a role (218). Low-dose warfarin (5 mg/day or less) throughout
pregnancy did not result in any adverse effects in 20 offspring (219).
Coumarin inhibits the formation of carboxyglutamyl residues from glutamyl
residues, decreasing the ability of proteins to bind calcium. The inhibition of
calcium binding by proteins during embryonic/fetal development, especially during
a critical period of ossification, could explain the nasal hypoplasia, stippled
calcification, and skeletal abnormalities of warfarin embryopathy (218).
Microscopic bleeding does not seem to be responsible for these problems early in
development (217).
One case report was unique in that the time of exposure to warfarin was between
8 and 12 weeks of gestation, and the infant presented Dandy-Walker
malformation, eye defects, and agenesis of the corpus callosum (220). This case
report is the clearest evidence for a direct effect of warfarin on the developing
CNS rather than an effect mediated by hemorrhage, because the exposure is well
defined and occurs before the appearance of vitamin K-dependent clotting factors.
Further supportive evidence for a direct pathogenic role of warfarin is the report of
an infant with an inherited deficiency of multiple vitamin K-dependent coagulation
factors whose congenital anomalies were similar to warfarin syndrome without
exposure to warfarin (221). The risk of stillbirths and spontaneous abortions is
increased in pregnant women treated with warfarin but the risk may be less if the
exposure is in the last half of pregnancy. The risk of adverse effects as a result of
hemorrhaging increases later in gestation.
Cyclophosphamide
Cyclophosphamide, a widely used antineoplastic agent, is also used in severe
rheumatic disease. Cyclophosphamide is likely to be teratogenic in the human but
the magnitude of
P.245
the teratogenic risk is uncertain. The reported defects include growth retardation,
ectrodactyly, syndactyly, cardiovascular anomalies, and other minor anomalies
(222,223,224). Ten normal pregnancies have been reported after
cyclophosphamide exposure (225).
The mechanism of cyclophosphamide teratogenesis was reviewed by Mirkes (24):
cytochrome P-450 monooxygenases convert cyclophosphamide to 4hydroxycyclophosphamide, which in turn breaks down to phosphoramide mustard
and acrolein. Phosphoramide mustard may produce teratogenic effects by
interacting with cellular DNA in an as yet undefined manner although acrolein acts
in a different manner, possibly by affecting sulfhydryl linkages in proteins (226).
Tissue sensitivity to phosphoramide mustard and acrolein is thought to be related
to such processes as detoxification and cellular repair.
Diethylstilbestrol
The first abnormality reported following exposure to diethylstilbestrol (DES)
during the first trimester was clitoromegaly in female newborns (227). Herbst and
associates (228,229) and Greenwald and associates (230) later reported an
association of vaginal adenocarcinoma in female offspring following first trimester
exposures. DES is the only drug with proven transplacental carcinogenic action in
the human. Almost all of the cancers occurred after 14 years of age and only in
those exposed before the 18th week of gestation. There is a 75 percent risk for
vaginal adenosis for exposures occurring before the ninth week of pregnancy; the
risk of developing adenocarcinoma is about 1:1,000 to 1:10,000 (231). While the
incidence of vaginal adenosis was related to the amount of DES administered, the
incidence of vaginal carcinoma does not appear to be related to the maternal dose.
Although there does not appear to be an adverse effect on menstrual cycle

functioning or on the rate of conception, the anatomic abnormalities of the uterus
and cervix induced by intrauterine exposure to DES, including T-shaped uterus,
transverse fibrous ridges and uterine hypoplasia, cause the increased incidence of
ectopic pregnancies, spontaneous abortions and premature delivery in
pregnancies of women exposed to DES in utero (232,233,234,235).
There have been reports that males exposed to DES in utero exhibited genital
lesions and abnormal spermatozoa (236,237). Other studies reported no increase
in the risk for the male for genitourinary abnormalities or infertility (238). An
association between in utero exposure to DES and testicular cancer in male
offspring has been suggested but the data are not conclusive (238,239). The
controversial nature of the effects of DES exposure on the male may be
attributable to study design or, more likely, to the fact that dose levels varied
greatly according to different regimens: exposures during the first half of
pregnancies varied from 1.5 to 150 mg per day with total doses from 135 mg to
18 g (231).
DES is a potent nonsteroidal estrogen and, as in the case of steroidal estrogens,
must interact with the receptor proteins present only in estrogen-responsive
tissues before exerting its effects by stimulating ribonucleic acid (RNA), protein,
and DNA synthesis. The carcinogenic effect of DES is most likely indirect: DES
exposure results in the presence of columnar epithelium in the vagina, and this
misplaced tissue may have a greater susceptibility to developing the
adenocarcinoma, much as teratomas and other misplaced tissues are more
susceptible to malignant degeneration.
Digoxin
Digoxin is used to correct fetal tachyarrhythmias with no substantiated adverse
fetal effects (240). However, the possibility of any adverse side effects must be
balanced with the fetal prognosis if the dysrhythmia persists or is likely to lead to
fetal cardiac failure.
Diphenylhydantoin
Hanson and Smith (241) characterized the fetal hydantoin syndrome in infants
whose mothers were treated for epilepsy with hydantoin anticonvulsants. Chronic
exposure to diphenylhydantoin has been suggested to present a maximum of 10%
risk for the full syndrome and a maximum of 30% risk for some anomalies
(242,243,244,245). Although cleft lip and palate, congenital heart disease, and
microcephaly have been reported, hypoplasias of the nails and distal phalanges
are possibly more common malformations in the exposed fetuses (246,247).
Hanson and associates noted that, although the hydantoin syndrome is observed
in 11% of the subjects in their study, three times that number exhibit mental
deficits (31,248). Prospective studies demonstrate a much lower frequency of

effects, and some do not demonstrate any effect; thus, the overall prospective
risk may be much lower for the classically reported effects.
Factors associated with epilepsy may contribute to the etiology of these
malformations (249). Based on the United States Collaborative Perinatal Project
and a large Finnish registry, the incidence of malformations was 10.5% when the
mother was epileptic, 8.3% when the father was epileptic and 6.4% when neither
parent was affected (250).
Cleft lip and palate, skeletal anomalies, and cardiac defects have been produced in
rabbits (31,251) mice (252,253,254,255), and rats (256,257), and the
malformation rate was dose-dependent (253,258).
Glucocorticoids
Glucocorticoids (dexamethasone, betamethasone, hydrocortisone,
methylprednisone) are effective in reducing the incidence of respiratory distress
syndrome in premature newborns by inducing early lung maturation as first
hypothesized by Liggins (259). Endogenous glucocorticoids mediate normal
pulmonary maturation. Exogenous glucocorticoids are used to stimulate the
production of surfactant. The adverse fetal effects observed in experimental
P.246
animals exposed to pharmacological doses are not seen in humans at therapeutic
levels (260,261,262).
Dexamethasone is used to suppress the fetal adrenal gland in cases of congenital
adrenal hyperplasia (263). 21-Hydroxylase deficiency impairs the conversion of
cholesterol to cortisol and results in excess 17-hydroxyprogesterone which in turn
results in excess levels of androgens. The masculinization of female fetuses with
congenital adrenal hyperplasia varies from clitoral hypertrophy to formation of a
phallus. Maternal replacement doses of dexamethasone suppress both the
maternal and fetal adrenal glands and prevent masculinization in most patients
(264).
Glucocorticoids are also used in the treatment of rheumatic diseases, other acute
and chronic inflammatory diseases, and organ transplantation. Although some
studies have reported increased frequency of perinatal death, prematurity and
intrauterine growth retardation with chronic therapy, women in these studies
typically had severe autoimmune or other diseases, or history of fetal loss
requiring treatment, thus these findings cannot be attributed to the glucocorticoid
therapy itself (223,265,266,267,268,269,270,271). Dexamethasone poses
minimal or no teratogenic risk at therapeutic doses in humans and the benefits of
glucocorticoid treatment, particularly dexamethasone, for prevention of
masculinization as a result of congenital adrenal hyperplasia, and betamethasone
for the prevention of the respiratory distress syndrome in premature infants, and
potentially other complications arising from prematurity are clear
(272,273,274,275,276).
Indomethacin
Oral administration of the prostaglandin synthetase inhibitor, Indomethacin, is a
prostaglandin synthetase inhibitor that is used as an analgesic, an
antiinflammatory, and an antipyretic agent. Indomethacin is also effective in the
treatment of polyhydramnios that is either idiopathic or related to maternal
diabetes mellitus. Administration of indomethacin during the first trimester has
not been shown to increase the frequency of congenital malformations (112). Late
in pregnancy maternal administration of indomethacin may cause
oligohydramnios, constriction of the ductus arteriosus (prostaglandins are
necessary to maintain the patency of the fetal ductus arteriosus), fetal hydrops
and persistent pulmonary hypertension in the newborn, potentially serious side
effects of indomethacin that warrant careful fetal and neonatal surveillance
(277,278,279,280). The risk for premature closure of the ductus arteriosus
increases with treatment after 32 weeks (281,282,283). Indomethacin may also
be used to prevent preterm labor or intraventricular hemorrhage; however its
efficacy when used for this purpose is controversial. Indomethacin may also
predispose the neonate to necrotizing enterocolitis when used as a tocolytic
(284,285).
Lithium Carbonate
Lithium carbonate, widely used for treatment of manic-depressive disorders, was
first associated with human congenital malformations in 1970 (286,287). The
malformations described include heart and large-vessel anomalies, Epstein's
anomaly, neural tube defects, talipes, microtia, and thyroid abnormalities
(288,289,290). Lithium readily crosses the placenta (291), and appears to be a
human teratogen at therapeutic dosages but it presents a small risk. Although
early reports suggested a strong association of prenatal lithium exposure with
cardiac defects, in particular Epstein's anomaly, more recent evidence from
controlled epidemiological studies suggests that the risk for malformations is
much lower than initially thought (292,293). The results of a retrospective study
suggest that lithium may also increase the risk for premature delivery (294), but
again the magnitude of the risk is likely to be small. Only one follow-up study has
been published examining long-term effects of lithium on early development. In
this study children exposed prenatally to lithium with no congenital abnormalities
at birth did not show any signs of developmental delay at 7.3 years follow-up
(295). Fetal toxicity has been associated with late gestational maternal lithium use
with and without obvious maternal toxicity. One reported side effect is
nephrogenic diabetes insipidus (296,297,298) and associated polyhydramnios
which may increase the likelihood of premature labor (294,297). Transient toxic
effects have also been reported in neonates exposed late in pregnancy. These
include hypothyroidism, lethargy, hypostomia, cardiac murmur, renal toxicity,
persistent fetal circulation and diabetes insipidus (299,300,301,302). To prevent
lithium intoxication in the neonate the lithium dosage of the patient should be
adjusted to avoid high serum levels in the second and third trimester (293).
Lithium can induce abnormal development in several laboratory animals, but the
mechanisms of the teratogenic action of lithium is not known (303,304,305). The
neurotropic activity of lithium suggests that CNS malformations may result from
cell membrane disturbances which affect neural tube closure (306).
Because of the value of lithium carbonate for treating manic-depressive psychosis,
the risk associated with psychiatric relapse on removing the drug may be more
important clinically than the teratogenic risk. Moreover, the risk of alternative
pharmacological agents for treatment of bipolar disorder may exceed the risk from
lithium carbonate (307).
Methylene Blue
Methylene blue has been used clinically for a variety of purposes including the
identification of anatomic structures, the treatment of methemoglobinemia, and to
mark the amniotic cavity during amniocentesis. Use of methylene blue in late
gestation to detect rupture of fetal membranes has been associated with adverse
fetal effects including hyperbilirubinemia, hemolytic anemia and staining of the
skin (308,309,310,311). There is currently not enough data to determine whether
respiratory distress in these infants may also result from late gestation methylene
blue exposure (312).
P.247
There have been several reports of an increased prevalence of small intestinal
atresia in twins with intraamniotic exposure to methylene blue. Twinning itself
results in an increased prevalence of intestinal atresia, increasing from
approximately 2 to 2.5 per 10,000 in singletons to 5 to 7.3 per 10,000 amongst
twins (313). However, in twins exposed to midgestational amniocentesis in which
methylene blue was used to mark the amniotic cavity the prevalence of small
intestinal atresia has been reported as high as 9.6% (314). The strongest
evidence indicating that methylene blue is a teratogen is a retrospective study
from Amsterdam. In this study methylene blue was injected into one amniotic
cavity of 86 twin pregnancies undergoing midgestation amniocentesis. Jejunal
atresia occurred in 17 infants, each from different pregnancies (314). In 15 of
these cases it was possible to determine which twin was exposed to methylene
blue, and in each case the twin exposed to methylene blue had jejunal atresia.
Based on this evidence and several other reports there appears to be a significant
risk of small intestinal atresia associated with exposure to methylene blue during
midgestation amniocentesis (311,313,315,316,317,318,319). Intraamniotic

exposure to methylene blue has not been associated with any other malformation
(312). Whether midgestation exposure to methylene blue increases the incidence
of fetal death has not been clearly established (320,321). In rats, intraamniotic
injection of methylene blue increased fetal loss (322).
Misoprostol
Misoprostol is a synthetic prostaglandin E1 methyl analogue used for the
prevention of gastric ulcers induced by nonsteroidal antiinflammatory drugs. It
has known, but not very effective, abortifacient properties. Gonzalez and
associates (323) recently reported seven newborns with vascular disruptive
phenomena (limb reduction defects, Moebius syndrome) whose mothers used
misoprostol early in pregnancy in an attempt to induce abortion. Although there is
evidence that misoprostol is used illegally by thousands of pregnant Brazilian
women as an abortifacient (324,325,326), controlled cohort or case control
epidemiological studies of the fetal outcome of failed abortions are not available.
Although the data available are not conclusive, the uterine bleeding produced by
misoprostol and type of malformations produced suggest a vascular disruption
mechanism for misoprostol induced teratogenesis.
If one is looking for vascular disruption, it will more likely be produced later in
gestation. Therefore, classical animal teratology experiments will not detect the
vascular disruptive effect of drugs or chemicals unless they are exposed beyond
the period of early organogenesis (327). Furthermore, it has become clear that if
an agent produces vascular disruption, it is a rare event and therefore large
populations would need to be studied before the effect may be discovered (328).
Previous case reports are also of little assistance. Collins and Mahoney (329)
reported an infant with hydrocephalus and attenuated digital phalanges after
exposure intravaginally to 15-methyl F2alpha prostaglandin five weeks after
conception. Schuler and associates (330) reported that 29% of women who used
misoprostol in Brazil as an abortifacient failed to abort. Seventeen children who
failed to abort were observed to have no malformations. Wood and associates
(331) reported an infant exposed to oxytocin and prostaglandin E2 for the
purpose of termination to have hydrocephaly and growth retardation. Schonhofer
(332) and Fonesca and associates (333) reported five Brazilian infants with
defects of the skull and overlying scalp who were exposed to misoprostol in utero.
These case reports indicate the low risk of misoprostol exposure and the
possibility that some of the features reported may or may not be as a result of
misoprostol (334). It is too early to know the extent of the effects of misoprostol,
but it is biologically plausible that they should include all of the features of
vascular disruption.
Oxazolidine-2,4-diones (Trimethadione,
Paramethadione)
Trimethadione and paramethadione are antiepileptic oxazolidine-2,4-diones that
distribute uniformly throughout body tissues and exert their effects by means of
the action of their metabolites. These drugs affect cell membrane permeability
and vitamin K-dependent clotting factors, but their primary mode of action is
unknown.
Zackai and associates (335) described the fetal trimethadione syndrome
characterized by developmental delay, V shaped eyebrows, low-set ears with
anteriorly folded helix, high arched palate and irregular teeth. Clinical
observations of these and other associated findings, such as cardiovascular,
genitourinary and gastrointestinal anomalies, have been reviewed
(335,336,337,338). The incidence of miscarriage, stillbirth and infant death was
also increased. There are wide variations in reported risk, with estimates as high
as 80% for major or minor defects. Because the number of exposures is small, the
actual risk could vary considerably from these figures. Although there are
variations in incidences reported amongst studies, the risk of malformation or
other adverse fetal effects is considered high, therefore the drug should not be
used in pregnant women.
D-Penicillamine
D-Penicillamine has been used in the treatment of rheumatoid arthritis and
cystinuria. D-Penicillamine is a copper chelator and copper deficiency appears to
be the mechanism for teratogenicity (339). Exposure to D-penicillamine can
induce a connective tissue defect including generalized cutis laxa, hyperflexibility
of the joints, varicosities and impaired wound healing (340,341,342). The
exposure must be long enough to induce a copper deficiency sufficient to inhibit
collagen synthesis and maturation. However, the condition appears to be
reversible and the risk is low, 5% or less.
P.248
Phenobarbital
Phenobarbital is used as a sedative and anticonvulsant. Risk of congenital
anomalies is unlikely with occasional exposure to phenobarbital in therapeutic
doses during pregnancy. Use of phenobarbital for long term treatment of seizure
disorders such as epilepsy may be associated with a small to minimal risk of
increase in congenital malformations such as cleft palate and congenital heart
defects (177,250,343,344,345,346,347). Some of these effects may relate
directly to the disease being treated itself, rather than to phenobarbital (348). The
risk appears to be greater in cases in which phenobarbital was combined with

other drugs such as phenytoin (344,346,349,350,351). Some studies report no
increased risk for congenital anomalies with use of phenobarbital for treatment of
epilepsy. However, other studies report an increased incidence of minor congenital
anomalies with phenobarbital monotherapy with a characteristic pattern of minor
dysmorphic features, termed the fetal anticonvulsant syndrome including nail
and midface hypoplasia and depressed nasal bridge (345,352,353,354).
In two prospective, studies of patients in high risk groups for neonatal
intraventricular hemorrhage, intravenous phenobarbital administered antenatally
resulted in a significant reduction in the incidence of severe cases of
intraventricular hemorrhage by increasing cerebral vascular resistance (355).
Phenobarbital reduces peak arterial blood pressure, thereby reducing the risk for
intraventricular hemorrhage, one of the leading complications of the very low
birth-weight preterm infant (356). Initiation of intravenous phenobarbital in the
early neonatal period is probably too late to derive the potential benefits of this
therapy.
Progestins
It is often overlooked that, although various progestins utilized therapeutically as
progestational agents act by means of similar receptors, their potential androgenic
effects can differ markedly. This point is critical to the evaluation of the virilizing
effects of these compounds in the human. It has been shown, for example, that
the pharmacokinetic parameters that estimate steroid bioavailability and
metabolism show great variability among subjects and between steroids
conveniently grouped together, such as progestins (357). One must assume that
these differences in bioavailability and metabolism reflect differences in the
biological activity of these steroids in humans.
In contrast to progesterone and 17-hydroxyprogesterone caproate, high doses of
some of the synthetic progestins have been reported to cause virilizing effects in
humans. Exposure during the first trimester to large doses of 17ethinyltestosterone has been associated with masculinization of the external
genitalia of female fetuses (358). Similar associations result from exposure to
large doses of 17-ethinyl-19-nortestosterone (norethindrone) (358) and 17ethinyl-17-OH-5 (10)estren-3-one (Enovid-R) (68). The synthetic progestins, like
progesterone, can influence only those tissues with the appropriate steroid
receptors. The preparations with androgenic properties may cause abnormalities
in the genital development of females only if present in sufficient amounts during
critical periods of development. In 1959, Grumbach and associates (68) pointed
out that labioscrotal fusion could be produced with large doses if the fetuses were
exposed before the thirteenth week of pregnancy, whereas clitoromegaly could be
produced after this period, illustrating that a specific form of maldevelopment can
be induced only when the embryonic tissues are in a susceptible stage of
development.
The World Health Organization (359) reported that there is a suspicion that
combined oral contraceptives or progestogens may be weakly teratogenic but that
the magnitude of the relative risk is small. In a large retrospective study,
Heinonen and associates (360) reported a positive association between
cardiovascular defects and in utero exposure to female sex hormones. A
revaluation of some of the base data by Wiseman and Dodds-Smith, (361)
however, did not support the reported association. Another retrospective study
conducted by Ferencz and associates (188) did not find a positive association
between female sex hormone therapy and congenital heart defects. Although
neither study disproved the positive association reported by Heinonen and
associates (360) their findings made the association less likely.
Epidemiological studies have reported an association between exposures to female
sex hormones, oral contraceptives or progestogens, and congenital neural tube
defects (362) and limb defects (363). Further studies and reevaluations have not
supported either of these associations (9,21,22).
Further support for the absence of a nongenital effect of progestins comes from
(a) a negative correlation between sex hormone usage during pregnancy and
malformations, (b) no increased incidence in malformations following
progesterone therapy to maintain pregnancy, and (c) no increased incidence in
malformations following first trimester exposure to progestogens (mostly
medroxyprogesterone) administered to pregnant women who had signs of
bleeding. The Food and Drug Administration has recognized that the evidence
does not support an increased risk of limb reduction defects, congenital heat
disease, or neural tube defects following exposure to oral contraceptives or
progestins (364).
It is generally accepted that the actions of steroid hormones are mediated by
specific steroid (365) and therefore only those tissues with the specific receptors
can be affected by steroid hormones.
Retinoids, Systemic Administration (Isotretinoin,

Etretinate)
Vitamin A congeners, including retinol, retinal, all-trans-retinoic acid (tretinoin)
and 13-cis-retinoic acid (isotretinoin), are all teratogenic in numerous species.
Both isotretinoin (Accutane, Roche Laboratories), marketed for treating severe
P.249
acne, and etretinate (Tigason, Sautier Laboratories), marketed for treating
psoriasis, contained warnings by the manufacturers against exposure during
pregnancy. Unfortu-nately, exposures occurred. Analyses of the resulting
malformations have been reviewed (366,367). Human malformations include CNS,
cardioaortic, microtia, clefting defects and, more controversially, limb defects.

Isotretinoin has a serum half-life of 10 to 12 hours; there is no apparent risk to
the fetus if maternal use is terminated before conception (368). However, the risk
for malformations and subnormal results on standard intelligence tests is high for
offspring of women who were treated with therapeutic doses of isotretinoin in the
first 60 days after conception (369).
Etretinate exposure during the first 60 days of pregnancy is associated with a high
risk of malformations. Etretinate persists in the body for up to 2 years (370,371).
Whether the lower concentrations of etretinate remaining in the mother's
circulation over these extended periods increases the risk of malformations is not
known but, if there is an increased risk, it is likely to be low.
Experimental evidence suggests that endogenous retinoic acid may act as a
natural morphogen. Exogenous retinoids act either directly, resulting in
cytotoxicity, or via receptor-mediated pathways to interact with DNA and alter
programmed cell death (372,373). It is likely that different specificities of retinoid
binding proteins (374) account for the variations in placental transfer. Although
retinoids can influence many types of cells, Lammer (366) emphasized that
neuroectodermally derived cells of the rhombencephalon are particularly sensitive
and that the resulting neural crest cell abnormality differs from that resulting in
oculoauriculo-vertebral dysplasia or Goldenhar syndrome. The susceptibility of
specific cell types to the effects of the retinoids may be determined by the
intracellular concentration of cellular retinoic binding protein (366).
Retinoids, Topical Administration (Tretinoin)

There have been a number of case reports of congenital malformations occurring
in the offspring of mothers who used topical tretinoin during their pregnancy. The
United States Food and Drug Administration received adverse reaction reports
involving approximately seventeen infants who were born from mothers who used
topical tretinoin during pregnancy, with a higher than expected representation of
holoprosencephaly (Rosa, F., personal communication). While there has not been
an abundance of epidemiological studies involving topical tretinoin, there are three
reports. Using Michigan Medicaid data, the incidence of birth defects in 147
pregnancies exposed to topical tretinoin was compared to the incidence of birth
defects in 104,092 nonexposed; the relative risk was 0.8 in the exposed
population (Rosa, F., personal communication). A relative risk of 0.7 for birth
defects was determined in 215 pregnancies exposed to topical tretinoin as
compared with 430 nonexposed mothers in data from Group Health of Puget
Sound (375). De Wals and associates (376) evaluated the association of the
occurrence of holoprosencephaly with topical tretinoin exposure. Among 502,189
births there were 31 infants with holoprosencephaly. Eight patients had an
abnormal karyotype and 16 had a normal karyotype. None of the patients with a
normal karyotype were exposed to topical tretinoin during pregnancy.
Since all teratogens that have been appropriately studied have a no-effect dose, it
would be paramount that topical administration of a known teratogen such as
tretinoin must be absorbed and produce teratogenic concentrations in the blood.
At conventional doses, the blood levels from topical administration are far below
the teratogenic dose. It would appear that prudent use of this topical medication
presents no risk to the embryo, because there would be no teratogenic exposure.
The pharmacokinetics, animal studies and human studies support this conclusion.
Rh Immune Globulin
After exposure to Rh(D)-positive red cells, usually resulting from a fetal
transplacental hemorrhage that occurs to some degree in 75% of pregnancies,
(377) the Rh(D)-negative mother becomes Rh immunized. Rh immunization
during a previous pregnancy results in brain damage of various degree or death in
an Rh(D)-positive newborn in approximately 50% of cases (377). Once maternal
Rh immunization has developed, it cannot be treated effectively, but it can be
prevented by antenatal prophylaxis with 300 ug of RhIg at 28 weeks' gestation
(377). No adverse fetal effects to immunoprophylaxis have been reported.
Smoking and Nicotine

Approximately 30% of all women of childbearing age smoke, and about 25% of all
women will continue to smoke after they become pregnant (378). The evidence in
humans indicates that smoking affects the fetus directly in a dose-related manner
and probably involves more than one component of smoke (379). Placental
lesions and fetal growth retardation have been consistently reported in
epidemiological studies involving pregnant women who smoke cigarettes
(380,381,382). Fetal death is 20%-80% higher among women who smoked
cigarettes although pregnant (382). However, despite suggestions that smoking
during pregnancy may increase the risk of limb anomalies, there is no proven
relationship between smoking and specific malformations or malformations in
general (383,384,385). Because of the large number of pregnant women who
smoke and the documented effects of smoking on the fetus, one can conclude that
smoking presents a significant risk to the fetus for growth retardation and
abortion.
Thalidomide
Lenz and Knapp (386) were the first to associate thalidomide exposure during
pregnancy with limb reduction defects and other features of the thalidomide
syndrome. Limb defects resulted from exposure limited to a 2-week period from
the 22nd to the 36th days postconception:
P.250
exposures from the 27th to the 30th days most often affected only the arm,
whereas exposures from the 30th to the 33rd days resulted in both leg and arm
abnormalities (387,388). Although there was no association of mental retardation,
brain malformations, or cleft palate, other abnormalities included facial
hemangioma, microtia, esophageal or duodenal atresia, deafness, and anomalies
of the eyes, kidneys, heart, and external ears and increased incidence of
miscarriages and neonatal mortality (386,387,389,390). A high proportion, about
20%, of the fetuses exposed during the critical period were affected. The current
use of thalidomide in Brazil for the treatment of leprosy has resulted in more
recent cases of embryopathy including at least 29 children born with thalidomide
syndrome (391,392). Although the mechanism of teratogenic action for
thalidomide is not yet defined, the subject has been critically reviewed by
Stephens (393).
Thyroid: Iodides, Antithyroid Drugs (Thioamides)

There are several case reports of congenital goiter as a result of in utero
exposures to iodide-containing drugs (394). Maternal intake of as low as 12 mg
per day may result in fetal goiter (394). Iodinated diagnostic x-ray contrast
agents used for amniofetography have been reported to affect fetal thyroid
function adversely (395).
Thioamides are antithyroid drugs used to treat fetal thyrotoxicosis. Fetal
thyrotoxicosis is induced by thyroid-stimulating immunoglobulins produced in
euthyroid or hypothyroid women. The adverse fetal effects, e.g., craniosynostosis,
intellectual impairment and increased mortality, are caused by excess fetal thyroid
hormone production. The thioamides block thyroid hormone synthesis by
inhibiting the oxidation of iodide or iodotyrosyl. Unlike other thioamides,
propylthiouracil also inhibits the peripheral deiodination of thyroxine to
triiodothyronine. All thioamides are associated with a significant risk of fetal goiter
and teratogenesis (290). However, in the case of propylthiouracil, the fetal goiter
can be reduced with intraamniotic injections of thyroxine (396) which also
prevents other abnormalities caused by inhibition of fetal thyroid function.
Tocolytics
Fetal distress can result from uterine hypertonus, umbilical cord compression,
premature rupture of the fetal membranes, oligohydramnios, placental abruption,
and uteroplacental insufficiency. In some cases of severe fetal hypoxemia or
acidosis, prompt delivery may be recommended. However, if immediate surgery is
not feasible, tocolytics may help to reduce fetal distress until delivery is possible.
In cases of hypoxemia as a result of reduced blood flow to the fetus, inhibiting
uterine activity should increase the delivery of oxygen to the fetus by increasing
uterine and intervillous perfusion. In addition to inhibiting uterine activity, adrenergic agonists both increase maternal cardiac output and dilate uterine
vessels resulting in a further increase in placental perfusion.
Ritodrine, a 2-adrenergic receptor agonist, may be administered as an
intravenous bolus for acute fetal distress (397). Ritodrine's mechanism of action
leads to a reduction in the intracellular calcium available for smooth muscle
contraction.
Terbutaline sulfate, a nonspecific beta-adrenergic agonist, is associated with a
higher incidence of cardiovascular side effects than ritodrine with prolonged use.
When the use of -adrenergic agonists is contraindicated in cases of intraamniotic
infection, uncontrolled maternal thyroid disease, diabetes mellitus, and
cardiovascular disease, magnesium sulfate may be used as a tocolytic agent.
Although its mechanism of action is unknown, it results in an uncoupling of the
actin-myosin interaction in smooth muscle (398). An advantage of magnesium
sulfate tocolysis is the absence of cardiovascular side effects.
The data concerning the fetal effects of tocolytic agents are restricted to case
reports, but, there are no reports of adverse fetal outcome resulting from
exposure to therapeutic doses of terbutaline (399), ritodrine (400), or magnesium
sulfate (401).
Toluene
Although occupational exposure to toluene has not been associated with
congenital malformations in offspring, there are case reports of malformations
resulting from the abuse of toluene. The first description of an infant with features
similar to FAS born to a chronic abuser of toluene appeared in 1979 (402). This
case and 22 additional cases have been described in detail (403). Thirty-nine
percent of the toluene-exposed infants were born prematurely and 9% died in the
perinatal period. In the surviving infants, 52% exhibited growth deficiency, 67%
were microcephalic, and 80% exhibited developmental delay. Craniofacial features
similar to those in the FAS were observed in 89%. An increased incidence of
prematurity, perinatal death, growth and developmental delay and phenotypic
features similar to FAS were reported in 35 pregnancies of 15 toluene abusers
(404). Pearson and associates (403) suggest that the clinical and experimental
data can be interpreted to imply that alcohol and toluene may have a common
mechanism of facial teratogenesis. Toluene appears to have the potential for
developmental toxicity in the human but the magnitude of the risk is minimal for
usual occupational exposures, although it may be moderate to high in inhalation
abusers.
Tranquilizers
The minor tranquilizers as a group are probably the most frequently prescribed
therapeutic agents. Within this group, the propanediol carbonates and the
benzodiazepines, the two most widely used classes, have been associated with
teratogenic effects (290). The strongest association has been between diazepam
and cleft lip with
P.251
or without cleft palate but even this association is not likely to be causal (405).
Since these drugs are widely used, even a small increased risk would be expected
to result in more reported adverse effects than has been the case. Continued
surveillance is warranted because so many pregnancies are exposed to these
drugs.
Valproic Acid
Valproic acid (dipropylacetic acid) is used for the treatment of various types of
epilepsy. Dalens and associates (406), were the first to report the association of
valproic acid and congenital malformations in the human. Although other reports
followed, Robert and colleagues (407) described the associated malformations,
consisting primarily of neural tube defects, and their incidence in detail. The
neural tube defect observed is usually spina bifida in the lumbar or sacral region
and increased risk appears to be correlated with higher serum levels (350,408).
Other anomalies include postnatal growth retardation, microcephaly, midface
hypoplasia, micrognathia and epicanthal folds. Therapeutic dosages during
pregnancy present a teratogenic risk for spina bifida of about 1% (409) but the
risk for facial dysmorphology may be greater. Valproic acid crosses the human
placenta (410) but the fetal serum concentrations are not known.
Vitamins
Biotin
Biotin-responsive multiple carboxylase deficiency is an inborn error of metabolism
in which there is a severe reduction in the activities of the mitochondrial biotindependent carboxylase enzymes. Affected individuals exhibit dermatitis, severe
metabolic acidosis and a characteristic pattern of organic acid excretion.
Metabolism in these patients is restored to normal levels by biotin
supplementation. Prenatal administration of 10 mg per day of oral biotin initiated
during the third trimester prevented neonatal complications with no adverse fetal
effects (390).
Vitamin A
Case reports have associated congenital defects in humans with massive vitamin
A ingestion during pregnancy (411). Although effects similar to those produced by

vitamin A congeners, namely CNS, cardioaortic, microtia and clefting defects (see
discussion of Retinoids above), may be predicted, no pattern of anomalies has
emerged. Another fourteen infants of women who ingested high doses of vitamin
A (25,000 IU or more) during pregnancy had no congenital anomalies, although
three additional pregnancies ended in miscarriage (412).
A review of literature concerning retinoids and birth defects entitled
Recommendations for Vitamin A Use During Pregnancy was published by the
Teratology Society (411). Supplementation of 8,000 IU vitamin A per day should
be the maximum during pregnancy and high dosages (25,000 IU or more) are not
recommended. Beta-carotene as a source of vitamin A is likely to be associated
with a smaller risk than an equivalent dose of vitamin A as retinol (411).
Vitamin B12
Ampola and associates (413) were the first to report prenatal treatment of a
vitamin-responsive inborn error of metabolism. Their report involved a fetus with
a vitamin B12-responsive variant of methylmalonic acidemia, a metabolic disease
involving a functional deficiency in the coenzymatically active form of vitamin B12.
Oral cyanocobalamin, 10 mg per day, initiated at 32 weeks gestation resulted in
only a slight increase in maternal serum B12 level. Oral therapy was therefore
stopped at 34 weeks' gestation and 5 mg per day of intravenous cyanocobalamin
was initiated. This regimen produced a progressive increase in maternal serum
B12 and a decrease in urinary methylmalonic acid excretion. The infant had no
acute neonatal complications after delivery at 41 weeks.
Vitamin D
There have been no large-scale studies examining the effects of ingestion of large
doses of vitamin D during pregnancy. In several small studies no increased risk of
malformations was seen in offspring of women who took large doses of vitamin D
during pregnancy (414,415). There have been some reports that huge doses of
vitamin D administered for rickets prophylaxis in pregnant women resulted in a
marked increased incidence of a syndrome consisting of supravalvular aortic
stenosis, elfin facies, and mental retardation similar to Williams syndrome in the
human (416,417). However, more recent studies have demonstrated that Williams
syndrome is caused by a gene deletion in most cases (418,419). Animal studies
and additional clinical reports suggest that the teratogenic risk of therapeutic
doses of vitamin D is none to minimal.
ACKNOWLEDGMENTS
The authors thank Yvonne Edney for her secretarial assistance.
REFERENCES
1. Hertig AT. The overall problem in man. In: K. Benirschke, ed. Comparative
Aspects of Reproductive Failure. Berlin: Springer-Verlag, 1967:11-41.
2. Robert CJ, Lowe CR. Where have all the conceptions gone? Lancet
1975;1:498-499.
3. Kajii T, Ferrier A, Niikawa N, et al. Anatomic and chromosomal anomalies in

639 spontaneous abortions. Hum Genet 1980; 55:87.
4. Simpson JL. Genes, chromosomes and reproductive failure. Fertil Steril

1980;33:116-778.
5. Copp AJ. Death before birth: clues from gene knockouts and mutations.
Trends Genet 1995;11.
6. Fraser FC. The multifactorial/threshold concept-uses and misuses. Teratology

1976;14:762-770.
P.252
7. World Health Organization. Spontaneous and induced abortion. World Health
Organization: Geneva, 1970.
8. Boue J, Boue A, Lazar P. Retrospective and prospective epidemiological

studies of 1,500 karyotyped spontaneous abortions. Teratology 1975;12:11-26.
9. Wilson JG. Environment and birth defects. New York: Academic Press, 1973.
10. Brent RL. Predicting teratogenic and reproductive risks in humans from
exposure to various environmental agents using in vitro techniques and in vivo
animal studies. Congenit Anom Kyoto 1988;28[Suppl]:S41-S55.
11. Sever JL. Infections in pregnancy: highlights from the collaborative

perinatal project. Teratology 1982;25.
12. Gwinn M, Pappaioanou M, George JR, et al. Prevalence of HIV infection in

childbearing women in the United States. JAMA 1991;265.
13. Braddick MR, Kreiss JK, Embree JE, et al. Impact of maternal HIV infection
on obstetrical and early neonatal outcome. AIDS 1990;4.
14. Blanche S, Rouzioux C, Moscato MLG, et al. A prospective study of infants

born to women seropositive for human immunodeficiency virus type 1. N Engl J
Med 1989;320.
15. Embree JE, Braddick MR, Datta P, et al. Lack of correlation of maternal
human immunodeficiency virus infection with neonatal malformations. Pediatr
Infect Dis J 1989;8.
16. European Collaborative Study. Mother-to-child transmission of HIV infection.

Lancet 1988;II.
17. Qazi QH, Sheikh TM, Fikrig S, et al. Lack of evidence for craniofacial
dysmorphism in perinatal human immunodeficiency virus infection. J Pediatrics
1988;112.
18. Van Allen MI. Structural anomalies resulting from vascular disruption.
Pediatr Clin N Am 1992;39:255-277.
19. Van Allen MI JS-B, Dixon J, et al. Construction bands and limb reduction
defects in two newborns with fetal ultrasound evidence for vascular disruption.
Am J Med Genet 1992;44.
20. Wilson JG, Brent RL, Jordan HC. Differentiation as a determinant of the
reaction of rat embryo to x-irradiation. Proc Soc Exp Biol Med 1953;82.
21. Briggs MH, Briggs M. Sex hormone exposure during pregnancy and
malformations. In: Briggs MH, Corbin A, eds. Advances in steroid biochemistry
and pharmacology. London: Academic Press, 1979:51-89.
22. Wilson JG, Brent RL. Are female sex hormones teratogenic? Am J Obstet
Gynecol 1981;114:567-580.
23. Fraser FC. Interactions and multiple causes. In: Wilson JG, Fraser FC, eds.
Handbook of teratology. New York: Plenum Press, 1977:445-463.
24. Mirkes PE. Cyclophosphamide teratogenesis: a review. Teratog Carcinog

Mutagen 1985;5:75-88.
25. Miller RK. Placental transfer and function: the interface for drugs and
chemicals in the conceptus. In: Fabro S, Scialli AR, eds. Drug and chemical
action in pregnancy: pharmacologic and toxicologic principles. New York: Marcel
Dekker, 1986:123-152.
26. Jackson MJ. Drug absorption. In: Fabro S, Scialli AR, eds. Drug and
chemical action in pregnancy: pharmacologic and toxicologic Principles. New
York and Basel: Marcel Dekker, 1986:15-36.
27. Mattison DR. Physiologic variations in pharmacokintics during pregnancy.

In: Fabro S, Scialli AR, eds. Drug and Chemical Action in Pregnancy:
pharmacologic and toxicologic principles. New York: Marcel Dekker 1986:37102.
28. Sonawane BR, Yaffe SJ. Physiologic disposition of drugs in the fetus and
newborn. In: Fabro S, Scialli AR, eds. Drug and chemical action in pregnancy:
pharmacologic and toxicologic principles. New York and Basel: Marcel Dekker,
1986:103-121.
29. Juchau MR, Rettie AE. The metabolic role of the placenta. In: Fabro S,
Scialli AR, eds. Drug and chemical action in pregnancy: pharmacologic and
toxicologic principles. New York and Basel: Marcel Dekker, 1989:153-169.
30. Juchau MR. Bioactivation in chemical teratogenesis. Ann Rev Pharmacol

Toxicol 1989;29.
31. Hanson JW. Teratogen Update: fetal hydantoin effects. Teratology

1986;33:349-353.
32. Laurence KM, James N, Miller MH, et al. Double-blind randomized controlled
trial of folate treatment before conception to prevent recurrence of neural tube
defects. Br Med J 1981;282: 1509-1511.
33. Smithells RW, Seller MJ, Nevin NC, et al. Further experience of vitamin
supplementation for prevention of neural tube defect recurrences. Lancet
1983;1:1027-1031.
34. Smithells RW, Shepard S, Schorah CJ, et al. Apparent prevention of neural
tube defects by periconceptional vitamin supplementation. Arch Dis Child
1981;56:911.
35. Smithells RW, Sheppard S, Wild J, et al. Prevention of neural tube defect
recurrences in Yorkshire: final report. Lancet 1989; 2:498-499.
36. Medical Research Council. Prevention of neural tube defects: results of the
Medical Research Council Vitamin Study. Lancet 1991;338:131-137.
37. Cziezel AE, Dudas I. Prevention of the first occurrence of neural-tube

defects by periconceptional vitamin supplementation. N Engl J Med 1992;327.
38. Brent RL. Environmental factors: miscellaneous. In: Brent RL, Harris MI,
eds. Prevention of embryonic fetal and perinatal disease. Bethesda: DHEW
(NIH), 1976:211-218.
39. Brent RL. Drugs and pregnancy: are the insert warnings too dire? Contemp
Ob/Gyn 1976;20:42-49.
40. Mirkin BI. Maternal and fetal distribution of drugs in pregnancy. Clin
Pharmacol Ther 1973;14:643-647.
41. McKusick VA. Mendalian inheritance in man: catalogs of autosomal

dominant, autosomal recessive, and X-linked phenotypes. Baltimore: Johns
Hopkins University Press, 1988.
42. Brent RL. Method of evaluating alleged human teratogens [Editorial].

Teratology 1978;17:83.
43. Brent RL. Definition of a teratogen and the relationship of teratogenicity to

carcinogenicity [Editorial]. Teratology 1986;34: 359-360.
44. Susser E. Spontaneous abortion and induced abortion: an adjustment for

the presence of induced abortion when estimating the rate of spontaneous
abortion from cross-sectional studies. Am J Epidemiol 1983;117:305-308.
45. Olsen J. Calculating the risk ratios for spontaneous abortions: the problem
of induced abortion. Int J Epidemiol 1984;13: 347-350.
46. Warner RH, Rosett HL. The effects of drinking on offspring: an historical
survey of the American and British literature. J Stud Alcohol 1975;36:1395.
47. Jones KL, Smith DW, Streissguth AP, et al. Outcome in offspring of chronic
alcoholic women. Lancet 1974;1:1076-1078.
48. Streissguth AP, Grant TM, Barr HM, et al. Cocaine and the use of alcohol
and other drugs during pregnancy. Am J Obstet Gynecol 1991;164:1239-1243.
49. Streissguth AP, Sampson PD, Marr HM. Neurobehavioral dose-response

effects of prenatal alcohol exposure in humans from infancy to adulthood. Ann
NY Acad Sci 1989;562:145-158.
50. Streissguth AP, Landesman-Dwyer C, Martin JC, et al. Teratogenic effect of

alcohol in humans and laboratory animals. Science 1980;209:353-361.
51. Day NL, Richardson GA. Prenatal alcohol exposure: a continuum of effects.
52. Autti-Ramo I, Granstrom M-L. The effect of intrauterine alcohol exposure of

various durations on early cognitive development. Neuropediatrics 1991;22:203210.
53. Autti-Ramo I, Granstrom M-L. The psychomotor development during the

first year of life of infants exposed to intrauterine alcohol of various durations.
Fetal alcohol exposure and development. Neuropediatrics 1991;22:59-64.
54. Clarren SK, Smith DW. The fetal alcohol syndrome. N Engl J Med
1978;298:1063-1067.
55. Skipper HT, Schabel FM Jr. Quantitative and cytokinetic studies in

experimental tumor models. In: Holland JF, Frei E III, eds. Cancer medicine.
Philadelphia: Lea and Bebieger, 1973:629-650.
56. Thiersch JB. Therapeutic abortions with a folic acid (4-amino PGA). Am J
Obstet Gynecol 1952;63:1298-1304.
57. Goetsch C. An evaluation of amniopterin as an abortifacient. Am J Obstet

Gynecol 1962;83:1474-1477.
58. Warkany J. Aminopterin and methotrexate: folic acid deficiency. Teratology

1978;17:353-358.
59. Warkany J, Beautry PH, Horstein S. Attempted abortion with amniopterin (4aminopteroylglutamic acid). Am J Dis Child 1959;97:274-281.
60. Jones KL. Fetal amniopterin/methotrexate syndrome. In: Jones KL, ed.
Smith's Recognizable patterns of human malformation. Philadelphia: WB
Saunders, 1997:570-571.
P.253
61. Del Campo M, Kosaki K, Bennett FC, et al. Developmental delay in fetal
aminopterin/methotrexate syndrome. Teratology 1999;60: 10-12.
62. Bawle EV, Conard JV, Weiss L. Adult and two children with fetal
methotrexate syndrome. Teratology 1998;57:51-55.
63. Shaw EB, Steinbach HL. Amniopterin-induced fetal malformation: survival of

infant after attempted abortion. Am J Dis Child 1968;115:477-482.
64. Skalko RG, Gold MP. Teratogenicity of methotrexate in mice. Teratology

1974;9:159-164.
65. Baranov VS. Characteristics of the teratogenic effect of aminopterin

compared to that of other teratogenic agents. Bull Exp Biol Med 1966;61:77-81.
66. Goeringer GC, DeSesso JM. Developmental toxicity in rabbits of the
antifolate aminopterin and its amelioration by leucovorin. Teratology

1990;41:560-561.
67. Feldkamp M, Carey JC. Clinical teratology counseling and consultation case
report: low dose methotrexate exposure in the early weeks of pregnancy.
Teratology 1993;47:533-539.
68. Grumbach MM, Conte FA. Disorders of sex differentiation. In: Williams RH,
ed. Textbook of endocrinology. Philadelphia: WB Saunders, 1981:422-514.
69. Moncrieff A. Nonadrenal female pseudohermaphroditism associated with

hormone administration in pregnancy. Lancet 1958; 2:267.
70. Hoffman F, Overzier C, Uhde G. Zur frage der hormonalen erzengung fotaler
zwittenbildugen beim menschen. Geburtshilfe Frauerheikd 1955;15:1061-1070.
71. Reschini E, Giustina G, D'Alberton A, et al. Female pseudohermaphroditism

due to maternal androgen administration: 25-year follow-up. Lancet
1985;1:1226.
72. Greene RR, Burrill MW, Ivy AC. Experimental intersexuality: the effect of
antenatal androgens on sexual development of female rats. Am J Anat
1939;65:415-469.
73. Raynaud A. Observations dur de development normal des ebauches de la

glande mammaire des foetus maleset femelle de souris. Ann Endocrinol
1947;8:349-359.
74. Brunner JA, Witschi E. Testosterone-induced modifications of sexual

development in female hamsters. Am J Anat 1946;79: 293-320.
75. Jost A. Problems of fetal endocrinology: the gonadal and hypophyseal

hormones. Recent Prog Horm Res 1953;8:379-418.
76. Wells, LJ, Van Wagenen G. Androgen induced female

pseudohermaphroditism in the monkey (macaca mulatta) anatomy of the
reproductive organs. Carnegie Institute Contrib Embryol 1954;35: 93-106.
77. Dohler KD, Hancke JL, Srivastava SS, et al. Participation of estrogens in
female sexual differentiation of the brain: neuroanatomical, neuroendocrine and
behavioral evidence. Prog Brain Res 1984;6:99-117.
78. Goy RW, Bercovitch FB, McBrair MC. Behavioral masculinization is

independent of genital masculinization in prenatally androgenized female rhesus
macaques. Horm Behav 1988;22:552-571.
79. Goy RW, Bridson WE, Young WC. Period of maximal susceptibility of the
prenatal female guinea pig to masculinizing actions of the testosterone
propionate. J Comp Physiol Psychol 1964;57:166-174.
80. Hoepfner BA, Ward IL. Prenatal and neonatal androgen exposure interact to
affect sexual differentiation in female rats. Behav Neurosci 1988;102:61-65.
81. Huffman L, Hendricks SE. Prenatally injected testosterone propionate and

sexual behavior of female rats. Physiol Behav 1981;26: 773-778.
82. Phoenix CH, Goy RW, Gerall AA, et al. Organizing action of prenatally
administered testosterone propionate on the tissues mediating mating behavior
in the female guinea pig. Endocrinology 1959;65:369-382.
83. Maxwell SRJ, Kendall MJ. ACE inhibition in the 1900s. Br J Clin Pract
1993;47:30-37.
84. German J, Kowal A, Ehlers KH. Trimethadione and human teratogenesis.

Teratology 1970;3:349-362.
85. Gavras I, Gavras H. Ace inhibitors: a decade of clinical experience. Hosp

Pract (Off Ed) 1993;28:117-127.
86. Guignard JP, Burgener F, Calame A. Persistent anuria in neonate: a side

effect of captopril. Int J Pediatr Nephrol 1981;2:133.
87. Hanssens M, Keirse MJNC, Vankelecom F, et al. Fetal and neonatal effects of
treatment with angiotensin-converting enzyme inhibitors in pregnancy. Obstet
Gynecol 1991;79:128-135.
88. Piper JM, Ray WA, Rosa FW. Pregnancy outcome following exposure to
angiotensin converting enzyme inhibitors. Obstet Gynecol 1992;80:429-432.
89. Pryde PG, Sedman AB, Nugent CE, et al. Angiotensin-converting enzyme
inhibitor fetopathy. J Am Soc Nephrol 1993;3:1575-1582.
90. Barr M Jr, Cohen MM Jr. ACE inhibitor fetopathy and hypocalvaria: the
kidney-skull connection. Teratology 1991;44:485-495.
91. Brent RL, Beckman DA. Angiotensin-converting enzyme inhibitors, an

embryopathic class of drugs with unique properties: information for clinical
teratology counselors. Teratology 1991;43: 543.
92. Guignard JP. Effect of drugs on the immature kidney. Adv Nephrol Necker
Hosp 1993;22:193-211.
93. Martin RA, Jones KL, Mendoza A, et al. Effect of ACE inhibition in the fetal
kidney: decreased renal blood flow. Teratology 1992; 46:317-321.
94. Cunniff C, Jones K, Phillipson J, et al. Oligohydramnios sequence and renal

tubular malformation associated with maternal enalapril use. Am J Obstet
Gynecol 1990;162:187-189.
95. Cox SM, Williams ML, Leveno KJ. The natural history of preterm ruptured
membranes: what to expect of expectant management. Obstet Gynecol
1988;71:558-562.
96. Snider DE, Layde PM, Johnson MW, et al. Treatment of turberculosis during
pregnancy. Am Rev Respir Dis 1980;122:65-79.
97. Warkany J. Antituberculosis drugs. Teratology 1979;20:133-138.
98. Heinonen OP, Slone D, Shapiro S, eds. Birth Defects and Drugs in
Pregnancy. Littleton: Publishing Sciences Group, 1977.
99. Jones HG. Intrauterine toxicity: a case report and review of literature. J Natl
Med Assoc 1973;65:201-203.
100. Nishimura H, Tanimura T, eds. Clinical aspects of the teratogenicity of

drugs. New York: American Elsevier, Excerpta Medica, 1976.
101. Bobrowitz ID. Ethambutol in pregnancy. Chest 1974;66:20-24.
102. Lewit T, Nebel L, Terracina S, et al. Ethambutol in pregnancy: observations

on embryogenesis. Chest 1974;66:25-26.
103. Monnet P, Kalb JC, Pujol M. Harmful effects of isoniazid on the fetus and
infants. Lyon Med 1967;218:431-455.
104. Varpela E. On the effect exerted by the first line turberculosis medicines on
the fetus. Acta Tuberc Pneumol Scand 1964;35:53-69.
105. Lenke RR, Turkel SB, Monsen R. Severe fetal deformities associated with
ingestion of excessive isoniazid in early pregnancy. Acta Obstet Gynecol Scand
1985;64:281-282.
106. Potworowska M, Sianoz-Ecka E, Szufladowicz R. Treatment with

ethionamide in pregnancy. Gruzlica 1966;34:341-347.
107. Zierski M. Effects of ethionamide on the development of the human fetus.

Gruzlica 1966;34:349-352.
108. Dluzniewski A, Gastol-Lewinska L. The search for teratogenic activity of

some tuberculostatic drugs. Diss Pharm Pharmacol 1971;23:383-392.
109. Nakamoto Y, Otani H, Tanaka O. Effects of aminoglycosides administered

to pregnant mice on postnatal development of inner ear in their offspring.
Teratology 1985;2:604-605.
110. Steen JSM, Stainton-Eldis DM. Rifampicin in pregnancy. Lancet

1977;2:604-605.
111. Friedman JM, Prolifka JE. Teratogenic effects of drugs (TERIS) 2nd ed.
Baltimore: Johns Hopkins University Press, 2000.
112. Aselton P, Jick H, Milunsky A, et al. First-trimester drug use and congenital
disorders. Obstet Gynecol 1985;65:451-455.
113. Burtin P, Taddio A, Ariburnu O, et al. Safety of metronidazole in

pregnancy: a meta analysis. Am J Obstet Gynecol 1995;172: 525-529.
114. Caro-Paton T, Carvajal A, Martin de Diego I, et al. Is metronidazole

teratogenic? A meta analysis. Br J Clin Pharmacol 1997;44: 179-182.
115. Morgan I. Metronidazole treatment in pregnancy. Int J Gynaecol Obstet

1978;15:501-502.
116. Piper J, Mitchel E, Ray W. Prenatal use of metronidazole and birth defects:
no association. Obstet Gynecol 1993;82:348-352.
117. Rosa F, Baum C, Shaw M. Pregnancy outcomes after first trimester

vaginitis drug therapy. Obstet Gynecol 1987;69:751-755.
118. Svare J JL-R, Anderson L, et al. Ampicillin-metronidazole treatment in

idiopathic preterm labor: a randomized controlled multicentre trial. Br J Obstet
Gynaecol 1997;104:892-897.
119. Norman K, Pattinson R, de Souza J, et al. Ampicillin and metronidazole

treatment in preterm labor: a multicentre, randomized controlled trial. Br J
120. Dobias L, Cerna M, Rossner P, et al. Genotoxicity and carcinogenicity of

metronidazole. Mutat Res 1994;317:177-194.
P.254
121. Roe F. Toxicologic evaluation of metronidazole with partucular reference to
carcinogenic, mutagenic, and teratogenic potential. Surgery 1983;93:158-164.
122. Roe F. Safety of nitroimidizoles. Scand J Infect Dis 1985;[Suppl 46]:72-81.
123. Schaefer C, Amoura-Elefant E, Vial T, et al. Pregnancy outcome after

prenatal quinolone exposure. Evaluation of a case registry of the European
Network of Teratology Information Services. Eur J Obstet Gynecol Reprod Biol
1996;69:83-89.
124. Loebstein R, Addis A, Ho E, et al. Pregnanacy outcome following

gestational exposure to fluoroquinolones: a multicenter prospective controlled
study. Antimicrob Agents Chemother 1998;42: 1336-1339.
125. Berkovitch M, Pastuszak A, Gazarian M, et al. Safety of the new quinolones

in pregnancy. Obstet Gynecol 1994;84:535-538.
126. Koren G. Use of the new quinolones in pregnancy. Can Fam Physician
1996;42:1097-1099.
127. Pagnini G, Pelagalli GV, Di Carlo F. Effect of nalidixic acid on the chick
embryo and on pregnancy and embryonic development in rabbits and rats. Atti
Soc Ital Sci Vet 1971;25:137-140.
128. Sato T, Kaneko Y, Saegusa T, et al. Reproduction studies of cinoxacin in

rats. Chemotherapy 1980;28[Suppl 4]:484-507.
129. Sato T, Kobayashi F. Teratological study on cinoxacin in rabbits.

Chemotherapy 1980;28 [Suppl 4]:508-515.
130. Murray EDS. Nalidixic acid in pregnancy. Br Med J 1981;282:224.
131. Ingham B, Brentnall DW, Dale EA, et al. Arthropathy induced by

antibacterial fused N-alkyl-4-pyridone-3-carboxylic acids. Toxicol Lett
1977;6:21-26.
132. Cohlan SQ, Bevelander G, Tiamsic T. Growth inhibition of prematures

receiving tetracycline: clinical and laboratory investigation. Am J Dis Child
1963;105:453-461.
133. Baden E. Environmental pathology of the teeth. In: Gorlin RJ, Goldman
HM, eds. Thomas' oral pathology. St Louis: Mosby, 1970: 189-191.
134. Rebich T, Kumar J, Brustman B. Dental caries and tetracycline-stained

dentition in an American-Indian population. J Dent Res 1985;64:462-464.
135. Colley DP, Kay J, Gibson GT. A study of the use in pregnancy of cotrimozazole and sulfamethizole. Aust J Pharm 1982;63: 570-575.
136. Bailey RR. Single-dose antibacterial treatment for bacteriuria in pregnancy.

Drugs 1984;27:183-186.
137. Williams JD, Brumfitt W, Condie AP, et al. The treatment of bacteriuria in
pregnanct women with sulphamethoxazole and trimethoprim. Postgrad Med J
1969;45[Suppl]:71-76.
138. Bailey RR, Bishop V, Peddie PA. Comparison of single dose with a 5-day
course of co-trimoxazole for asymptomatic (covert) bacteriuria of pregnancy.
Aust NZ J Obstet Gynaecol 1983;23: 139-141.
139. Hernandez-Diaz S, Mitchell AA. Folic acid antagonists during pregnancy

and risk of birth defects. N Engl J Med 2001;344:934- 935.
140. Hernandez-Diaz S, Werler MM, Walker AM, et al. Folic acid antagonists
during pregnancy and the risk of birth defects. N Engl J Med 2000;343:16081614.
141. Hernandez-Diaz S, Werler MM, Walker AM, et al. Neural tube defects in
relation to use of folic acid antagonists during pregnancy. Am J Epidemiol
2001;153:961-968.
142. Cziezel AE, Rockenbauer M, Sorensen HT, et al. The teratogenic risk of
trimethoprim-sulfonamides: a population based case control study. Reprod
Toxicol 2001;15:637-646.
143. Elmazar MMA, Nau H. Trimethoprim potentiates valproic acid-induced

neural tube defects (NTDs) in mice. Reprod Toxicol 1993;7:249-254.
144. Kreutz R. Investigation on the influence of trimethoprim at the intrauterine

development of the rat. Anat Anz 1981;149.
145. Udall V. Toxicology of sulphonamide-trimethoprim combinations. Postgrad

Med J 1969;45[Suppl]:42-45.
146. Chassard D, Mathon L, Dailler F, et al. Extradural clonidine combined with

sufentanil and 0.0625% bupivacaine for analgesia in labour. Br J Anaesth
1996;77:458-462.
147. Cigarini I, Kaba A, Bonnet F, et al. Epidural clonidine combined with

bupivicaine for analgesia in labor. Effects on mother and neonate. Reg Anesth
1995;20:113-120.
148. Boutroy MJ, Gisonna CR, Legagner M. Clonidine: placental transfer and
neonatal adaptation. Early Hum Dev 1988;17: 275-286.
149. Horvath JS, Phippard A, Korda A, et al. Clonidine hydrochloride-a safe and
effective antihypertensive agent in pregnancy. Obstet Gynecol 1985;66:634638.
150. Widerlov E, Karlman I, Storsater J. Hydralazine-induced neonatal

thrombocytopenia [Letter]. N Engl J Med 1980;301:1235.
151. Bott-Kanner G, Schweitzer A, Reisner SH, et al. Propranolol and

hydralazine in the management of essential hypertension in pregnancy. Br J
152. Derham RJ, Robinson J. Severe preeclampsia: is vasodilation therapy with

hydralazine dangerous for the preterm fetus? Am J Perinatol 1990;7:239-244.
153. Kirshon B, Wasserstrum N, Cotton DB. Should continuous hydralazine

infusions be utilized in severe pregnancy-induced hypertension? Am J Perinatol
1991;8:206-208.
154. Spinnato JA, Sibai BM, Anderson GD. Fetal distress after hydralazine
therapy for severe pregnanacy-induced hypertension. South Med J
1986;79:559-562.
155. NHBPEP. Working Group Report on High Blood Pressure in Pregnancy,
Public Health Service. In: National Institutes of Health publication, no. 91-2039.
Bethesda, MD: NIH, 1991, p. 38.
156. Cziezel AE. Teratogenicity of ergotomine. J Med Genet 1989;26: 69-70.
157. Eliahou HE, Silverberg DS, Reisen E. Propranalol for the treatment of
hypertension in pregnancy. Br J Obstet Gynaecol 1978;85: 431-436.
158. Lieberman BA, Stirrat GM, Cohen SL, et al. The possible adverse effect of
propranolol on the fetus in pregnancies complictaed by severe hypertension. Br
J Obstet Gynaecol 1978;85: 678-683.
159. Oakley GD, McGarry K, Limb DG, et al. Management of pregnancy in

patients with hypertrophic cardiomyopathy. Br Med J 1979;1:1749-1750.
160. Pruyn SC, Phelan JP, Buchanan GC. Long-term propranolol therapy in
pregnancy: maternal and fetal outcome. Am J Obstet Gynecol 1979;135:485489.
161. Redmond GP. Propranolol and fetal growth retardation. Semin Perinatol
1982;6:142-147.
162. Paran E, Holzberg G, Mazor M, et al. Beta-adrenergic blocking agents in

the treatment of pregnancy-induced hypertension. Int J Clin Pharmacol Ther
1995;33:119-123.
163. Witter FR, King TM, Blake DA. Adverse effects of cardiovascular drug
therapy on the fetus and neonate. Obstet Gynecol 1981;58: 100S-105S.
164. Tunstall ME. The effect of propranolol on the onset of breathing at birth. Br
J Anaesth 1969;41:792.
165. Habib A, McCarthy JS. Effects on the neonate of propranolol administered

during pregnancy. J Pediatr 1977;91:808-811.
166. Rubin PC. Beta blockers in pregnancy. N Engl J Med 1981;305: 1323-1326.
167. Jankowski A, Skublicki S, Wichlinski LM, et al. Clinical pharmacokinetic

investigations of acetylsalicylic acid in cases of imminent premature delivery. J
Clin Hosp Pharm 1985; 10:361.
168. Trudinger BJ, Cook CM, Giles WB, et al. Low-dose aspirin in pregnancy.
Lancet 1989;1:410.
169. Wallenburg HC, Rotmans N. Prevention of recurrent idiopathic fetal growth

retardation by low-dose aspirin and dipyridamole. Am J Obstet Gynecol
1987;157:1230.
170. Barton JR, Sibai BM. Low-dose aspirin to improve perinatal outcome. Clin
171. Werler MM, Mitchell A, Shapiro S. The relation of aspirin use during the
first trimester of pregnancy to congenital cardiac defects. N Engl J Med
1989;321:1639.
172. McParland P, Pearce JM, Chamberlain GVP. Doppler ultrasound and aspirin
in recognition and prevention of pregnancy-induced hypertension. Lancet
1990;335:1552.
173. Lewis RB, Schulman JD. Influence of acetylsalicylic acid, an inhibitor of

prostaglandin synthesis, on the duration of human gestation and labour. Lancet
1973;2:1159.
174. Kullander S, Kallen B. A prospective study of drugs and pregnancy. Acta

Obstet Gynecol Scand 1976;55:25-33.
175. Czeizel A. Lack of evidence of teratogenicity of benzodiazepine drugs in

Hungary. Reprod Toxicol 1988;1:183-188.
176. Hartz SC, Heinonen OP, Shapiro S, et al. Antenal exposure to

meprobamate and chloridiazepoxide in relation to malformations, mental
development, and childhood mortality. N Engl J Med 1975;292:726-728.
P.255
177. Rothman K, Fuyler D, Goldblatt A, et al. Exogenous hormones and other

drug exposures of children with congenital heart disease. Am J Epidemiol
1979;109:433-439.
178. Zierler S, Rothman KJ. Congenital heart disease in relation to maternal use
of Bendectin and other drugs in early pregnancy. N Engl J Med 1985;313:347352.
179. Safra MJ, Oakley GP. Association between cleft lip with or without cleft
palate and prenatal exposure to diazepam. Lancet 1975; 2:478-479.
180. Tikkanen J, Heinonen OP. Risk factors for conal malformations of the
heart. Eur J Epidemiol 1992;8:48-57.
181. Laegried L, Hagberg G, Lundberg A. Neurodevelopment in late infancy

after prenatal exposure to benzodiazepinesa prospective study.
Neuropediatrics 1992;23:60-67.
182. Rementeria JL, Bhatt K. Withdrawal symptoms in neonates from

intrauterine exposure to diazepam. J Pediatr 1977;90:123-126.
183. Milkovich L, van den Berg BJ. Effects of prenatal meprobamate and
chlordiazepoxide hydrochloride on human embryonic and fetal development. N
Engl J Med 1974;291:1268-1271.
184. Saxen I. Association between oral clefts and drugs taken during
pregnancy. Int J Epidemiol 1975;4:37-44.
185. Barone JJ, Roberts H. Human consumption of caffeine. In: Dewes PB, ed.
Caffeine. New York: Springer-Verlag, 1984:59-73.
186. Morris MB, Weinstein L. Caffeine and the fetusis trouble brewing? Am J
187. Beaulac-Baillargeon L, Desrosiers C. Caffeine-cigarette interaction on fetal

growth. Am J Obstet Gynecol 1987;157:1236-1240.
188. Ferencz C, Matanoski GM, Wilson PD, et al. Maternal hormone therapy and
congenital heart disease. Teratology 1980;21:225-239.
189. Srisuphan W, Bracken MB. Caffeine-consumption during pregnancy and

association with late spontaneous abortion. Am J Obstet Gynecol 1986;154:1420.
190. Watkinson B, Fried PA. Maternal caffeine use before, during and after
pregnancy and effects upon offspring. Neurobehav Toxicol Teratol 1985;7:9-17.
191. Wilcox AJ, Weinberg CR, Baird DD. Risk factors for early pregnancy loss.
Epidemiology 1990;1:382-385.
192. Berger A. Effects of caffeine consumption on pregnancy outcomea

review. J Reprod Med 1988;33:945-956.
193. Kurppa K, Holmberg PC, Kuosma E, et al. Coffee consumption during

pregnancy and selected congenital malformations: a nationwide case-control
study. Am J Public Health 1983;73: 1397-1399.
194. van't Hoff W. Caffeine in pregnancy. Lancet 1982;1:1020.
195. Rosenberg L, Mitchell AA, Shapiro S, et al. Selected birth defects in

relation to caffeine-containing beverages. JAMA 1982;247: 1429-1432.
196. Felts JH. Coffee arabica. N C Med J 1981;42:281.
197. Jones KL, Lacro RV, Johnson KA, et al. Pattern of malformations in the
children of women treated with carbamazepine during pregnancy. N Engl J Med
1989;320:1661-1666.
198. Nielsen M, Froscher W. Finger and toenail hypoplasia after carbamazepine

monotherapy in late pregnancy. Neuropediatrics 1985;16:167-168.
199. Bertollini R, Kallen B, Mastroiacovo P, etal. Anticonvulsant drugs in

monotherapy: effect on the fetus. Eur J Epidemiol 1987;3:164-167.
200. Rosa FW. Spina bifida in infants of women treated with carbamazapine
during pregnancy. N Engl J Med 1991;10:674-677.
201. Janz D. Antiepileptic drugs and pregnancy: alered utilization patterns and
teratogenesis. Epilepsia 1982;23:S53-S63.
202. Little BB, Santos-Ramos R, Newell JF, et al. Megadose carbamazapine

during the period of neural tube closure. Obstet Gynecol 1993;82:705-708.
203. Slutsker L. Risk associated with cocaine use during pregnancy. Obstet
Gynecol 1992;79:778-779.
204. Young SL, Vosper HJ, Phillips SA. Cocaine: its effects on maternal and child
health. Pharmacotherapy 1992;12:2-17.
205. Jones KL. Developmental pathogenesis of defects associated with prenatal

cocaine exposure: fetal vascular disruption. Clin Perinatol 1991;18:139-146.
206. Koren G, Graham K. Cocaine in pregnancy: analysis of fetal risk. Vet Hum
Toxicol 1992;34:263-264.
207. Woods JR, Plessinger MA. Maternal-fetal cardiovascular system: a target of

cocaine. NIDA Res Monogr 1991;108:7-27.
208. Plessinger MA, Woods JR. Maternal, placental, and fetal pathophysiology of
cocaine exposure during pregnancy. Clin Obstet Gynecol 1993;36:267-278.
209. Chasnoff IJ, Chisum GM, Kaplan WE. Maternal cocaine use and
genitourinary tract malformations. Teratology 1988;37:201- 204.
210. Dogra VS, Menon PA, Poblete J, et al. Neurosonographic imaging of small
for gestational age neonates exposed and not exposed to cocaine and
cytomegalovirus. J Clin Ultrasound 1994; 22:93-102.
211. Hoyme EH, Jones KL, Dixon SD. Prenatal cocaine exposure and prenatal
vascular disruption. Pediatrics 1990;85:743.
212. Chavez GF, Mulinare J, Cordero JF. Maternal cocaine use during early
pregnancy as a risk factor for congenital urogenital anomalies. JAMA
1989;262:795-798.
213. Viscarello RR, Ferguson DD, Nores J, et al. Limb-body wall complex
associated with cocaine abuse: further evidence of cocaine's teratogenicity.
214. Lutiger BK, Graham K, Einarson TR, et al. Relationship between gestational
cocaine use and pregnancy outcome: a meta analysis. Teratology 1991;44:405414.
215. DiSaia PJ. Pregnancy and delivery of a patient with a Starr-Edwards mitral
valve prosthesis: report of a case. Obstet Gynecol 1966;29:469-472.
216. Kerber IJ, Warr OS, Richardson C. Pregnancy in a patient with prosthetic
mitral valve. JAMA 1968;203:223-225.
217. Barr M, Burdi AR. Warfarin-associated embryopathy in a 17-week abortus.

Teratology 1976;14:129-134.
218. Hall JG, Pauli RM, Wilson RM. Maternal and fetal sequelae of
anticoagulation during pregnancy. Am J Med 1980;68:122-140.
219. Cotrufo M, deLuca TSL, Calabro R, et al. Coumarin anticoagulation during

pregnancy in patients with mechanical valve prostheses. Eur J Cardiothorac
Surg 1991;5:300-305.
220. Kaplan LC. Congenital Dandy Walker malformation associated with first
trimester warfarin: a case report and literature review. Teratology 1985;32:333337.
221. Pauli RM, Lian JB, Mosher DF. Association of congenital deficiency of
multiple vitamin K-dependent coagulation factors and the phenotype of the
warfarin embryopathy: clues to the mechanism of coumarin derivatives. Am J
Hum Genet 1987;41:566- 583.
222. Greenberg LH, Tanaka KR. Congenital anomalies probably induced by
cyclophosphamide. JAMA 1964;188:423-426.
223. Scott JR. Fetal growth retardation associated with maternal administration
of immunosuppressive drugs. Am J Obstet Gynecol 1977;128:668-676.
224. Toledo TM, Harper RC, Moser RH. Fetal effects during cyclophosphamide
and irradiation therapy. Ann Intern Med 1971;74:87-91.
225. Blatt J, Mulvihill JJ, Ziegler JL, et al. Pregnancy outcome following cancer
chemotherapy. Am J Med 1980;69:828-832.
226. Hales BF. Effects of phosphoramide mustard and acrolein, cytotoxic

metabolities of cyclophosphamide, on mouse limb development in vitro.
Teratology 1989;40:11-20.
227. Bongiovanni AM, DiGeorge AM, Grumbach MM. Masculinization of the

female infant associated with estrogenic therapy alone during gestation: four
cases. J Clin Endocrinol Metab 1959; 19:1004-1011.
228. Herbst AL, Ulfelder H, Poskanzer DC. Adenocarcinoma of the vagina:

association of maternal stilbestrol therapy with tumor appearance in young
women. N Engl J Med 1971;284:878-881.
229. Herbst AL, Kurman RJ, Scully RE, et al. Clear-cell adenocarcinoma of the
genital tract in young females. N Engl J Med 1972; 287:1259-1264.
230. Greenwald P, Barlow JJ, Nasca PC, et al. Vaginal cancer after maternal
treatment with synthetic estrogens. N Engl J Med 1971; 285:390-392.
231. Herbst AL, Robboy SJ, Scully RE, et al. Clear-cell adenocarcinoma of the
vagina and cervix in girls: analysis of 170 registry cases. Am J Obstet Gynecol
1974;119:713-724.
232. Barnes AB, Colton T, Gundersen J, et al. Fertility and outcome of

pregnancy in women exposed in utero to diethylstilbestrol. N Engl J Med
1980;302:609-613.
233. Berger MJ, Goldstein DP. Impaired reproductive performance in DESexposed women. Obstet Gynecol 1980;55:25-27.
234. Herbst AL, Hubby MM, Blough RR, et al. A comparison of pregnancy
experience in DES-exposed and DES-unexposed daughters. J Reprod Med
1980;24:62-69.
235. Linn S, Liberman E, Schoenbaum SC, et al. Adverse outcome of pregnancy

in women exposed to diethylstilbestrol in utero. J Reprod Med 1988;33:3-7.
P.256
236. Gill WB, Schumacher GFB, Bibbo M, et al. Association of diethylstilbestrol
exposure in utero with cryptorchidism, testicular hypoplasia and semen
abnormalities. J Urol 1979;122:36-39.
237. Shy KK, Stenchever MA, Karp LE, et al. Genital tract examinations and
zona-free hamster egg penetration tests from men exposed in utero to
diethylstilbestrol. Fertil Steril 1984;42:772- 778.
238. Vessey MP. Epidemiological studies of the effects of diethylstilbestrol. IARC

Sci Publ 1989;96:335-348.
239. Gershman ST, Stolley PD. A case-control study of testicular cancer using
Connecticut tumour registry data. 1988;17:738- 742.
240. Pinsky WW, Rayburn WF, Evans MI. Phamacologic therapy for fetal
arythmias. Clin Obstet Gynecol 1991;34:304-309.
241. Hanson JW, Smith DW. The fetal hydantoin syndrome. J Pediatr
1975;87:285-290.
242. Speidel BD, Meadow SR. Maternal epilepsy and abnormalities of the fetus
and newborn. Lancet 1972;2:839-843.
243. Frederick J. Epilepsy and pregnancy: a report from Oxford record linkage
study. Br Med J 1973;2:442-448.
244. Monson RR, Rosenberg L, Hartz SC, et al. Diphenylhydantoin and selected
malformations. N Engl J Med 1973;289:1049.
245. Albengres E, Tillement JP. Phenytoin in pregnancy: a review of the

reported risks. Biol Res Pregnancy Perinatol 1983;4:71-74.
246. Barr M, Pozanski AK, Shmickel RD. Digital hypoplasia and anticonvulsants
during gestation, a teratogenic syndrome. J Pediatr 1974;4:254-256.
247. Hill RM, Verland WM, Horning MG, et al. Infants exposed in utero to
antiepileptic drugs. Am J Dis Child 1974;127:645-653.
248. Hanson JW, Myrianthopoulos NC, Harvey MAS, et al. Risks to the offspring
of women treated with hydantoin anticonvulsants, with emphasis on the fetal
hydantoin syndrome. J Pediatr 1976;89:662-668.
249. Keneko S. Antiepileptic drug therapy and reproductive consequences:

functional and morphological effects. Reprod Toxicol 1991;5:179-198.
250. Shapiro S, Slone D, Hartz SC, et al. Anticonvulsants and parental epilepsy
in the development of birth defects. Lancet 1976;1: 272-275.
251. McClain RM, Langhoff L. Teratogenicity of diphenylhydantoin in the New

Zealand White rabbit. Teratology 1980;21:371-379.
252. Collins MD, Fradkin R, Scott WI. Induction of postaxial forelimb

ectrodactyly with anticonvulsant agents in A/J mice. Teratology 1990;41:61-70.
253. Finnell RH, Abbott LC, Taylor SM. The fetal hydantoin syndrome: answers
from a mouse model. Reprod Toxicol 1989;3: 127-133.
254. Elshave J. Cleft palate in the offspring of female mice treated with
phenytoin. Lancet 1969;2:1074.
255. Harbinson RD, Becker BA. Relation of dosage and time of administration of
diphenylhydantoin to its teratogenic effect in mice. Teratology 1969;2:305-312.
256. Rowland JF, Binkerd PE, Hendrickx AG. Developmental toxicity and
pharmacokinetics of oral and intravenous phenytoin in the rat. Reprod Toxicol
1990;4:191-202.
257. Zengel AE, Keith DA, Tassinari MS. Prenatal exposure to phenyltoin and its
effect on postnatal growth and craniofacial proportion in the rat. J Craniofac
Genet Dev Biol 1989;9:147-160.
258. Finnell RH. Phenytoin-induced teratogenesis: a mouse model. Science

1981;211:483-484.
259. Liggins GC, Howie RN. A controlled trial of antepartum glucocorticoid

treatment for prevention of the respiratory distress syndrome in premature
infants. Pediatrics 1972;50:515-525.
260. Collaborative Group on Antenatal Steroid Therapy. Effect of antenatal

dexamethasone administration in the prevention of respiratory distress
syndrome. Am J Obstet Gynecol 1981;141:276-287.
261. Cziezel AE, Rockenbauer M. Population based case-control study of

teratogenic potential of corticosteroids. Teratology 1997;56: 335-340.
262. Liu D-L, Zhou Z-L. Enhancement of fetal lung maturity by intra-amniotic
instillation of dexamethasone. Clin Med J 1985;98: 915-918.
263. Chrousos GP, Evans MI, Loriaux DL, et al. Prenatal therapy in congenital
adrenal hyperplasia. Attempted prevention of abnormal external genital
masculinization by pharmacologic suppression of the fetal adrenal gland in
utero. Ann N Y Acad Sci 1985;458:156-164.
264. Pang S. Congenital adrenal hypoplasia. Endocrinol Metab Clin North Am

1997;26:853-891.
265. Bar J, Fisch B, Wittenberg C, et al. Prednisone dosage and pregnancy

outcome in renal allograft recipeints. Nephrol Dial Transplant 1997;12:760-763.
266. Kallen B. Drug treatment of rheumatic diseases during pregnancy: the

teratogenicity of antirheumatic drugs-what is the evidence? Scand J Rheumatol
Suppl 107:114-124.
267. Kwak J, Gilman-Sachs A, Beaman K, et al. Reproductive outcome in

women with recurrent spontaneous abortion of alloimmune and autoimmune
causes: preconception versus postconception treatment. Am J Obstet Gynecol
1992;166:1787-1798.
268. Laskin C, Bombardier C, Hannah M, et al. Predisone and aspirin in women

with autoantibodies and unexplained fetal loss. N Engl J Med 1997;337:148-153.
269. Le Thi Huong D, Wechsler B, Vauthier-Brouzes D, et al. Outcome of

planned pregnancies in systemic lupus erythematosis: a prospective study on
62 pregnancies. Br J Rheumatol 1997;36: 772-777.
270. Cowchuck FS, Reece EA, Balaban D, et al. Repeated fetal losses associated
with antiphopholipid antibodies: a collaborative randomized trial comparing
prednisone with low-dose heparin treatment. Am J Obstet Gynecol
1992;166:1318-1323.
271. Reinisch JM, Simon NG. Prenatal exposure to prednisone in humans and
animals retards intrauterine growth. Science 1978;202:436-438.
272. Higgins R, Mendelsohn A, DeFeo M, et al. Antenatal dexamethasone and

decreased severity of retinopathy of prematurity. Arch Opthalmol
1998;116:601-605.
273. Leylek O, Ergur A, Senocak F, et al. Prophylaxis of the occurrence of

hyperbilirubinemia in relation to maternal oxytocin infusion with steroid
treatment. Gynecol Obstet Invest 1998;46:164-168.
274. Spinello A, Capuzzo E, Ometto A, et al. Value of antenatal corticosteroid

therapy in preterm birth. Early Hum Dev 1995;42: 37-47.
275. Silver R, Vyskocil C, Solomon S, et al. Randomized trial of antenatal

dexamethsone in surfactant-treated infants delivered before 30 weeks
gestation. Obstet Gynecol 1996;87:683-691.
276. Ward RA, Pharmacologic enhancement of fetal lung maturation. Clin
Perinatol 1994;21:523-542.
277. Moise KJ. Indomethacin therapy in the treatment of symptomatic

polyhydramnios. Clin Obstet Gynecol 1991;24:310.
278. Bessinger, R, Niebyl J, Keyes W, et al. Randomized comparative trial of

indomethacin and ritrodine for the long-term treatment of preterm labor. Am J
279. Bivens H, Newman R, Fyfe D, et al. Randomized trial of oral indomethacin

and terbutaline sulfate for the long-term supression of preterm labor. Am J
Obstet Gynecol 1993;169:1065-1070.
280. Mohen D, Newnham J, D'Orsonga L. Indomethacin for the treatment of

polyhydramnios: a case of constriction of the ductus arteriosus. Aust N Z J
281. Eronen M. The hemodynamic effects of antenatal indomethacin and a betasympathomimetic agent of the fetus and the newborn: a randomized study.
Pediatr Res 1993;33:615-619.
282. Moise KJ. Effect of advancing gestational age on the frequency of fetal
ductal constriction in association with maternal indomethacin use. Am J Obstet
Gynecol 1993;168:1350-1353.
283. Norton M. Teratogen update: fetal effects of indomethacin administration

during pregnancy. Teratology 1997;56: 282-292.
284. Fejgin MD, Delpino ML, Bidiwala KS. Isolated small bowel perforation
following intrauterine treatment with indomethacin administration. Am J
Perinatol 1994;11:295-296.
285. Major CA, Lewis DF, Harding JA, et al. Tocolysis with indomethacin
increases in incidence of necrotizing enerocolitis in the low-birth-weight
neonate. Am J Obstet Gynecol 1994;170: 102-106.
286. Lewis WH, Suris OR. Treatment with lithium carbonate: results in 35
cases. Tex Med 1970;66:58-63.
287. Vacaflor L, Lehmann HE, Ban TA. Side effects and teratogenicity of lithium
carbonate treatment. J Clin Pharmacol 1970;10: 387-389.
288. Frankenberg RR, Lipinski JF. Congenital malformations. N Engl J Med

1983;309:311-312.
289. Warkany J. Teratogen Update: lithium. Teratology 1988;38: 593-596.
290. Schardein JL, ed. Chemically induced birth defects. New York: Marcel
Dekker, 1993.
P.257
291. Rane A, Tomson G, Bjarke B. Effects of maternal lithium therapy in a
newborn infant. J Pediatr 1974;93:296-297.
292. Jacobson SJ, Jones K, Johnson K, et al. Prospective multicentre study of

pregnancy outcome after lithium exposure during first trimester. Lancet
1992;339:530-533.
293. Cohen LS, Friedman JM, Jefferson JW, et al. A reevaluation of risk of in
utero exposure to lithium. JAMA 1994;271:146-150.
294. Troyer WA, Pereira G, Lannon RA, et al. Association of maternal lithium
exposure and premature delivery. J Perinatol 1993;13: 123-127.
295. Schou M. What happened later to the lithium babies? A follow-up study of
children born without malformations. Acta Psychiatr Scand 1976;54:193-197.
296. Holtzman EJ, Ausiello DA. Nephrogenic diabetes insipidus: causes

revealed. Hosp Pract 1994;29:89-93,97-98,103-104.
297. Krause S, Ebbsen F, Lange AP. Polyhydramnios with maternal lithium

treatment. Obstet Gynecol 1990;75:504-506.
298. Lam SS, Kjellstrand C. Emergency treatment of lithium-induced diabetes
insipidus with non-steroidal anti-inflammatory drugs. Renal Failure

1997;19:183-188.
299. Nars PW, Girad J. Lithium carbonate intake during pregnancy leading to a
large goiter in a premature infant. Am J Dis Child 1977;131:123-127.
300. Filtenborg JA. Persistent pulmonary hypertension after lithium intoxication

in the newborn. Eur J Pediatr 1982;138:321-323.
301. Morrell P, Sutherland GR, Buamah PK, et al. Lithium toxicity in the
neonate. Arch Dis Child 1983;58:539-541.
302. Wilson N, Forfar JD, Godman MJ. Atrial flutter in the newborn resulting
from lithium ingestion. Arch Dis Child 1983;58: 538-539.
303. Weinstein MR, Goldfield M. Cardiovascular malformations with lithium use

during pregnancy. Am J Psychiatry 1975;132:529- 531.
304. Hansen DK, Walker RC, Grafton TF. Effect of lithium carbonate on mouse
and rat embryos in vitro. Teratology 1990;41:155-160.
305. Klug S, Collins M, Nagao T, et al. Effect of lithium on rat embryos in

culture: growth, development, compartmental distribution and lack of protective
effect of inositol. Arch Toxicol 1992;66:719-728.
306. Jurand A. Teratogenic activity of lithium carbonate: an experimental

update. Teratology 1988;38:101-111.
307. Llewellyn A, Stowe ZN, Strader JR. The use of lithium and management of
women with bipolar disorder during pregnancy and lactation. J Clin Psychiatry
1998;59:57-64.
308. Cowett RM, Hakanson DO, Kocon RW, et al. Untoward neonatal effect of
intraamniotic administration of methylene blue. Obstet Gynecol 1976;48:745755.
309. Serota FT, Bernbaum JC, Schwartz E. The methylene blue baby. Lancet
1979;2:1142-1143.
310. Crooks J. Haemolytic jaundice in a neonate after intra-amniotic injection of

methylene blue. Arch Dis Child 1982;57:872-886.
311. Dolk H. Methylene blue and atresia or stenosis of ileum and jejunum.
Lancet 1991;338:1021-1022.
312. Cragan JD. Teratogen Update: methylene blue. Teratology 1999; 60:42-48.
313. Cragen JD, Martin L, Waters et al. Increased risk of small intestinal atresia
among twins in the United States. Arch Pediatr Adolesc Med 1994;148:733-739.
314. van der Pol JG, Wolf H, Boer K, et al. Jejunal atresia related to the use of
methylene blue in genetic amniocentesis in twins. Br J Obstet Gynecol
1992;99:141-143.
315. Moorman-Voestermans CGM, Heig HA, Vos A. Jejunal atresia in twins. J

Pediatr Surg 1990;25:638-639.
316. Moorman-Voestermans CGM, Heij HA, Vos A. Letter to the Editor. J Pediatr
Surg 1992;27:133.
317. Nicolini U, Monni G. Intestinal obstruction in babies exposed in utero to

methylene blue. Lancet 1990;336:1258-1259.
318. Cragen JD, Martin L, Khoury MJ, et al. Dye use during amniocentesis and
birth defects [Letter]. Lancet 1993;341:1352-1353.
319. Gluer S. Intestinal atresia following intraamniotic use of dyes. Eur J Pediatr
Surg 1995;5:240-242.
320. Kidd SA, Lancaster PA, Anderson JC, et al. Fetal death after exposure to
methylene blue dye during mid-trimester amniocentesis. Prenat Diagn
1996;16:39-47.
321. Kidd SA, Lancaster PA, Anderson JC, et al. A cohort study of pregnancy
outcome after amniocentesis in twin pregnancy. Paediatr Perinat Epidemiol

1997;11:200-213.
322. Piersma AH, Verhoet A, DeLiefde A, et al. Embryotoxicity of methylene

blue in the rat. Teratology 1991;43:458-459.
323. Gonzalez CH, Vargas FR, Perez ABA, et al. Limb deficiency with or without
Moebius sequence in seven Brazilian children associated with misoprotol use in
the first trimester of pregnancy. Am J Med Genet 1993;46:59-64.
324. Coelho HLL, Misago C, Fonsecam WVC, et al. Selling abortifacients over
the counter in pharmacies in Fortaleza, Brazil. Lancet 1991;338:247.
325. Costa SH, Vessey MP. Misoprostol and illegal abortion in Rio de Janeiro,
Brazil. Lancet 1993;341:1258-1261.
326. Luna-Coelho HL, Teixeria AC, Santos AP, et al. Misoprostol and illegal
abortion in Fortaleza, Brazil. Lancet 1993;341:1261-1263.
327. Brent RL. Relationship between uterine vascular clamping, vascular

disruption, and cocaine teratogenicity [Editorial]. Teratology 1990;41:757-760.
328. NICHD Workshop. CVS and limb reduction defects. Teratology 1993;48:713.
329. Collins FS, Mahoney MJ. Hydrocephalus and abnormal digits after failed
first trimester prostaglandin abortion attempt. J Pediatr 1983;102:620-621.
330. Schuler LS, Ashto PW, Sanseverino MT. Teratogenicity of misoprostol.

Lancet 1992;339:437.
331. Woods JR, Plessinger MA, Clark KE. Effect on cocaine on uterine blood flow
and fetal oxygenation. JAMA 1987;257:957- 961.
332. Schonhofer PS. Brazil: misuse of misoprostol as a abortifacient may induce

malformations. Lancet 1991;337:1534.
333. Fonseca W, Alencar AJC, Mota FSB, et al. Misoprostol and congenital
malformations. Lancet 1991;336:56.
334. Castilla EE, Orioli IM. Teratogenicity of misoprostol: data from the LatinAmerican collaborative study of congenital malformations (ECLAMC). Am J Med
Genet 1994;51:161-162.
335. Zackai EH, Melmen WJ, Neiderer B, et al. The fetal trimethadione
syndrome. J Pediatr 1975;87:280-284.
336. Cohen MM. Syndromologyan updated conceptual overview. VII. Aspects

of teratogenesis. Int J Oral Maxillofac Surg 1990;19:26-32.
337. Feldman GL, Weaver DD, Lovrien EW. The fetal trimethadione syndrome.
Am J Dis Child 1977;131:1389-1392.
338. Smith ES, Dafoe CS, Miller JR, et al. An epidemiological study of congenital
reduction deformities of the limbs. Br J Prev Soc Med 1977;31:39-41.
339. Keen CL, Mark-Savage P, Lonnerdal B, et al. Teratogenesis and low copper
status resulting from D-penicillamine in rats. Teratology 1982;26:163-165.
340. Harpey J-P, Jaudon M-C, Clavel J-P, et al. Cutix laxa and low serum zinc
after antenatal exposure to penicillamine. Lancet 1983;2:858.
341. Linares A, Zarranz JJ, Rodriguez-Alarcon J, et al. Reversible cutix laxa due
to maternal D-penicillamine treatment. Lancet 1979; 2:43.
342. Solomon L, Abrams G, Dinner M, et al. Neonatal abnormalities associated

with D-pencillamine treatment during pregnancy. N Engl J Med 1977;296:54-55.
343. Greenberg G, Inman W, Weatherall J, et al. Maternal drug histories and

congenital abnormalities. Br Med J 1977;2:853-856.
344. Nakane Y, Okuma T, Takahashi R, et al. Multi-institutional study on the

teratogenicity and fetal toxicity of antieplieptic drugs: a report of a collaborative
study group in Japan. Epilepsia 1980; 21:663.
345. Robert E, Lofkvist E, Mauguiere F, et al. Evaluation of drug therapy and

teratogenic risk in a Rhone-Alps district population of pregnant epileptic women.
Eur Neurol 1986;25:436-443.
346. Dansky L, Finnell R. Parental epilepsy, anticonvulsant drugs, and

reproductive outcome: epidemiologic and experimental findings spanning three
decades; 2, Human studies. Reprod Toxicol 1991; 5:301-335.
347. Waters C, Belai Y, Gott P, et al. Outcomes of pregnancy associated with

antiepileptic drugs. Arch Neurol 1994;51:250-253.
348. Kelly T. Teratogenicity of anticonvulsant drugs I: Review of the literature.

Am J Med Genet 1984;19:413-434.
349. Dravet C, Julian C, Legras C, et al. Epilepsy, antiepileptic drugs, and

malformations in children of women with epilepsy: a French prospective cohort
study. Neurology 1992;42:75-82.
350. Lindhout D, Meinardi H, Meijer JWA, et al. Antiepileptic drugs and

teratogenesis in two consecutive cohorts: changes in prescription policy
paralleled by changes in pattern of malformations. Neurology 1992;42:94-110.
P.258
351. Tanganelli P, Regesta G. Epilepsy, pregnancy, and major birth anomalies:
an Italian prospective controlled study. Neurology 1992;42:89-93.
352. Seip M. Growth retardation, dysmorphic facies and minor malformations

following massive exposure to phenobarbitone in utero. Acta Paediatr Scand
1976;65:617-621.
353. Jones K, Johnson K, Chamber C. Pregnancy outcome in women treated

with phenobarbitol monotherapy. Teratology 1992;45: 452-453.
354. Koch S, Losche G, Jager-Roman E, et al. Major and minor birth

malformations and antiepileptic drugs. Neurology 1992;42: 83-88.
355. Morales WJ. Antenatal therapy to minimize neonatal intraventricular

hemorrhage. Clin Obstet Gynecol 1991;34:328-335.
356. Morales WJ. Effect of intraventricular hemorrhage on the one-year mental

and neurologic handicaps of the very low birth weight infant. Obstet Gynecol
1987;70:111-114.
357. Fotherby K. A new look at progestins. Clin Obstet Gynecol 1984; 11:701722.
358. Wilkins L. Masculinization due to orally given progestins. JAMA

1960;172:1028-1032.
359. World Health Organization. The effect of female sex hormones on fetal
development and infant health. Geneva: World Health Organization, 1981.
360. Heinonen OP, Slone D, Monson RR, et al. Cardiovascular birth defects and
antenatal exposure to female sex hormones. N Engl J Med 1977;296:67-70.
361. Wiseman RA, Dodds-Smith IC. Cardiovascular birth defects and antenatal
exposure to female sex hormones: a reevaluation of some base data.
Teratology 1984;30:359-370.
362. Gal I. Risks and benefits of the use of hormonal pregnancy test tablets.
Nature 1972;240:241-242.
363. Janerich DT, Piper JM, Glebatis DM. Oral contraceptives and congenital
limb reduction defects. N Engl J Med 1974;291:697-700.
364. Brent RL. The magnitude of the problem of congenital malformations. In:
Marois M, ed. Prevention of physical and mental congenital defect part a basic
and medical Science, education and future strategies. New York: Alan R. Liss,
1985:55-68.
365. O'Malley BW, Schrader DT. The receptors of steroid hormones. Sci Am
1976;234:32-43.
366. Lammer EJ. Developmental toxicity of synthetic retinoids in humans. Prog

Clin Biol Res 1988;281:193-202.
367. Rosa FW. Teratogen Update: penicillamine. Teratology 1986;33: 127-131.
368. Dai WS, Hsu M, Itri LM. Safety of pregnancy after discontinuation of
isotretinoin. Arch Dermatol 1989;125:362-365.
369. Dai WS, LaBraico JM, Stern RS. Epidemiology of isotretinoin exposure
during pregnancy. J Am Acad Dermatol 1992;26: 599-606.
370. DiGiovanna JJ, Zech LA, Ruddel ME, et al. Etretinate: persistent serum
levels of a potent teratogen. Clin Res 1984;32:579A.
371. Rinck G, Gollnick H, Organos CE. Duration of contraception after

etretinate. Lancet 1989;1:845-846.
372. Alles AJ, Sulik KK. Retinoic-acid-induced limb-reduction defects:

perturbation of zones of programmed cell death as a pathogenetic mechanism.
Teratology 1989;40:163-171.
373. Yasuda Y, Konishi H, Kihara T, et al. Developmental anomalies induced by

all-trans-retinoic acid in fetal mice: II. Induction of abnormal neuroepithelium.
Teratology 1987;35:355-366.
374. Ong DE, Chytil F. Changes in levels of cellular retinol-and retinoic-acidbinding proteins of liver and lung during perinatal development rat. Proc Natl
Acad Sci U S A 1976;73:3976- 3978.
375. Jick SS, Terris BZ, Jick H. First trimester topical tretinoin and congenital
disorders. Lancet 1993;341:1181-1182.
376. DeWals P, Bloch D, Calabro A, et al. Association between

holoprosencephaly and exposure to topical retinoids: results of the EUROCAT
survey. Paediatr Perinat Epidemiol 1991;5:445-447.
377. Bowman JM. Antenatal suppression of Rh alloimmunization. Clin Obstet
Gynecol 1991;34:296-303.
378. Prager K, Malin H, Speigler D, et al. Smoking and drinking behavior before
and during pregnancy of married mothers of liveborn and stillborn infants.
Public Health Rep 1984;99:117-127.
379. Naeye RL. Effects of maternal cigarette smoking on the fetus and placenta.
Br J Obstet Gynaecol 1979;85:732-737.
380. Chattingius S. Does age potentiate the smoking-related risk of fetal growth
retardation? Early Hum Dev 1989;20:203-211.
381. Hjortdal JO, Hjortdal VE, Foldspang A. Tobacco smoking and fetal growth:
a review. Scand J Soc Med 1989;45:1-22.
382. Stillman RJ, Rosenberg MJ, Sachs BP. Smoking and reproduction. Fertil
Steril 1986;46:545-566.
383. Erickson JD. Risk factors for birth defects: data from the Atlanta defects
case-control study. Teratology 1991;43:41-51.
384. Tikkanen J, Heinonen OP. Maternal exposure to chemical and physical

factors during pregnancy and cardiovascular malformations in the offspring.
Teratology 1991;43:591-600.
385. Werler MM, Pober BR, Holmes LB. Smoking and pregnancy. Teratology
1985;32:473-481.
386. Lenz W, Knapp K. Thalidomide embryopathy. Arch Environ Health

1962;5:100-105.
387. Brent RL, Holmes LB. Clinical and basic science lessons from the
thalidomide tragedy: what have we learned about the causes of limb defects?
Teratology 1988;38:241-251.
388. Lenz W. A short history of thalidomide embryopathy. Teratology

1988;38:203-215.
389. Kida M. Thalidomide embryopathy in Japan. Kodansha, Tokyo: 1987.
390. Ruffing L. Evaluation of thalidomide children. Birth Defects Orig Artic Ser
1977;13:287-300.
391. Cutler J. Thalidomide revisted. Lancet 1994;343:795-796.
392. Jones GRN. Thalidomide: 35 years on and still deforming. Lancet

1994;343:1041.
393. Stephens TD. Proposed mechanisms of action in thalidomide embryopathy.

Teratology 1988;38:229-239.
394. Carswell F, Kerr MM, Hutchinson JH. Congenital goiter and hypothyroidism
produced by maternal ingestion of iodides. Lancet 1970;1:1241-1243.
395. Rodesch F, Camus M, Ermans AM, et al. Adverse effect of amniofetography

on fetal thyroid function. Am J Obstet Gynecol 1976;126:723-726.
396. Clewell WP. In utero treatment of thyrotoxicosis. In: Evans MI, et al, eds.
Fetal diagnosis and therapy: science, ethics, and the law. Philadelphia: JB
Lippincott, 1984:124.
397. Smith CV. Reversing acute intrapartum fetal distress using tocolytic drugs.
Clin Obstet Gynecol 1991;34:353-359.
398. Caritis SN, Darby MJ, Chan L. Pharmacologic treatment of preterm labor.
Clin Obstet Gynecol 1988;31:635-651.
399. Egarter CH, Husslein PW, Rayburn WF. Uterine hyperstimulation after lowdose prostagladin E2 therapy: tocolytic treatment in 181 cases. Am J Obstet
Gynecol 1990;163:794-796.
400. Mendez-Bauer C, Shekarloo A, Cook V, et al. Treatment of acute

intrapartum fetal distress by beta 2-symphatomimetics. Am J Obstet Gynecol
1987;148:104-106.
401. Reece EA, Chervenak FA, Romero R, et al. Magnesium sulfate in the
management of acute intrapartum fetal distress. Am J Obstet Gynecol
1984;148:104-106.
402. Toutant C, Lippman S. Fetal solvents syndrome. Lancet 1979;1:1356.
403. Pearson MA, Hoyme HE, Seaver LH, et al. Toluene embryopathy:
delineation of the phenotype and comparison with fetal alcohol syndrome.
Pediatrics 1994;93:211-215.
404. Arnold GL, Kirby RS, Langendoerfer S, et al. Toluene embryopathy: clinical
delineation and developmental follow-up. Pediatrics 1994;93:216-220.
405. Safra MJ, Oakley GP. Valium: an oral cleft teratogen? Cleft Palate J
1976;13:198-200.
406. Dalens B, Raynaud E-J, Gaulme J. Teratogenicity of valproic acid. J Pediatr

1980;97:332-333.
407. Robert E. Valproic acid as a human teratogen. Congenit Anom Kyoto

1988;28:S71-S80.
408. Omtzigt JGC, Nau H, Los FJ, et al. The disposition of valproate and its
metabolites in the late first trimester and early second trimester of pregnancy in
maternal serum, urine and amniotic fluid: effect of dose, co-medication, and the
presence of spina bifida. Eur J Clin Pharmacol 1992;43:381-388.
409. Lammer EJ, Sever LE, Oakley GP. Valproic acid. Teratology 1987;35:465473.
410. Dickinson RG, Hapland RC, Lynn RK, et al. Transmission of valproic acid
across the placenta: half-lives of the drug in mother and baby. J Pediatr
1979;94:832-835.
411. Teratology Society. Teratology Society position paper: recommendation for

vitamin A use during pregnancy. Teratology 1987;35:269-275.
412. Zuber C, Librizzi RJ, Vogt BI. Outcomes of pregnancies exposed to high
doses of vitamin A. Teratology 1987;35:42A.
P.259
413. Ampola MG, Mahoney MJ, Nakamura E, et al. Prenatal therapy of a patient
with vitamin B responsive methylmalonic acidemia. N Engl J Med 1975;293:313317.
414. Goodenday L, Gordon G. No risk from vitamin D in pregnancy. Ann Intern

Med 1971;75:807-808.
415. O'Brien J, Rosenwasser S, Feingold M, et al. Prenatal exposure to milk with

excessive vitamin D supplementation. Teratology 1993;47:387.
416. Friedman WF. Vitamin D and the supravalvular aortic stenosis syndrome.
In: Woollam DHM, ed. Advances in teratology. New York: Academic Press,
1968:83-96.
417. Garcia RE, Friedman WF, Kaback MM, et al. Idiopathic hypercalcemia and
supravalvular stenosis: documentation of a new syndrome. N Engl J Med
1964;271:117-120.
418. Ewart A, Morris C, Atkinson D. Hemizygosity at the elastin locus in a

developmental disorder, Williams syndrome. Nature Genet 1993;5:11-16.
419. Mari A, Amati F, Mingarelli R. Analysis of the elastin gene in 60 patients

with clinical diagnosis of Williams syndrome. Hum Genet 1995;96:444-448.
Chapter 16
Obstetric Anesthesia and Analgesia: Effects on the
Fetus and Newborn
Judith Littleford
Many drugs and various techniques have been used to provide anesthesia and analgesia for surgery during
pregnancy, for labor and delivery, and for breastfeeding. The following quote, which refers to the first
administration of inhalational analgesia in childbirth, is as relevant to the practice of obstetric anesthesia today
as it was in 1847, It will be necessary to ascertain anesthesia's precise effect, both upon the action of the
uterus and on the assistant abdominal muscles; its influence, if any, upon the child; whether it has the tendency
to hemorrhage or other complications (1). Between the mid-1800s and 1950s, descriptive reports of the
presumed effect of maternally administered medication on the fetus and newborn appeared sporadically in the
literature. Two developments eventually encouraged physicians to acknowledge the potential problems
associated with placental transmission of anesthetic drugs:
Recognition that morphine, a popular ingredient of patent medicines, was addictive, and that signs of
withdrawal could be identified in the fetus (violent fetal movements and/or sudden fetal death) when the
mother's heavy opioid use was decreased.
Confirmation of the structure and dynamic function of the placenta and demonstration of the presence of
chloroform in the umbilical blood of neonates.
In 1952 the pioneering work of anesthesiologist Virginia Apgar converted an intangible phenomenon, the clinical
condition of a newly born baby, into a formally defined measurement (2). Thereafter, the well-being of the infant
became a major criterion for evaluation of the obstetric and anesthetic management of pregnant women.
This chapter introduces the neonatal practitioner to the clinical aspects of obstetric anesthesia and analgesia and
examines their effects on the fetus and newborn.
EVALUATION OF WELL-BEING
Several methods of evaluation have been adopted into common usage as anesthesiologists attempt to separate
out the fetal/neonatal effects of their interventions from concomitant medical and nursing management, and
from the influence of preexisting maternal conditions.
The Apgar Score

The Apgar score is a convenient method of reporting the status of the newborn and the effectiveness of
resuscitation. It rates each of five physical signs traditionally used by anesthesiologists to monitor a patient's
condition: heart rate, respiratory effort, muscle tone, reflex irritability, and color. Apgar demonstrated that her
score was sufficiently sensitive to detect differences among newborns whose mothers had received spinal versus
general anesthesia (GA) for cesarean section (2).
Elements of the score are partly dependent on the physiologic maturity of the infant. Likewise, neonatal
conditions such as bradyarrhythmias affect heart rate, while infection, neuromuscular conditions, and certain
medications affect respiratory effort and tone. Additional information is required to interpret Apgar scores
meaningfully in infants receiving ongoing resuscitation.
Wider application has resulted in incorrect use of the Apgar score to evaluate infant morbidity/mortality and to
predict the occurrence of intrapartum events and long-term disability. Apgar scores at 1 and 5 minutes correlate
poorly with both cause and neurologic outcome. A joint statement from the American Academy of Pediatrics
(AAP) and American College of Obstetricians and Gynecologists emphasizes the appropriate use of the Apgar
Score (3).
P.261
When used as originally intended, the Apgar score remains a valuable tool to assess the condition of the infant
at birth (4), but it is not specific for the effects of anesthesia on the newborn.
Umbilical Cord Blood Gas Analysis

Cord blood gas analysis is the gold standard for assessing fetal acid-base status and uteroplacental function at
birth. Umbilical artery pH, base excess, and pCO2 reflect fetal and immediate neonatal condition whereas
umbilical vein values reflect maternal acid-base status and placental function.
Normal values vary depending on the definition of normality and the influence of factors (e.g., altitude, parity,
breech vaginal delivery, and duration of labor) on the population studied (5). Helwig and associates (6)
retrospectively examined the records of 15,000 vigorous newborns with a 5-minute Apgar score of >7. Median
umbilical artery values, with the 2.5th percentile value in parentheses, were pH 7.26 (7.10) and base excess -4
mmol/L (-11 mmol/L). The means 2 standard deviations were similar. The generally accepted lower limit of
normal umbilical artery pH extends to 7.10 and base excess to -12 mmol/L (5,7).
Values for pH, pCO2, and base excess also vary with differences in sampling technique. Preanalytical error can
be introduced if the cord is not clamped immediately, there is an excess quantity of heparin in relation to the
amount of blood collected, air is present in the syringe, or the sample is kept at room temperature for longer
than 15 minutes (8).
The supply of oxygen and the removal of volatile (CO2) and fixed (e.g., lactate) acids by the placenta for
excretion by the maternal lungs and kidneys, respectively, allow the fetus to maintain acid-base balance within a
narrow range. Interruption of these processes can lead to acidemia in the fetus. In general, respiratory acidosis
alone is not associated with newborn complications; rather, it reflects a sudden decrease in uteroplacental or
umbilical perfusion such as placental abruption or cord prolapse immediately preceding delivery. Base excess
values have greater usefulness than pH values because base excess does not change significantly with
respiratory acidosis and demonstrates linear, rather than logarithmic, correlation to the degree of metabolic
acidosis. Umbilic artery base excess is the most direct measure of fetal metabolic acidosis. Human and animal
studies have confirmed normal values before and during labor and rates of base excess change in relation to
events causing fetal hypoxemia (maternal hypoxemia, umbilical cord occlusion, and reduced uterine blood flow)
(7).
The process of normal labor and delivery without anesthetic intervention stresses the fetus such that mild
acidosis develops in almost all labors (9). Reynolds and associates (10) recently completed a metanalysis
comparing epidural with systemic opioid analgesia to determine the effect of these anesthetic interventions on
acid-base status at birth. They concluded that epidural analgesia was associated with an improvement in base
excess, suggesting that placental exchange is well preserved in association with this technique.
While umbilical artery pH, base excess, and pCO2 are considered sensitive and objective indicators of fetal
hypoxia during labor, the results represent a snapshot of fetal status and depict the mixed effluent of all fetal
tissues. Cord gases do not distinguish between primary fetal pathologic conditions, fetal effects of maternal
conditions (e.g., acid-base disorders), or the influence of inadequate placental blood flow. They also do not
indicate in which direction the condition of the fetus is moving, or at what rate, nor do they reflect events that
occurred remote from delivery.
Fetal Scalp Blood Gas Analysis

This method is advocated to improve the specificity of fetal heart rate monitoring (11) but has not been used by
anesthesiologists to measure the fetal effects of their interventions.
Evaluation of Newborn Neurobehavior

Brazelton, Scanlon, and Amiel-Tison assessed the effect of anesthetic medications on the neurobehavior of term,
healthy newborns (12,13,14). Their belief was that central nervous system depression from drugs administered
to the mother during labor and delivery could be distinguished from effects associated with perinatal asphyxia
and trauma at birth. Several review articles have described these tests in detail and compared the Brazelton
Neonatal Behavioral Assessment Scale (NBAS) (12) and Early Neonatal Neurobehavioral Scale (ENNS) (13) to
the Neurologic and Adaptive Capacity Score (NACS) (14).

Researchers in obstetric anesthesia have tended to favor the NACS. It emphasizes muscle tone, avoids aversive
stimuli, can be completed quickly, and is considered easier to learn than the NBAS and ENNS. The 20-item
instrument is organized into two scales, adaptive capacity and neurologic assessment. The latter is further
divided into four subscales, passive tone, active tone, primary reflexes, and general neurologic status
assessment. A systematic review of the literature regarding use of NACS in obstetric anesthesia research
concluded that reliability and validity evaluations of the tool were lacking, and it was unclear whether NACS
could detect the existence of subtle neurobehavioral effects (15). The topic was the subject of an editorial that
examined the widespread acceptance of this test notwithstanding the criticism that surrounded its initial
publication (16). Subsequently, it has been determined that NACS has poor reliability when used to detect the
effects of intrapartum drugs and other interventions on the neonate (17).
The NBAS is a much more detailed examination that may be more sensitive to environmental influences on
behavior and reflect more accurately the capabilities of the neonate (18). Early studies involving concentrated
epidural solutions tended to show that babies delivered to medicated mothers were less alert, had poorer muscle
tone, and had difficulty habituating to repeated stimuli. This finding
P.262
is much less likely to occur with modern obstetric analgesia (19), where emphasis is placed on administering the
lowest effective dose of medication or combination of medications needed to achieve the desired maternal effect
and reduce the likelihood of adverse influences on the fetus/newborn.
Electronic Fetal Monitoring

Electronic fetal monitoring, one technique in the overall strategy of intrapartum fetal surveillance (20,21), aims
to improve outcomes by identifying fetuses with hypoxic acidemia at a point when the process is still completely
reversible by intrauterine resuscitation or expedited delivery. The fetal heart rate (FHR), including variability,
accelerations, and decelerations, if any occur, is recorded electronically on a paper trace. A reactive (normal or
reassuring) cardiotocogram (CTG) is defined by the presence of accelerations. Baseline FHR normally ranges
between 110 and 160 beats per minute. Reduced variability and the presence of decelerations are abnormal
findings.
A systematic review of studies comparing the efficacy and safety of routine continuous electronic fetal
monitoring versus intermittent auscultation (IA) of the FHR reveals the former to be associated with a
significantly higher rate of cesarean and operative vaginal delivery (22). In the antecedent Cochrane review of
this same topic, the incidence of GA to facilitate obstetric intervention was also increased. This aspect of
management was not examined in the current review. The only significant benefit of the use of routine,
continuous fetal monitoring was in the reduction of neonatal seizures; otherwise, neither IA nor continuous
monitoring has been shown to decrease morbidity or mortality (22).
Difficulty in the visual interpretation of CTG patterns during labor can result in unnecessary operative
intervention, while some significant changes go unrecognized. Computerized CTG systems, which do not rely on
observer reading of the FHR tracing, are more accurate and reliable. Evaluation of the FHR pattern is given
online continuously, and warnings are displayed if there is signal loss, reduction in fetal movements, or an
abnormally flat or decelerative trace (23). Retrospective analysis of several thousand records enabled
investigators to conclude that the most reliable single parameter of fetal condition was variability (short- and
long-term). Absence of accelerations, presence of decelerations, decrease in the number of movements, and
changes in baseline FHR all occurred occasionally in normal fetuses (24).
Recently, automatic ST segment and T-wave analysis of the fetal electrocardiogram (recorded continuously from
the fetal scalp electrode used for FHR measurement) has been combined with conventional CTG. Elevation or
depression of the ST segment occurs when the fetus is exposed to a hypoxic stress. When CTG is accompanied
by computerized analysis of the ST segment, the number of operative deliveries for the indication of fetal
distress can be reduced without jeopardizing fetal outcome (24). This technique of fetal surveillance improves
the practitioner's ability to identify cases of significant fetal hypoxia necessitating obstetric intervention, thus
reducing the risk of an umbilical cord arterial metabolic acidosis at birth (25).
Anesthesiologists have used continuous fetal CTG recordings to evaluate the effect of maternal analgesia on FHR
and variability. Solt and associates (26) demonstrated that intrapartum intravenous (IV) administration of
meperidine (pethidine) 50 mg and promethazine 25 mg resulted in decreased variability and fewer accelerations
on computerized CTG for the duration of their 40-minute recording. It is unclear whether the effect can be
attributed to one of the two drugs or the combination, although this is a typical fetal response to systemic
maternal opioid administration.
A randomized study was conducted to examine the effect of continuous epidural anesthesia with or without
narcotic on intrapartum FHR characteristics as measured by computer analysis. The narcotic epidural consisted
of an initial bolus of 10-12 mL 0.125% bupivacaine with fentanyl 50 g, followed by continuous bupivacaine
0.125% with fentanyl 1.7g/mL infusion. The nonnarcotic epidural was initiated with a bolus of 10-12 mL
bupivacaine 0.25% and followed by a bupivacaine 0.125% infusion. Investigators found no difference in preand postepidural baseline FHR, accelerations, or variability between the groups (27). These solutions are
considerably less concentrated than those used 10-20 years ago and are in keeping with modern obstetric
anesthesia practice, which aims to reduce motor block by using dilute local anesthetic (LA) plus narcotic epidural
solution combinations.
In a double-blind randomized study of bolus epidural opioid effect on FHR variability, butorphanol 2 mg, fentanyl
50 g, sufentanil 15 g, or saline in combination with bupivacaine 0.25% did not change FHR short- or longterm variability (28).
FHR decreases in response to compression of the fetal head during passage through the birth canal and in
response to umbilical cord compression or reduced uterine blood flow secondary to maternal hypotension or
prolonged uterine contraction. This bradycardia is vagally mediated.
Biophysical Profile Score

This ultrasound-based method combines measures of acute biophysical variables, fetal breathing, heart rate
accelerations, gross body movements, and fetal tone with amniotic fluid volume (29). The first four variables
reflect acute fetal condition, whereas the last variable reflects chronic fetal condition. The observation period
lasts 30 minutes because fetuses are known to sleep for intervals lasting approximately 30 minutes. When
normal (8 out of a possible 10), the biophysical profile score (BPS) is a direct, reliable, and accurate measure of
normal tissue oxygenation. A normal score is never associated with an abnormal fetal pH. Scores 6/10 during
routine antepartum evaluation indicate the amount of oxygen
P.263
delivered to target organs is insufficient to maintain function. The lower the score, the greater the likelihood that
central acidemia is present (30).
Placental transfer to the fetus of maternally administered intramuscular (i.m.) narcotic medication (diamorphine
10 mg or morphine 10 or 15 mg administered with dimenhydrinate) resulted in transiently reduced fetal activity
and, consequently, a lower BPS for the duration of the drug effect (31,32). A small dose of IV fentanyl (50 g)
given in early labor was associated with abolishment of fetal breathing at 10 minutes postdosing, fewer body
movements, and reduced variability (33). The effect lasted approximately 30 minutes, in keeping with the
pharmacodynamic profile of fentanyl. This should be taken into account when fentanyl is administered close to
the time of delivery. In this study, none of the neonates required resuscitation, and all had umbilical artery pH
values >7.2.
Fetal Pulse Oximetry

Fetal oxygen saturation monitoring by pulse oximetry (SpO2) is poised to become the most significant advance
for intrapartum fetal assessment in recent years. It relies on established, noninvasive technology and is based
on the following assumptions: fetal well-being is dependent on perfusion of vital organs with oxygenated blood;
SpO2 correlates with saturation as measured by blood gas analysis; and a critical threshold for SpO2 exists,
above which the fetus will be nonacidemic and below which, acid-base decompensation may occur (34).
Based on animal and human data, fetal SpO2 values 30% can be considered reassuring, whereas SpO2 values
<30% may be associated with acidosis. Low SpO2 values warrant further assessment if they persist for periods
of >2 minutes and necessitate intervention (intrauterine resuscitation or expedited delivery) if they persist for
>10 minutes (34,35). The predictive value of intrapartum fetal SpO2 compares favorably with fetal scalp blood
analysis (36). The former has the advantage of being a continuous monitoring technique.
The United States Food and Drug Administration approved the Nellcor N-400/FS14 fetal pulse oximeter for
clinical use in 2000 as an adjunct to electronic fetal monitoring in the presence of a nonreassuring tracing (Fig.
16-1). At present, this technology is limited to singleton, term fetuses in the vertex position.
Results of studies designed to investigate the effects on fetal oxygen saturation of administering epidurals to
healthy parturients are beginning to appear in the literature. Neither an initial epidural bolus of 15 mL of
ropivacaine 0.1% with sufentanil 10 g or intermittent repeat plain ropivacaine boluses affected SpO2 in healthy
fetuses (37). In a study designed to account for the possible influence of maternal position, diastolic blood
pressure, and preexisting FHR pattern, SpO2 values were not affected by boluses of dilute epidural infusion
solutions but did decrease with bolus administration of more concentrated LA at epidural insertion or top-up
(38). Other than to comment on the variety of combinations and strengths of analgesic agents, details of the
epidural solutions were not contained in the report. Further clinical trials are needed to assess and characterize
the effect of all types of anesthetic intervention on fetal SpO2 during normal and abnormal labor, in addition to
various maternal conditions associated with fetal compromise (e.g., pregnancy-induced hypertension).
Figure 16-1 Oxifirst Fetal Pulse Oximetry. Reprinted with permission from TycoHealthcare. Copyright 2000
Mallinckrodt, Inc. All rights reserved.
Fetal Doppler
Flow velocity waveforms from maternal vessels (uterine arteries), placental circulation (umbilical arteries), and
fetal systemic vessels (e.g., middle cerebral artery), collectively known as Doppler evaluation, provide
prognostic and diagnostic detail about placentation and fetal adaptation (39).
Regulation of the circulation is a complex fetal behavior, influenced by gestational age and the maternal
environment. Under normal circumstances, the reduction in sympathetic tone created by epidural analgesia does
not affect Doppler flow characteristics of either the uterine or umbilical artery vessels because the spiral
arterioles are maximally dilated and the fetoplacental circulation is stable and tolerant of environmental
changes. However, epidural analgesia has been shown to improve uteroplacental perfusion and effectively
reduce maternal blood pressure in laboring patients with pregnancy-induced hypertension (40). This offers
potential benefits for both the fetus and mother: when uteroplacental perfusion improves, fetal oxygenation and
acid-base balance improve, and when blood pressure is restored to normal levels, the risk of vascular accidents
and organ damage is reduced.
Static Charge-Sensitive Bed

A new method for long-term monitoring of respiration, heart rate, cardiac function, and body movements in
newborn infants was presented in 1984, and possible clinical
P.264
applications were discussed (41). The technology has been used to evaluate neonatal sleep states following
unmedicated vaginal delivery or elective cesarean section delivery under spinal anesthesia (42) and, in
combination with SpO2 and ECG recording, to examine the influence on the newborn baby of maternal fentanyl
analgesia in labor (43).
In the first study, mode of delivery did not affect sleep state distribution during the first day of life (42).
Vaginally born neonates had fewer body movements and more episodes of SpO2 <95% in the first 24 hours
after birth. This result surprised the investigators since babies born by cesarean section are known to develop
respiratory problems more often than infants who are delivered vaginally. In the second study, fentanyl (50 g
of IV bolus q5min until pain relief, then fentanyl patient-controlled analgesia) was compared with paracervical
block (10 mL of bupivacaine 0.25%) in a prospective, randomized fashion. The trial was interrupted after
enrolment of the twelfth healthy, term newborn because there was a significant decrease in SpO2 to 59% in one
of the babies. Interestingly, the SpO2 improved with naloxone administration even though later analysis showed
the concentration of fentanyl in the umbilical vein to be below the detection limit of the assay. Intrapartum
electronic fetal monitoring did not reveal any difference in variability or heart rate between groups. As well,
Apgar scores and analyses of umbilical artery pH were similar. The SpO2 values were lower and the percentage
of minimum SpO2 values between 81% and 90% were more prevalent in the fentanyl group. The static chargesensitive bed (SCSB) method proved sensitive enough to detect lower heart rates and less quiet sleep in the
fentanyl group, suggesting a salutary effect of the opioid on delivery stress (43).
Summary
The nature of the association of anesthetic medications and interventions is complex and can be confounded by
a myriad of factors. As yet, there is no one test that clearly separates effects on the fetus/newborn, if any, of
maternally administered medication during labor and delivery, although newer technologies show some promise.
PAIN MANAGEMENT
For most women, childbirth is likely one of the most painful events in their lifetimes. There are both physiologic
and psychologic aspects to pain and its management (44).
Labor pain evokes a generalized neuroendocrine stress response that has widespread physiologic effects on the
parturient and fetus (45). The neuroendocrine model, presented in Figure 16-2, examines the potential
detrimental consequences of untreated pain. The sequelae of hyperventilation, secretion of stress-related
hormones, and increased oxygen consumption can be prevented, obtunded, or abolished by central neuraxial
blockade (epidural or spinal anesthesia).
Research in humans supports elements of this model (46), but studies are not necessarily designed to consider
the effects of simultaneously occurring care practices on these same physiologic responses. This critique is
needed because it is somewhat counterintuitive that the procreative physiologic process of labor and birth would
by nature have detrimental effects on a healthy mother and fetus (47). An example of a concurrent care
practice is the administration of isotonic sport drinks versus water only during labor (48). Sports drinks were
shown to prevent the development of maternal ketosis without increasing gastric volume, although there was no
difference between the groups in neonatal outcome.
Labor Pain: Implications for the Fetus

Neural pathways and neurochemical systems involved in pain perception are functional from mid-gestation and
are well developed by the third trimester (49,50). Gitau and associates (51) conducted a parallel study of the
fetal and maternal hormonal responses to fetal blood transfusion. They confirmed that the fetus mounts a
hypothalamic-pituitary-adrenal response to transfusion via the intrahepatic vein, which involves piercing the
fetal trunk, but not to transfusion in the umbilical vein at the placental cord insertion, which has no sensory
innervation. The rise in fetal cortisol and endorphin occurred independently of the maternal reaction.
Pretreatment of the fetus with fentanyl for this same procedure attenuated the rise in -endorphin (52).
Hormonal stress responses do not provide a direct index of pain. While it is true that a rise in cortisol and
endorphin is seen as a consequence of painful stimuli in children, other nonpainful situations (e.g., exercise) are
also associated with an increase in the levels of these hormones. Nonetheless, editorial review of Fisk's fentanyl
pretreatment study suggests that fetal analgesia should be given during invasive in utero procedures (53).
At present, there is no literature on fetal pain during labor or delivery.
Labor and Delivery Pain: Implications for the Mother
Visceral pain predominates during the first stage of labor. Nociceptive information arising from uterine
contractions, distention of the lower uterine segment, and cervical dilation is relayed in C afferent fibers to the
dorsal horn of the spinal cord at the T10 to L1 levels. As labor progresses, a mixture of visceral and somatic (Adelta fibers) pain results from traction on the pelvic floor structures surrounding the vaginal vault, and
eventually from distention and stretch of the vagina and perineum (L2-S1). Delivery pain (Stage II) is somatic in
nature and transmitted along the pudendal nerve (S2-4). Synaptic input at the dorsal horn, mediated by
neurotransmitters and chemicals (e.g., excitatory amino acids), is relayed via the spinothalamic tract to higher
centers including the reticular formation, hypothalamus, and limbic system. Dorsal horn neurons also initiate
segmental spinal reflexes. Descending spinal tracts, endogenous opioids, and other inhibitory systems modulate
P.265
nociception centrally in the spinal cord. The neural mechanism of labor shares features with other forms of acute
pain (54).
Figure 16-2 Potential adverse effects of untreated maternal pain on the fetus. (Modified from Brownridge P,
Cohen SE, Ward ME. Neural blockade for obstetrics and gynecologic surgery. In: Cousins MJ, Bridenbaugh PO,
eds. Neural blockade in clinical anesthesia and management of pain, 3rd ed. Philadelphia: Lippincott Williams &
Wilkins, 1998:557-604, with permission.)
To view labor pain only as a neuroendocrine, sensory experience is limiting and undermines the complexity of
this phenomenon (47). Pain is just one component of the totality of the labor and birth experience. Assisting
women to cope with the affective or distress components of labor and birth in a supportive environment has
been shown to reduce the need for pain-relieving drugs, decrease the incidence of operative delivery, result in
higher Apgar scores, and improve breastfeeding success (44,55).
Practice guidelines, including a section devoted to specific analgesia techniques, have been developed to
enhance the quality of anesthetic care for obstetric patients (56). For the obstetrician, analgesia options are
outlined in a practice bulletin that was written to facilitate communication with patients and anesthesia and
neonatology colleagues (57).
The management of pain and anxiety in labor is a worthwhile goal whether the techniques used are
nonpharmacologic, pharmacologic, or include a combination of both. The choice depends on patient preferences,
medical status of the mother and fetus, progress of labor, and resources available at the facility for pain
management and treatment of potential complications.
ANALGESIC TECHNIQUES FOR LABOR: EFFECTS ON THE FETUS AND

NEWBORN
Analgesia refers to pain relief without loss of consciousness. Regional analgesia denotes partial sensory blockade
in a specific area of the body, with or without partial motor blockade. The term neuraxial analgesia pertains to
the administration of pain-relieving medications using caudal, spinal, and/or epidural techniques.
Not all methods of pain relief are available or desirable in all centers, and certain methods are more popular in
different parts of the world (58).
Nonpharmacologic Methods
Proponents of nonpharmacologic methods claim that these methods reduce requirements for analgesia during
the first stage of labor. This does not necessarily imply that women who use these techniques have less pain,
rather that they are able to cope with labor using less analgesia.
In a systematic review of comfort measures, Simkin and O'Hara (59) commented on five methods scientifically
evaluated for their effectiveness in reducing indicators of labor pain. This review also mentioned painrelatedoutcomes
P.266
such as obstetric interventions and duration of labor. Continuous labor support was associated with a decrease
in duration of labor, requests for analgesia, rates of instrumental and cesarean deliveries, and occurrence of
lower Apgar scores. The use of baths offered temporary pain relief and was considered safe provided water
temperatures were maintained at or below maternal body temperature and that immersion duration was
controlled. Perinatal morbidity and mortality did not increase, even if membranes were ruptured. The authors
concluded that there had been insufficient study to provide clear conclusions regarding touch/massage, although
emotional and physical relief was demonstrated with this intervention. Intradermal water blocks were effective
in reducing severe back pain, and one randomized study reported a decrease in cesarean deliveries. Lastly,
maternal movement and positioning was reported to impact pain relief in labor and impact several variables
related to fetal and neonatal well-being. In this systematic review, no trials compared a policy of freedom to
move spontaneously with a policy of restriction to a bed for outcomes such as comfort, labor progress, or fetal
welfare. Mechanisms by which dystocia may be prevented or corrected through the use of maternal positioning
have been discussed elsewhere (60).
Systemic Opioids
From the maternal perspective, efficacy and incidence of side effects with systemic opioid analgesia is largely
dose- rather than drug-dependent. There is little evidence to suggest one agent is intrinsically superior. Most
often, the choice is based on institutional tradition or personal preference.
Opioids may affect the fetus directly as a result of placental transfer and/or indirectly, for example, by altering
maternal minute ventilation or uterine tone. As a group, these low-molecular-weight drugs are lipid-soluble
weak bases (61) that readily cross the placenta. This implies that maternal to fetal concentration gradients are
important; only free, not protein-bound, drug is available for transfer. The amount of free drug delivered to
the placenta depends on placental blood flow and the degree of maternal protein binding. The amount of drug
available to the fetus depends on the degree of placental uptake, metabolism, and clearance (62). In singledose drug studies, key factors influencing umbilical vein/maternal drug ratio are lipid solubility and transit time
through the placental bed. In multidrug dosing (e.g., patient-controlled narcotic analgesia [PCA] delivery
systems), key factors influencing fetal drug levels are the degree of ionization and degree of fetal protein
binding (Fig. 16-3).
Figure 16-3 Factors influencing fetal drug levels with PCA narcotic administration.
Fetal pH is lower than maternal pH; consequently, the fraction of opioid (and other basic drugs) existing in the
ionized state is higher in the fetus than in the mother. Ionization results in drug trapping. The degree of
ionization depends on the drug's pKa; the effect is greater for meperidine (pKa ~8.5) than morphine (pKa
~8.0), and more significant when the fetus is acidotic. This is a simplistic, albeit true, application of opioid
pharmacokinetics, a complex, difficult to predict, and incompletely evaluated topic.
All opioids have the potential to decrease baseline FHR and reduce variability, making interpretation of fetal CTG
recordings potentially problematic. It has been documented from observational studies that parenteral narcotics
can be associated with neonatal respiratory depression, decreased neonatal alertness, inhibition of sucking, and
delay in effective feeding. When evidence related to the use of parenteral opioids for labor pain relief was
subjected to a systematic review (63), it was noted that none of the studies was sufficiently powered to address
the primary outcome measure of neonatal resuscitation, a measure of safety. Intramuscular opioid was
compared to placebo, different i.m. opioid, same i.m. opioid but different dose, and same opioid given
intravenously; IV opioid was compared to different IV opioid and same IV opioid but different modes of
administration. There was insufficient pooled information to draw conclusions regarding any of the secondary
outcome measures, including fetal distress administration of naloxone, Apgar score <7 at 5 minutes, baby
death, admission to a special care setting, feeding problems, and problems with mother-baby interaction.
The concept of genetic imprinting at birth for opiate or amphetamine addiction in later life has been associated
with systemically administered pain-relieving labor medications (narcotics, barbiturates, or nitrous oxide) (63).
The original studies that led to this conclusion were criticized regarding the matching of controls and the
imprinting hypothesis proposed to explain the finding (64). Although a more recent study of drug-abusing
subjects confirmed the phenomenon, these results will be considered controversial until there is more
confirmatory evidence.
Meperidine is the most commonly used opioid for labor analgesia worldwide. It has been shown that, as the time
increases from administration of single-dose, i.m. meperidine 1.5 mg/kg during labor to delivery of the baby, so
too does the level of meperidine in the fetus (65). Maximum fetal concentrations reach a plateau between 1 and
5 hours after dosing; therefore, babies born within 1 to
P.267
5 hours after meperidine is given to the mother have the greatest chance of narcotic-induced depression. In
contrast to single-dose studies, multiple doses of meperidine administered over many hours lead to
accumulation of meperidine's metabolite, normeperidine, in the mother and fetus (66). Half-lives of 17-25 hours
for this metabolite are common in the mother, whereas the half-life exceeds 60 hours in the fetus/newborn.
Normeperidine is associated with respiratory depression, not reversible by naloxone, and seizures. Because of
concerns about meperidine, research has focused on the newer, shorter-acting opioids with no active
metabolites as alternatives.
Fentanyl has been available clinically for more than 20 years. It offers prompt analgesia coupled with a short
duration of action and no active metabolites. Both maternal and fetal drug levels decline in a parallel fashion
following a single dose of the drug (67). In the first report of its administration to laboring patients, fentanyl (50
to 100 g IV q1h) was compared with meperidine (25 to 50 mg IV q2-3hr) (68). More mothers were nauseated
and sedated and more babies required naloxone in the meperidine group.
Sufentanil is the most lipid soluble (octanol:water partition coefficient 1778) of the commonly used opioids (61).
This feature should enhance placental transfer after a single dose, but transfer is impeded by the extent of
maternal plasma protein binding (1-acid glycoprotein) and uptake by the placenta. Sufentanil concentration in
the fetus rises slowly, reaching a plateau between 45 and 80 minutes postadministration (69). It is a useful
maternal analgesic for pain relief during second stage, when fetal delivery is imminent (<45 minutes).
Remifentanil is a novel, ultra-short-acting opioid. It has the most rapid onset of peak effect (~1 minute),
shortest context-sensitive half time (~3-5 minutes), and greatest clearance (40 mL/kg/min) of the commonly
used opioids (61). Although the maternal cardiovascular and side effect profiles are similar to other fentanyl
congeners, remifentanil is chemically distinct because of its ester linkages. This ester structure renders it
susceptible to hydrolysis by red cell- and tissue-nonspecific esterases, resulting in rapid metabolism.
Remifentanil concentration decreases by 50% within 3 to 5 minutes of stopping drug administration, regardless
of the duration of the infusion (61).

Labor pain occurs at intervals and increases in intensity over time. Rapid recovery between contractions and
after delivery is desirable. Therefore, the most effective means of administering remifentanil to the laboring
patient, taking advantage of the drug's characteristics, is via PCA with a background infusion. Bolus dose,
lockout time, and rate of infusion can be titrated (70). The drug rapidly crosses the placenta and is quickly
metabolized by fetal esterases (71). Maternal oxygen desaturation and sedation and reduced FHR variability
have been observed (72). Newborns exposed to remifentanil in utero up until the time of delivery have been
vigorous. There have been no reports of lower Apgar scores, unacceptable umbilical cord gas analyses, or
respiratory depression necessitating naloxone use. (70,71,73)
Naloxone is used to reverse respiratory depression in narcotic-exposed newborns. It is contraindicated for
infants of narcotic-dependent mothers, as administration may precipitate acute withdrawal and seizures.
Naloxone has never been shown to reduce the need for assisted mechanical ventilation or reduce admission
rates to special-care nurseries (74). No studies have evaluated the effect of naloxone on time to spontaneous
effective respiration or long-term outcome. Naloxone use in adults has been associated with reports of
hypertension, pulmonary edema, arrhythmias, and cardiac arrest (61).
Agonist-Antagonist Opioids
Nalbuphine is commonly used as a systemic analgesic during labor. Reports of severe perinatal cardiovascular
and respiratory depression prompted Nicolle and associates (75) to carry out a study designed to delineate
placental transfer and disposition of nalbuphine in the neonate. The estimated half-life was 4.1 hours (versus
0.9 hour in infants and 2 hours in adults). Given that the liver extensively metabolizes nalbuphine, the authors
speculated that the slower neonatal plasma disappearance rate compared to the infant or adult could be due in
part to immature hepatic function or bypass of the liver via the ductus venosus. All 28 babies had 5-minute
Apgar scores of 10. Fifty-four percent of the FHR tracings showed reduced variability lasting 10-35 minutes after
maternal injection.
One potential use for this class of drugs is in the treatment of opiate-dependent pregnant women. Babies born
to mothers on a buprenorphine maintenance program showed little or no clinically measurable neonatal
abstinence syndrome in contrast to findings with methadone, morphine, or heroin maintenance programs (76).
Nitrous Oxide
Nitrous oxide (N2O) is an odorless inhalational agent that exerts weak but prompt analgesic activity. It is a
relatively insoluble gas at room temperature, and therefore equilibrates rapidly between the alveoli, blood, and
brain. To be fully effective, inhalation needs to be timed with contractions such that the patient begins to
breathe the gas about 10 to 15 seconds in advance of the next contraction. This synchronizes the peak effect of
N2O with the zenith of pain, assuming the average contraction lasts 60 seconds and peaks at the midpoint. N2O
is combined with oxygen in a 50:50 mixture for obstetric use and is self-administered by the patient through a
specialized breathing circuit equipped with a demand valve. The negative pressure generated at the onset of
inspiration opens the valve, which remains open during inspiration and closes when the patient begins to exhale.
Inhalation analgesia with N2O during labor and delivery by itself, as a coanalgesic, or as a temporizing measure
pending other forms of pain relief is less common in the United States than in other developed countries.
Any patient at risk of vitamin B12 deficiency (e.g., pernicious anemia or vegetarian) should not use N2O as it
P.268
irreversibly oxidizes vitamin B12, reducing the activity of methionine synthetase (necessary for myelin
formation) and other B12-dependent enzymes. Many countries have set maximal environmental limits for N2O,
which necessitates the use of ventilation systems that allow exhaled gas to be scavenged.
N2O readily crosses the placenta. The maternal-fetal concentration ratio reaches 0.8 within 15 minutes of
continuous inhalation. It has no effect on uterine contractions or FHR. It is not metabolized and is eliminated
quickly and entirely by the lungs with the onset of respiration at birth. This is true whether the mother inhales
N2O for 5 minutes or 5 hours. N2O does not affect Apgar scores or sucking behavior (77).
Paracervical Block
This peripheral block provides a therapeutic alternative for first-stage labor pain when central neuraxial blockade
is contraindicated or unavailable. The technique involves transvaginal injection of LA on either side of the cervix
to interrupt pain transmission at the level of the uterine and cervical plexuses (located at the base of the broad
ligament). Paracervical block (PCB) is relatively easy to perform and, when effective, it provides good to
excellent analgesia that lasts 1-2 hours.
Since the introduction of PCB in the 1940s, reports of serious adverse sequelae, including injection of LA directly
into the uterine arteries or fetal head, fetal death, and profound bradycardia, have resulted in modification of
the injection technique and changes to the concentration and type of LA used. There are statements cautioning
against employing this block in situations of uteroplacental insufficiency or nonreassuring FHR tracings.
Overall, in current practice the incidence of fetal bradycardia postblock is about 15% (78), with the onset
beginning 2 to 10 minutes after injection and the bradycardia lasting 15 to 30 minutes. The exact etiology is
unknown; however, a recent investigation comparing epidural with PCB using Doppler flow velocity waveforms of
the maternal femoral and uterine arteries, and umbilical and fetal middle cerebral arteries has shed some light
on this phenomenon (79). PCB was associated with a small but significant increase in uterine artery impedance,
indicating uterine artery vasoconstriction. In this study, Apgar scores and umbilical arterial and venous pH
determinations were within the normal range in both groups.
Neuraxial Analgesia
Spinal, epidural, and combined spinal-epidural (CSE) techniques are commonplace for managing childbirth pain.
They are used to administer opioids, LAs, and other pain-modulating adjuvants. Collectively, these methods are
considered the most effective forms of pain relief available to laboring women.
Although spinal anesthesia has been in use since 1899, spinal analgesia only became a viable possibility in the
1970s following the discovery of specific opioid receptors in the brain and spinal cord. From a practical
standpoint, however, it was not an option for laboring women at that time for two reasons: there was an
unacceptably high incidence of postspinal headache (also known as postdural puncture headache) in the young
female population; and a single injection technique could not be relied upon to provide analgesia for more than
1 to 2 hours. The advantage of having an epidural catheter in place, either for subsequent bolus dosing or for
continuous infusion, is the provision of uninterrupted analgesia between placement of the catheter and delivery
of the baby.
The (re-)introduction of fine-gauge, pencil-point, atraumatic (noncutting) spinal needles in the late 1980s
fostered a renewed interest in subarachnoid (intrathecal) injection (80). With the advent of CSE equipment and
needle-through-needle technique, single-level subarachnoid injection, followed immediately by epidural
catheter placement at the same site, became possible (81). The CSE procedure has become synonymous with
subarachnoid injection of opioid ( a small dose of LA) and simultaneous initiation of a low-dose epidural
infusion. The perceived advantages of CSE compared with more traditional methods continue to be debated,
along with the consequences of routine dural puncture (81,82,83).
Epidural catheter analgesia alone has been popular for many years. Depending on the choice of agents used, it
provides superior pain relief during the first and second stages of labor and can be extended, if necessary, for
cesarean section, instrumented vaginal delivery, manual placental removal, or episiotomy repair.
Neuraxial Opioids
Opioids injected into the lumbar intrathecal space distribute between nerve tissue and cerebrospinal fluid (CSF)
on the basis of their partition coefficients (lipid solubility). Opioids injected into the epidural space first diffuse
across the dura to reach the subarachnoid space, and then behave as their intrathecal counterparts. Morphine,
the least lipid soluble of the commonly used opioids, diffuses slowly from the CSF into the substantia gelatinosa
of the dorsal horn to activate opioid receptors. This accounts for its delayed onset and prolonged duration of
action. Morphine also spreads rostrally, moving by bulk flow with CSF to reach vasomotor, respiratory, and
vomiting centers in the brainstem. In contrast, the highly lipophilic fentanyl and sufentanil penetrate nerve
tissue quickly. They have a faster onset of activity coupled with a shorter duration of action. Remifentanil is not
approved for use in the intrathecal space because it contains a glycine preservative.
All opioids have some minor intrinsic LA properties, but these effects are marked with meperidine (61), allowing
it to be used as the sole agent, even for cesarean section, in the rare event of amide LA allergy. Intrathecal
meperidine produces significant sympathetic and motor blockade as well as typical opioid side effects such as
pruritus. It has little value as an adjunct to regional analgesia for labor.
P.269
The purported advantages of using opioid drugs alone to induce neuraxial analgesia for labor include:
Preservation of motor function, sustaining the ability to ambulate during first stage and to push during
second stage
Avoidance of LA-induced sympathectomy which can be associated with undesirable cardiovascular

sequelae such as hypotension
Reduction in the systemic side effects of opioids themselves, given the receptor-specific route and minute
amount of drug needed to exert an effect. Less total opioid means less chance of drug transfer to the fetus
and fewer unpleasant maternal side effects such as nausea, vomiting, pruritus, urinary retention, and
sedation.
Fetal bradycardia may develop following administration of intrathecal opioid, although its occurrence can follow
any type of effective labor analgesia (84). While technique-specific etiologies can occur (e.g., maternal
hypotension or fetal LA toxicity), uterine hypertonus as the mechanism inciting the transient but profound drop
in heart rate was first reported in 1994 (85). This topic has been the subject of study (86) and systematic
review (87).
Pain relief can affect uterine function. Hunter (88) reported in 1962 that bilateral lumbar sympathetic block for
first-stage labor pain caused abnormal uterine contraction patterns to normalize and previously normal patterns
to become hyperactive. Although there is likely more than one mechanism, the etiology is thought to be related
to a change in the balance of circulating catecholamines occurring with the advent of analgesia, favoring - over
-activation of smooth muscle receptors (46,89). Uterine muscle tone and vascular resistance increase as a
result of the contraction-inducing norepinephrine influence predominating over the contraction-relaxing
epinephrine effect. FHR decreases because of a reduction in uteroplacental blood flow. The effect may be more
pronounced in the face of oxytocin stimulation.
From an anesthetic perspective, any block that includes segments T10 to T12 (uterine afferent pain fibers enter
the spinal cord at the T10-L1 level) will affect efferent nerves to the adrenal medulla (46). There is a temporal
relationship between the speed of onset of labor pain relief and the appearance, if any, of bradycardia. It occurs
faster with spinal analgesia (<10 minutes) and more slowly with epidural analgesia (15 to 30 minutes) (84).
However, as Van de Velde and associates (86) point out, speed of onset of labor pain relief cannot be the only
factor at work. Nonreassuring FHR tracings did not occur after CSE using a mixture of bupivacaine and sufentanil
(1.5 g) when compared to a larger dose of intrathecal sufentanil (7.5 g) alone, despite equally fast pain relief.
Metanalysis performed during a well-conducted systematic review of this topic revealed a significant increase in
the risk of fetal bradycardia due to intrathecal opioid (odds ratio 1.8, 95% confidence interval 1.0 to 3.1) (87).
The clinical implications are not obvious because the occurrence of FHR changes in response to labor analgesia
has not prompted an increase in the rate of interventional delivery (87). The hypertonus usually lasts less than
10 minutes and can be relieved by administration of a nitric oxide donor such as nitroglycerin (50 to 100 g IV),
or a 2 agonist such as terbutaline (125 to 250 g IV), in cases of prolonged fetal bradycardia. The observed
changes in FHR have not correlated with observable clinical differences in neonatal outcome, including Apgar
scores, cord pH, prevalence of cord pH <7.15, or admission rate to a neonatal intensive care unit (86).
Despite the utility of opioids as neuraxial agents during labor, their use without LAs is confined to early labor; by
themselves, they do not provide adequate relief as labor progresses and pain intensifies (81,90).
Neuraxial Local Anesthetics

LAs exert their effect by penetrating the epineurium and neural cell membrane in the lipid-soluble (un-ionized)
base form to reach the axoplasm. Cellular integrity and metabolism are unaffected. Once inside the cell, the
molecules revert to ionized and un-ionized forms in equilibrium. The charged form blocks sodium channels,
thereby disrupting sodium conductance and preventing depolarization. This interaction between LA and channel
is reversible and ends when the concentration of LA falls below a critical minimum level.
There are several factors that influence the choice of LA agent and concentration employed for epidural or
caudal analgesia. These include the delivery system used (intermittent bolus, continuous infusion, or patientcontrolled epidural bolus background infusion), desired speed of onset, nature of the pain, progress of labor,
degree of motor block tolerable for the patient and anesthesiologist, local practice and experience, and cost. In
general, when LA is used alone, concentrated solutions are required to afford pain relief. Since block density is
dose dependent, the more concentrated the solution used, the greater the degree of motor block expected; this
has been implicated in pelvic muscle relaxation-induced fetal malposition, maternal inability to push, and need
for instrumental delivery (91). In a recent metanalysis comparing randomized clinical trials of techniques using
epidural LA alone with intrathecal opioid alone (92), the authors were unable to find sufficient information to
comment on either similarities or differences in maternal and fetal outcomes. This study was accompanied by an
insightful editorial that will assist readers to critically interpret the results of this and other systematic reviews
and metanalyses (93).
LAs alone are not normally used for spinal analgesia, and use of the caudal technique for pain relief in labor is
uncommon.
Combining Opioids, Local Anesthetics, and Pain-Relieving Adjuvants

When spinal, epidural, and CSE techniques involvecombinations of pain-modulating drugs (sodiumchannelblocking
P.270
agents, opioid-receptor agonists, -adrenergic agonists, and/or acetylcholinesterase inhibitors) acting by
different mechanisms, there is potential to afford analgesia with minimal motor blockade and other relevant side
effects (19,83,94). Lower doses of individual agents in combination often result in at least additive and, at
times, synergistic analgesic effect.
Labor Epidurals and Outcome

An erudite and interesting historical review of anesthesia for childbirth by Caton and associates (95) provides
the foundation for a discussion of the controversies and unresolved issues surrounding the influence of epidural
analgesia on maternal outcomes that affect the fetus and newborn.
There have been five systematic reviews (63,96,97,98,99) comparing epidural to nonepidural analgesia
methods. All focused in some manner on the clinical dilemma of balancing the alleviation of maternal pain with
the possible increase in side effects and/or adverse outcomes for the mother and baby. The search strategies,
choice of inclusion and exclusion criteria, assessment of individual trial validity, primary and secondary
outcomes, and approach to the synthesis of information differed. Commentary about two of the reviews that
were conducted in parallel is thought provoking (100).
The following is a simplified summary of the information provided by the reviews:
Epidural solution concentration and technique varied tremendously over the 38-year timespan covered. As
a result, there were some qualitative differences in the effects of treatment (heterogeneity).
All reviewers concluded that there was insufficient evidence to support an increased incidence of cesarean
section with the use of epidural analgesia.
Data on babies were scanty, apart from gross measures such as Apgar scores and results of umbilical cord
blood gas analysis. Some issues remained unproven; e.g., the suggestion that epidural analgesia is
associated with less naloxone use and higher one-minute Apgar scores. No consistent picture emerged
about the incidence of neonatal adverse effects associated with epidural. There was little evidence
regarding the effects of epidural on fetal physiologic mechanisms.
Epidural analgesia was associated with (not necessarily a causal relationship):
longer second-stage labor (approximately 14 minutes)
more fetal malposition (occiput posterior), possibly due to failure to rotate or because laboring with
a fetus in this position results in more pain, which brings about the request for epidural analgesia
higher rate of instrumental vaginal deliveries
maternal fever
increased use of oxytocin augmentation. This serves as an example of the association versus
causation debate (101). Active labor management protocols, including the routine use of oxytocin,
may simply be associated with a higher demand for epidurals.
Halpern and associates noted (96) that the quality of the clinical trials included in their review improved with
time; all trials after 1995 reported outcomes with patients grouped by intent to treat. Analysis conducted in this
way (i.e., by groups to which patients were randomized) is vital because women who choose epidurals differ
demographically from those who choose other methods of labor analgesia. The former are more likely to be
nulliparous, be admitted to the hospital earlier in labor with higher fetal head positions, have slower rates of
cervical dilation, bear heavier babies, and need oxytocin augmentation more frequently. All of these factors and
degree of maternal pain independently predict the need for cesarean delivery for dystocia (failure to progress
or prolonged labor) (102).
In the five reviews cited above, the nonepidural analgesia control methods primarily consisted of i.m. or IV
administration of opioids. More recently, continuous infusion epidural has been compared to CSE (103) and
patient-controlled epidural analgesia (104). These publications serve to highlight epidural technique
modifications, such as choice of drug, dosage (volume and concentration), and method of administration, that
have taken place over time.
Obstetric management practices have an important role to play in terms of the progress and outcome of labor
(91). For instance, the presence of an epidural block may sometimes decrease the obstetrician's threshold for
performing instrument-assisted deliveries, as well as for allowing instrument-assisted delivery for the purpose of
teaching residents. Active management of labor (routine oxytocin augmentation and/or artificial rupture of
membranes), delayed pushing in second stage, and promotion of ambulation have all been suggested as
methods to reduce obstetric intervention and increase the number of spontaneous vaginal births (94).
Instrumental Vaginal Delivery

Attempts at operative vaginal delivery (vacuum extraction or forceps) put the baby at risk of injury, including
hemorrhagic morbidities such as subgaleal hematoma and cerebral hemorrhage (105). Forceps use is also
associated with facial nerve and brachial plexus injury (106). In a study comparing use of forceps and vacuum
extractors, babies delivered by vacuum extraction showed a higher incidence of cephalohematomas, although
need for resuscitation at birth, admission to a neonatal intensive care unit, and neonatal death rate were not
different (107).
Epidural analgesia has been associated with a higher rate of operative vaginal deliveries (96,97,98,99).
Recently, the Comparative Obstetric Mobile Epidural Trial (COMET) study group (108) showed that there was a
decreased incidence of instrumental vaginal delivery when lower doses of medication (CSE, low-dose LA plus
opioid infusion) were used in labor compared to conventional doses of LA (bupivacaine 0.25%). The presumed
mechanism is the preservation of motor tone and bearing-down reflex. Mild neonatal
P.271
depression necessitating neonatal resuscitation was more common among babies whose mothers were in lowdose LA plus opioid epidural infusion group, most likely due to the cumulative effect of the opioid over time.
Neither the Apgar scores at 5 minutes nor the rates of admission to a neonatal intensive care unit differed
between groups.
Newer methods of epidural analgesia offer the best chance of spontaneous delivery with satisfactory pain control
(109).
Delayed Pushing
The Pushing Early or Pushing Late with Epidural (PEOPLE) study, a multicenter, randomized, controlled trial,
compared conventional early pushing commencing at full cervical dilation with pushing late, 2 hours after full
dilation. Operative delivery was reduced with delayed pushing; however, umbilical arterial pH <7.10 occurred
more frequently among babies whose mothers were in the delayed pushing group. The two groups had similar
rates of neonatal morbidity, including asphyxia. Protocols advocating delayed pushing result in longer second
stages and increased incidences of maternal fever (110).
Maternal Fever
Mothers who choose epidural are more likely to develop a fever during labor (111). The link between epidural
analgesia, maternal fever, and the purported increase in neonatal septic work-ups is less clear
(111,112,113,114); the concern is that babies are more likely to be treated with antibiotics since it is not
currently possible to distinguish between maternal fever from infectious and noninfectious causes during labor.
Many investigators believe the association of an epidural with fever is probably attributable to noninfectious
causes, e.g., altered thermoregulation resulting from epidural analgesia. Neonates born to mothers who receive
epidural analgesia do not have an increased risk of sepsis (91,111).
Breastfeeding
A succinct and thorough review of this topic concludes that intrapartum epidural analgesia does not adversely
affect a baby's or mother's ability to breastfeed (115). The most critical factors for breastfeeding success are
support of and education for the mother.
SURGERY DURING PREGNANCY

Urgent surgery during pregnancy requires modification of anesthetic and surgical approaches to address the
safety of the mother and her fetus (116,117,118). Anesthetic considerations include:
Management of maternal risk factors resulting from physiologic adaptation to the demands of a growing
fetus and ongoing support of the placental unit
Maintenance of the pregnant state
Optimization of uteroplacental perfusion and fetal oxygenation, and maintenance of a stable intrauterine
environment
Attention to the direct and indirect actions of maternally administered medications on fetal well-being
The choice of anesthetic technique is guided by maternal indications, taking into account the site and nature of
surgery. Efforts are made to reduce fetal drug exposure and, with reassurance, allay maternal anxiety. When
possible, regional techniques are preferred because managing the airway of a pregnant patient poses unique
challenges. Edema, weight gain, and increased breast size make intubation of the trachea technically difficult.
Decline in functional residual capacity coupled with increased oxygen consumption predisposes the mother to
rapid desaturation during induction of GA. Lower esophageal sphincter laxity leads to reflux of stomach contents,
increasing the risk of aspiration once protective airway reflexes are abolished. Most abdominal procedures,
however, require GA to provide sufficient muscle relaxation to facilitate surgical exposure.
Approximately 2% of pregnant women require surgery during pregnancy (117). The procedure may be directly
related (e.g., cervical cerclage), indirectly related (e.g., ovarian cystectomy), or unrelated (e.g., appendectomy)
to pregnancy. Semielective procedures should be delayed until the second trimester. Surgery at this time avoids
the vulnerable period of organogenesis (approximately 15- to 60-day gestation) and technical difficulties of
maneuvering around a large, gravid uterus or managing the maternal airway in an advanced stage of
pregnancy. Special techniques, including laparoscopy, cardiopulmonary bypass, transplantation, and induced
hypothermia have all been performed safely during pregnancy (116).
Premature labor represents the greatest risk to the fetus in the perioperative period. Neonatal mortality in the
developed world is approximately 50% at 25 weeks, dropping to about 10% at 30 weeks (119). Postponing
surgery during this period of rapid fetal maturation should weigh the advantages to the fetus against the
hazards that delay poses to the mother. There is no evidence to suggest that any anesthetic agent, dose, or
technique influences the risk of preterm labor (117). Rather, it is more likely to be related to the surgery itself,
manipulation of the uterus, or the underlying condition of the mother (e.g., infection). The more advanced the
pregnancy, the greater the probability of uterine irritability. Certain medications can be used as part of the
anesthetic technique to promote uterine quiescence (e.g., magnesium sulfate, inhalational anesthetic agents, or
2 agonists), and surgical strategies can be employed to avoid handling the uterus. IV, sublingual, or
transcutaneous administration of nitroglycerin is usually reserved for uterine relaxation during brief procedures
or to manage refractory uterine activity (120).
The decision to monitor the fetus during surgery necessitates that someone be available for ongoing
interpretation
P.272
of fetal well-being and that there be a plan for intervention should fetal distress be diagnosed or suspected.
Indicators of fetal distress are often indistinct because technical limitations at various gestational ages eclipse
data acquisition, and FHR variability is reduced or eliminated by certain anesthetic drugs. Intervention may
include delivery, reassessment of anesthetic depth, or a more aggressive approach to maximize uterine blood
flow, tocolysis, and/or maternal oxygenation (116). If delivery of the fetus is planned to occur at the same time
as surgery, a coordinated team approach involving anesthesia, obstetrics, surgery, nursing, respiratory therapy,
and neonatology is vital.
The well-being of the fetus is dependent on the adequacy of the maternal blood supply to the placenta, which is
mainly derived from the uterine arteries (121). Uterine artery blood flow increases during pregnancy and
approaches 500 to 800 mL/min (10% to 15% of maternal cardiac output) at term. The uterine vascular bed is a
low-resistance system, not capable of further dilatation and devoid of autoregulation. Therefore, placental blood
flow varies directly with net perfusion pressure (uterine artery pressure - uterine venous pressure) across the
intervillous space and inversely with uterine vascular resistance. When faced with maternal hypotension, to
preserve uteroplacental perfusion in a pressure-passive system, a more aggressive approach to management
(rapid fluid loading, vasopressor therapy, Trendelenburg and left lateral positioning) is required compared to
strategies for the nonpregnant patient (118). Hypotension may be due to many different etiologies but
commonly results from aortocaval compression in the supine position, general or high spinal anesthesia, or
hemorrhage. Bleeding from the uterine vessels can be very brisk and can quickly lead to life-threatening
hemorrhage. Left lateral decubitus positioning prevents aortocaval compression in the second and third
trimesters. This can be accomplished by having the mother lie on her left side or by elevating the right hip with
a wedge, as illustrated in Figure 16-4. Maintaining homeostasis in the intrauterine environment also requires
attention to maternal oxygenation, temperature, and acid-base balance (respiratory and metabolic).
Most anesthetic agents are not known to be teratogens. When evaluating the possibility of teratogenicity from
maternally administered anesthetic medications, points to be considered include (116,122):
The incidence of congenital anomalies in the developed world is 3%.
Human teratogenicity studies are impossible to perform for ethical reasons.
Extrapolation from animal studies may not be valid.
Hypoxemia and hypotension cause physiologic derangement that may be teratogenic.
Figure 16-4 Left lateral tilt to relieve aortocaval compression.
Drugs that are usually avoided during anesthesia for long surgical procedures in early pregnancy include N2O
and benzodiazepines. N2O is avoided because it causes oxidation of vitamin B12, rendering it incapable of
functioning as a cofactor for methionine synthetase, an enzyme necessary for DNA synthesis in humans.
Benzodiazepines are avoided because epidemiologic studies have shown a link to the development of congenital
inguinal hernia (122).
Postoperative pain management may include plexus blocks or epidurals, when appropriate, to limit fetal
exposure to drugs. Opioids and acetaminophen are used widely. Prolonged use of nonsteroidal antiinflammatory
drugs (NSAIDs) is avoided due to concerns about premature constriction of the ductus arteriosus and
development of oligohydramnios.
Fetal Surgery
Fetal surgery is defined as the performance of procedures on the fetus or placenta designed to alter the natural
history of a fetal disease that is diagnosed in utero (123). Surgery can vary from minimally invasive
percutaneous procedures, facilitated by local, spinal, or epidural anesthesia, to direct fetal operations following a
hysterotomy incision. The latter requires maternal GA and attention to the possibility of inflicting pain on the
fetus (124). Anesthetic considerations are identical to those for nonobstetric surgery during pregnancy. Once
again, care needs to be
P.273
provided to two patients simultaneously, although the fetus is the primary patient in these circumstances.
Fetal sedation by placental transfer of maternally administered medication is not reliable and does not ensure an
anesthetized or immobile fetus. Given enough time and subject to their individual solubilities, inhalation
anesthetic agents used for maternal GA and uterine relaxation equilibrate in fetal tissues. Deep maternal
inhalation anesthesia may result in progressive fetal acidosis by an uncertain mechanism. Fetal blood pressure,
heart rate, oxygen saturation, and base excess can decrease due to direct impairment of fetal myocardial
contractility, redistribution of fetal blood flow, or changes in uterine perfusion. Fetal distress and response to
maneuvers can be recognized and managed by measuring heart rate, blood pressure, and umbilical blood flow,
and by monitoring pH, pCO2, pO2, base deficit, glucose, and electrolytes. Vascular access facilitates this and the
administration of fluid, blood products, and/or drugs (125).
Additional fetal anesthesia can be provided by direct i.m. or intravascular (via the umbilical vein) administration
of opioids and neuromuscular blocking agents. Pancuronium is often chosen for fetal paralysis because of its
long duration and vagolytic properties, which help elevate the FHR and maintain cardiac output. Fentanyl, in
relatively large doses (12.5 to 25 g/kg estimated fetal weight), attenuates the autonomic and hormonal stress
response during potentially painful procedures (52,125). In the face of intense uterine tocolysis, maintenance of
maternal blood pressure may require concomitant vasopressor therapy.
The ex utero intrapartum treatment (EXIT) procedure was developed for fetuses that have a predictably
compromised airway, either because of prior in utero surgery (e.g., to treat congenital diaphragmatic hernia) or
due to an obstructing mass, such as cystic hygroma or thyroid goiter. Delivery occurs by planned cesarean
section with an anesthetic approach that maintains uterine relaxation. A hysterotomy incision is made with a
device that limits uterine bleeding, and the fetus is partially delivered through the incision. The surgeon
performs laryngoscopy or tracheotomy and secures the airway (endotracheal tube or tracheotomy tube) while
the fetus is still attached to the umbilical cord and maintained on uteroplacental perfusion. Attention is paid to
avoiding fetal hypothermia. The fetal lungs are expanded and surfactant administered if the infant is premature.
The cord is then clamped and the remainder of the cesarean section proceeds as usual. Fetal well-being and
operating conditions have been maintained for up to 2 hours during EXIT procedures (123,125).
Whether the fetus feels pain, and from what gestational age, has been the subject of vigorous debate
(124,126,127). Prior to 22 weeks, the fetus does not have the neuro-anatomic pathways in place to feel pain;
between 22 and 26 weeks, thalamocortical fibers, considered to be crucial for nociception, are forming; and
after 26 weeks, the fetus has the necessary neurologic development to feel pain. Investigators have used
surrogate end points, including fetal reflex movement away from and biochemical stress response to noxious
stimuli in an attempt to define markers of pain. Hormonal and circulatory stress responses to invasive
procedures are observed by 20 weeks (49,50,51,52,53). Further definition of the neuroanatomic and
neurophysiologic maturation of sensory pathways involved in pain transmission in the human fetus may provide
more direct information about the fetal pain experience.
RESUSCITATION ISSUES DURING PREGNANCY

Intrauterine Resuscitation of the Compromised Fetus
The term fetal distress is used to identify a state of progressive fetal asphyxia with hypoxia and acidosis,
which, if not corrected, results in decompensation and permanent organ damage. It is diagnosed or suspected
when characteristically abnormal features emerge in one or more of the tests used to evaluate fetal well-being.
Intrauterine resuscitation consists of a series of maneuvers designed to reverse treatable causes of fetal
asphyxia, restore fetal oxygenation, and correct fetal acidosis. These maneuvers are fundamental to the practice
of obstetric anesthesia and are summarized, using an evidence-based template, by Thurlow and Kinsella (128).
The circumstances surrounding the onset of fetal distress direct the order of application and determine which
aspects of intrauterine resuscitation are appropriate for a particular patient. The goals and measures are:
Increase blood flow to the placenta
treat maternal hypotension aggressively by administering a nonglucose-containing crystalloid/colloid

IV fluid bolus, stopping the epidural infusion, and giving vasopressor therapy
alleviate aortocaval compression by changing the mother's position until an improvement in FHR
occurs (left lateral, followed by right lateral, and finally knee chest position)
reverse uterine artery vasoconstriction induced by hypocapnia from maternal hyperventilation by

managing pain and providing verbal reassurance
Relax uterine muscle
stop oxytocin infusion
administer a tocolytic such as 50 to 100 g of IV nitroglycerin
Increase fetal oxygenation
administer 100% oxygen to the mother by face mask
relieve umbilical cord compression by changing the mother's position or, if oligohydramnios is
present, consider amnioinfusion
Rule out umbilical cord prolapse or, if present, provide manual elevation of the presenting part per vagina,
maintaining warmth and moisture for the cord until emergent delivery
Confirm fetal asphyxia by an alternate test
Prepare for emergent delivery
P.274
Maternal Cardiac Arrest

The management of cardiac arrest in pregnancy is outlined in the Guidelines 2000 for Cardiopulmonary
Resuscitation and Emergency Cardiovascular Care (129). Acute cardiac life support protocols are modified in
pregnancy, but the standard adult algorithms for medication, intubation, and defibrillation still apply.
The key to resuscitation of the fetus is resuscitation of the mother. Relief of aortocaval compression is
paramount. Chest compressions are performed higher on the sternum to adjust for the shift of abdominal
contents toward the head. Consideration of arrest etiologies unique to pregnancy (e.g., amniotic fluid embolism)
and diagnoses exacerbated by the physiologic changes of pregnancy (e.g., peripartum cardiomyopathy) is
important if response to resuscitative efforts is lacking. All medication infusions such as magnesium sulfate,
oxytocin, or epidural are discontinued and early intubation is encouraged to reduce the risk of aspiration. Fetal
surveillance monitors (e.g., scalp electrode lead) must be discontinued prior to defibrillation. The decision to
perform open-chest cardiac massage or emergent cesarean section should occur earlier rather than later if
circulation is not restored by usual measures. Perimortem cesarean section has the highest chance of improving
outcome for both mother and baby when uterine size is >20 weeks (some would say >24 weeks) and delivery
occurs within 5 minutes of the onset of the arrest (130).
Perimortem Cesarean Section

Perimortem cesarean section refers to emergent operative delivery of the fetus through a midline, classical
uterine incision in situations in which the mother is moribund or pulseless and is being actively resuscitated.
Emptying the uterine cavity increases maternal cardiac output (contraction of the myometrium results in
autotransfusion of blood previously directed to the uterus and improved preload secondary to relief of inferior
vena cava compression) and lung volume and makes it easier to perform effective chest compressions.
Accurate data regarding the incidence and outcome of this procedure for the mother or neonate are difficult to
obtain. For example, the British Confidential Enquiry does not report deliveries during which the mother has
been successfully resuscitated (131). There have been several case reports of expedient cesarean delivery
during which both mother and baby have survived fully intact. Perimortem cesarean section is an established
part of the resuscitation process in near-term pregnant women (129). In the largest series to date, Katz and
associates (132) reviewed the literature from 1900 to 1985 for papers in which time from cardiopulmonary
arrest to delivery was listed. Most fetal survivors were delivered within 5 minutes. The general consensus is that
the likelihood of perimortem cesarean section resulting in a living and neurologically normal baby is related to
the interval between onset of maternal cardiac arrest and delivery. Also, the chances of normal survival are
good following delivery if the fetus lives past the first few days (131). For this technique to be an option, it must
be considered in the context of immediately available staff and equipment to perform the cesarean section, and
staff and equipment to conduct the neonatal resuscitation.
ANESTHETIC TECHNIQUES FOR CESAREAN SECTION: EFFECTS ON

THE FETUS AND NEWBORN
General Anesthesia
GA is a reversible state characterized by loss of consciousness and lack of pain with or without skeletal muscle
relaxation. Going to sleep, staying asleep, and waking up are known as induction, maintenance, and
emergence. The state of GA is achieved through the use of drugs administered in a specific order, namely,
induction agents (possibly including narcotics), neuromuscular blockers, inhalation agents, analgesics, and
reversal agents.
Given enough time, all medication administered to the mother crosses the placenta and enters the umbilical
vein, so gauging drug administration during the induction and maintenance phases of GA is important. An
important factor affecting neonatal outcome is the elapsed time between the induction of anesthesia and
clamping of the umbilical cord, as this represents the time of fetal exposure to maternally administered
medication. A second factor is the time from uterine incision to delivery of the baby. A long incision-to-delivery
time is associated with an increased incidence of fetal acidosis, presumably caused by uteroplacental
vasoconstriction. If possible, induction-to-clamp time should be <10 minutes and uterine incision-to-delivery
time <3 minutes (133).
The three determinants of placental transfer of drugs to the fetus include the physical-chemical properties of the
drug, characteristics of the maternal, placental, and fetal circulations, and placental anatomy and physiology.
Fetal and neonatal pharmacologic effects of anesthetic agents given to the mother during a cesarean section
conducted under GA depend on the amount of drug reaching the fetus. Estimating this is not an easy task.
There are difficulties associated with human in vivo studies of placental transfer during pregnancy (134). The
fetoplacental unit is inaccessible in situ, and there are ethical considerations in conjunction with maternal and
fetal safety. In vivo studies are most commonly performed at birth by collecting maternal venous and umbilical
cord arterial and venous blood samples. It is difficult to draw conclusions based on one set of measurements.
Likewise, the applicability of animal placentas as models for the human placenta is limited because the structure
and function of the placenta is species specific. Many studies of anesthetic pharmacology to date have been
conducted using animal models. The alternative is to use a human ex vivo placental perfusion model. Ala-Kokko
and associates
P.275
(134) make a strong case for the human placental perfusion model as the best method available for studying
transplacental passage of drugs.
It will take time to study anesthesia drugs using this methodology. The conclusions reached in most of the
studies cited in the following discussion derive from animal and in vivo blood sampling data.
The standard method by which anesthesia is induced for cesarean section is rapid sequence induction. The
process consists of giving 100% oxygen by mask, IV administration of an induction agent narcotic and
neuromuscular-blocking drug followed by application of cricoid pressure and intubation of the trachea. The
induction agents used to initiate GA include sodium thiopental, methohexital, ketamine, propofol, and
midazolam. Keta-mine is usually reserved for situations involving maternal hemodynamic instability because it
preserves sympathetic outflow. The others have been studied and compared (135,136,137). Midazolam and
propofol have been associated with longer induction times, a lighter plane of maternal anesthesia (as measured
by electroencephalogram), and lower Apgar scores. Both sodium thiopental and methohexital are highly lipid
soluble. They share pharmacokinetic properties with thiamylal, another barbiturate that peaks in umbilical
arterial plasma at 3 to 5 minutes and declines rapidly until 11 minutes (133,137). Induction to umbilical cord
clamp times of approximately 10 minutes coincide with declining fetal levels of these agents and therefore little
neonatal depression.
Neuromuscular-blocking drugs share a structural similarity, a quaternary ammonium ion, which slows but does
not eliminate transfer of these drugs across the placenta (133). Succinylcholine is the only depolarizing drug
available for clinical use. In the normal parturient, it is degraded so rapidly by plasma cholinesterase that
virtually none reaches the fetus, whereas the percentage of nondepolarizing neuromuscular-blocking drug (e.g.,
rocuronium, pancuronium, and atracurium) that crosses the placenta ranges from 7% to 22%, depending on the
drug. The literature is vague with respect to effects on the neonate. However, in the setting of high-dose
nondepolarizing neuromuscular blockade (e.g., EXIT procedures), it may be necessary to support neonatal
ventilation for a period of time or to administer reversal agents.
Inhalation anesthetics (not including N2O) are also known as volatile agents. Halothane, enflurane, isoflurane,
and the newer desflurane and sevoflurane are examples of drugs used to maintain anesthesia during cesarean
section. Achieving adequate depth of inhaled anesthesia depends on how quickly partial pressures of a particular
volatile agent equilibrate in alveolar, blood, and brain compartments. The less soluble the agent, the faster a
deep plane of anesthesia is attained. Desflurane and sevoflurane are much less soluble than the other agents so
theoretically they would be expected to cross the placenta and equilibrate in fetal tissues more rapidly than their
more soluble counterparts, potentially resulting in a more depressed neonate. Equally expected, however, once
the newborn establishes ventilation, is that the lungs more quickly would excrete (blow off) these relatively
insoluble drugs. Desflurane is more pungent and irritating to the airway and may result in laryngospasm. This
should be considered when suctioning the neonate whose mother has received desflurane (133). When
compared with isoflurane 0.5%, sevoflurane 1% (equianesthetic concentration) was found to produce similar
maternal and neonatal results (138). Cord blood gases and Apgar scores were equivalent. Desflurane in a
subanesthetic dose (3%), mixed with N2O-O2, was considered safe and effective for cesarean section in healthy
parturients when compared to enflurane 0.6%. Higher doses of desflurane delayed time to sustained respiration
in the newborn (139).
Administration of GA to a parturient is demanding and dissimilar in terms of the drugs and techniques used to
achieve the same state in an elective surgical patient. With the exception of a planned, elective cesarean
section, circumstances surrounding labor and delivery are difficult to control. Consequently, the obstetric patient
rarely comes to the operating room in optimal condition. Combine this with the effect on the parturient of the
physiologic changes of pregnancy, add in the fact that mother and baby have unrelated anesthetic needs, and
even a healthy woman becomes a high anesthetic risk.
Complications of GA for cesarean section remain the leading cause of anesthesia-related mortality. The case
fatality ratio for GA versus regional anesthesia (RA) during obstetric delivery for the period 1991 to 1996 was
6.7:1 (140). In parturients who die from complications of GA, airway problems (failed intubation or aspiration)
represent the most frequent cause of death (141). The incidence of failed intubation in obstetric patients is
1:250 to 1:280 compared with 1:2,230 for provision of GA in the main operating room (140,141); thus, the
anesthesiologist working in the labor and delivery suite is seven times more likely to encounter failed intubation.
The use of GA for cesarean section is declining in favor of RA techniques (142,143), prompting concern over the
number of opportunities remaining for trainees to learn, and anesthesiologists to maintain, their obstetric airway
management skills (144).
GA continues to be indicated in certain situations including, but not limited to, expedited delivery, technically
impossible or failed RA, coagulopathy, cardiovascular instability, anticipated hemorrhage, tethered spinal cord,
and patient preference.
Regional Anesthesia
RA is the loss of all sensation, motor function, and reflex activity in a specific area of the body. The surgical
conduct of cesarean section under RA requires a sensory block to the fourth to sixth thoracic dermatome levels
(T4-T6). Although the incision is most commonly in the lower abdomen, traction on the peritoneum uterine
exteriorization can cause discomfort unless the area extending from mid-thoracic to sacral levels is blocked.
Either an epidural or spinal approach will suffice, but larger doses of
P.276
LA and opioid medications are required than when the same techniques are used to provide analgesia in labor.
As a result, the fetus/neonate may be affected directly by placental transfer of LAs and opioids, or indirectly by
sympathetic blockade, with resultant alterations in uteroplacental perfusion (133).
LA toxicity manifests systemically as a progressive continuum of symptoms beginning with tongue numbness
and tinnitus, moving to visual disturbance and muscle twitching, and finally escalating into convulsions, coma,
arrhythmias, and respiratory arrest. Toxic LA levels are most often produced by inadvertent IV injection;
however, toxicity can result through absorption from the epidural space, particularly given the large dose of LA
required to produce T4 blockade. The search for LAs with low toxic profiles has led to four modern-day agents:
lidocaine, bupivacaine, ropivacaine, and levobupivacaine (145,146). Lidocaine is subject to tachyphylaxis and is
often mixed with low-dose epinephrine to delay its otherwise rapid absorption from the epidural space. Its
toxicity profile includes central nervous system rather than cardiac hyperactivity. Bupivacaine is marketed as a
racemic mixture. In the 1970s and early 1980s epidural anesthesia with higher concentrations of this drug was
associated with lethal ventricular arrhythmias and cardiovascular collapse. It became apparent that the Renantiomer was responsible. Bupivacaine is still in widespread use today, but much smaller doses are employed
for labor analgesia and it is not commonly used in the epidural space to provide anesthesia for cesarean section.
Ropivacaine and levobupivacaine are single enantiomer (L-form) LAs. Evidence from a variety of experiments in
several species supports a reduction in toxicity (145,146).
Labor epidural analgesia can be extended and made denser if the parturient is delivered by cesarean section.
Under normal maternal and fetal conditions, skillfully conducted GA and RA are almost equivalent with respect to
neonatal well-being (147,148). Nevertheless, given the risks to the mother and the association of lower Apgar
scores with GA, RA for elective, and sometimes emergent, cesarean section is preferred (19,142,149,150). A
compromised fetus may even benefit from anticipatory maternal epidural catheter placement in labor when
there is a high risk of cesarean section (151) or primary epidural or spinal anesthesia for elective cesarean
section (150,152). RA results in less neonatal exposure to drugs (especially when the spinal technique is used),
allows the mother and her partner to participate in the birth of their baby, and provides better maternal
postoperative pain relief (143).
For all the advantages of spinal anesthesia such as simplicity of technique, rapid onset, reduced risk of systemic
toxicity, density of anesthetic block, and postoperative pain relief afforded by neuraxial morphine, the potential
for hypotension with this technique poses the greatest threat to the mother and fetus (19). The incidence of
hypotension is similar between epidural and spinal anesthesia but occurs earlier and more rapidly with the spinal
approach. Hypotension results from temporary sympathectomy, an inevitable but undesirable component of midthoracic blockade. Reduced preload (increased venous capacitance and pooling of blood volume in the splanchnic
bed and lower extremities) and reduced afterload (decreased systemic vascular resistance) lower maternal mean
arterial pressure (MAP), leading to nausea, lightheadedness and dysphoria, and reduced uteroplacental
perfusion. When maternal MAP is maintained, maternal symptoms are averted and uteroplacental perfusion
improves.
In their epidemiologic study of 5,806 cesarean deliveries, Mueller and associates (153) concluded that fetal
acidemia was significantly increased after spinal anesthesia and maternal arterial hypotension was by far the
most common problem encountered. The prevalence of fetal acidemia with RA for cesarean section has been
confirmed in another study (150). However, isolated acidemia does not correlate with Apgar scores and is a poor
indicator of outcome. Low umbilical artery pH reflects both the respiratory and metabolic components of
acidosis, whereas base excess reflects only the metabolic component. It is base excess that correlates with
neonatal outcome, values more negative than -12 mmol/L having an association with moderate to severe
newborn encephalopathy (7). However, prevention of hypotension is advantageous to minimize any influence on
neonatal acid-base status.
The routine measures used to maintain uteroplacental perfusion include left lateral tilt position, lower leg
compressive stockings, and IV fluid loading (121,154). Vasopressor therapy is reserved for the treatment of
hypotension. Prophylactic use of ephedrine in one study (155) and therapeutic use in another (150) possibly
contributed to fetal acidemia. Likewise, ephedrine use was associated with lower umbilical arterial pH values
when compared with phenylephrine in a systematic review (156). The literature is replete with debate regarding
which vasopressor, a mixed agonist (e.g., ephedrine) or a pure agonist (e.g., phenylephrine), would be
more appropriate for the management of hypotension during spinal anesthesia for cesarean delivery
(157,158,159,160). The controversy revolves around the etiology of fetal acidemia: Is it due to the metabolic
effects of -stimulation in the fetus or insufficient maintenance of uteroplacental perfusion by failure to reclaim
sequestered blood from the splanchnic bed to augment preload? Regardless, the choice of vasopressor drug is
perhaps less important than the avoidance of hypotension (161).
Fetal Effects of Maternal Oxygen Administration

Providing supplemental oxygen to the parturient during cesarean section is common clinically, whether the
procedure is elective, urgent, or emergent (162). This practice has been justified by the belief that raising the
oxygen reserve of the mother is universally beneficial for the fetus. Since the advent of SpO2, patients who may
benefit from oxygen therapy are more easily identified and clinicians can be more selective about administering
oxygen therapeutically.
P.277
For example, SpO2 readings obtained during elective cesarean section conducted under spinal anesthesia
showed that maternal saturation was well maintained on room air, and administration of 35% oxygen by
facemask failed to significantly change umbilical vein pH or partial pressure of oxygen (163). This finding was
further substantiated in a study by Cogliano and associates (164).
Oxygen given to the mother does not increase fetal oxygenation to the same extent because of usage during
intermediary placental metabolism. In comparison, neonatal oxygen therapy has the potential to raise newborn
PaO2 substantially. There is a growing body of clinical and experimental evidence that seems to indicate that the
practice of routine supplemental oxygen during neonatal resuscitation may be injurious (165,166,167).
Oxidative stress is implicated as a common underlying mechanism in several neonatal conditions, including
necrotizing enterocolitis, retinopathy of prematurity, periventricular leukomalacia, and chronic lung disease
(168,179). It occurs when free radical generation exceeds the body's antioxidant defense mechanisms, and it is
not exclusive to newborns; free radicals and antioxidants also play a role in adult diseases (170,171). The
interest in free radicals as harbingers of disease prompted investigators to explore potential fetal effects of
giving oxygen to mothers. Accordingly, the process of routine oxygen supplementation for healthy parturients
undergoing elective cesarean delivery with RA has been questioned. (172).
Free radicals have a brief lifespan, making their detection difficult. Therefore, studies investigating oxidative
stress usually measure surrogate markers, namely products of the attack by free radicals on lipids, proteins, and
nucleotides. This methodology was used to examine the effect on the newborn of administering air or oxygenenriched air to parturients undergoing elective cesarean section (173). There was a clear difference between
groups, with greater free radical activity in the babies born to mothers breathing oxygen-enriched air. The main
site of free radical generation was the placenta, as evidenced by the higher concentration of free radicals in the
umbilical vein compared to the artery. As Backe and Lyons point out, at present, we have no means of linking
free radical formation with neonatal outcome following elective cesarean section. in a low-risk situation such as
[this], a favorable outcome is unlikely to be influenced by maternal hyperoxia (174). The significance of [the
Khaw and associates study] relates to the use of high inspired maternal oxygen fractions (60%) for the delivery
of compromised and premature [babies]. There are no published trials addressing maternal oxygen therapy for
fetal distress (175).
POSTPARTUM PAIN MANAGEMENT

The genesis of postpartum pain is multifactorial. Postsurgical etiologies include episiotomy, manual placenta
removal, postpartum tubal ligation, and/or visceral/ incisional pain following cesarean delivery. For some
women, cramping originating from uterine involution or localized pain resulting from perineal tears can result in
distressing discomfort. The mother may suffer a postspinal headache as a complication of regional analgesia or
anesthesia. Any method that reduces the magnitude of these painful experiences is desirable. This section is
limited to pharmacologic therapy and its implications for the breastfeeding mother.
The goal of a pain management regimen is to prevent or control pain to a degree that is acceptable to the
patient while facilitating activities of daily living and promoting quality of life. There are additional challenges for
treating pain in women who have just given birth. New mothers are motivated to ambulate early and care for
their babies, which requires them to be clear-headed and to experience minimal side effects from analgesics.
Concern that postpartum analgesia could have implications for the baby may make the mother hesitant to take
an analgesic, and therefore suffer unnecessarily. Breastfeeding calls for use of analgesic agents that are
transferred minimally into breast milk and have little effect on the neonate (176).
The current trend is to use a balanced, multimodal approach (177,178). Additive or synergistic analgesic effects,
with fewer side effects, have been demonstrated by combining agents with different mechanisms of action in
smaller doses than would be required if the individual agents were used alone. This concept has become so
important that, as part of the accreditation process, health care organizations are surveyed regarding pain
assessment and management, looking for treatment initiatives that include a multimodal approach (179).
The act of breastfeeding results in milk production, secretion, let-down, and increased breast blood flow.
Therefore, the timing of breastfeeding relative to drug administration influences the amount of drug that
appears in the breast milk. Drug exposure can be minimized if the mother takes medication immediately after
nursing or just before the baby is due to have a lengthy sleep, and uses short-acting medications. The neonatal
dose of most medications obtained through breastfeeding is 1% to 2% of the maternal dose (122).
Breast milk content also affects the extent of drug transfer from maternal plasma. For instance, lipid-soluble
drugs are less likely to accumulate in colostrum, which contains little fat. Colostrum has the same pH as
maternal plasma, which is advantageous as far as narcotics are concerned. These weak bases are not
sequestered in colostrum via ion trapping and therefore do not accumulate. Colostrum morphine concentrations
have been measured during the first 48 hours of postcesarean IV patient-controlled analgesia (180). The milk-tomaternal plasma ratio for morphine was <1. Morphine was seen in very small concentrations in less than half of
the milk samples. The authors concluded that the amount of drug likely to be transferred to the breastfed
neonate was negligible. Only small amounts of colostrum are secreted during the first few postpartum days (10
to 120 mL/day) so newborn exposure from volume is limited. Ultimately, the concentration of drug in the
P.278
baby's plasma is more important than the concentration of the drug in colostrum or breast milk. This depends on
absorption across the gastrointestinal tract, volume of distribution, and extent of metabolism and excretion in
the newborn. Little is known about the bioavailability of analgesics and their metabolites because of ethical
issues involved in repeated blood sampling from babies.
Postcesarean section pain peaks on the second day after surgery (181), and analgesic usage begins to decline.
Milk composition continues to change over the first 10 days after birth. There is a gradual increase in fat and
lactose content and a reduction in protein and pH. By day 10 postpartum, factors such as high lipid solubility,
low molecular weight, minimal protein binding, and the un-ionized state facilitate secretion of medications into
mature breast milk (182). Women who breastfeed and require GA for surgery are usually counseled to feed their
baby before the surgery and temporarily interrupt feeding postoperatively by wasting the first milk sample
(express with a breast pump and discard). After that, if the mother feels well enough and there are no surgical
contraindications, she is encouraged to resume feeding. Most anesthetics are rapidly cleared from the mother;
some authors argue that no portion of human milk need be wasted (182).
Postdelivery analgesia should be tailored to match the changing severity of pain over time; as well, prompt
recognition and treatment of side effects help to optimize pain management. Epidural or intrathecal morphine is
commonly administered when neuraxial blockade is used for delivery. The analgesic effect following a single
neuraxial dose can last up to 18 to 24 hours; however, in keeping with the multimodal approach, fixed regimen
or on-request analgesics such as NSAIDs or acetaminophen (paracetamol) are usually prescribed
concomitantly.
Mild analgesics (acetaminophen and NSAIDs) provide background pain relief to which opioids and/or adjuvant
analgesics can be added (178,183). Fixed-dose combinations (e.g., acetaminophen plus codeine) have
established efficacy and safety. They are widely used for postpartum pain management (184). Individual
titration of opioids is essential. Different routes of administration, including oral, IV, and patient-controlled IV or
patient-controlled epidural infusion should be available. A significant reduction in postpartum narcotic use can be
achieved through implementation of a self-medication program (185). The AAP published a statement on drug
transfer into human milk and possible effects on the infant or on lactation to assist prescribing practices (186).
The AAP considers acetaminophen, most NSAIDs, and morphine compatible with breastfeeding.
CONCLUSION
The ideal analgesia/anesthesia for labor and delivery would meet the following criteria (82):
Provide fast, effective, and continuous pain relief while maintaining the parturient's ability to move and
ambulate throughout labor, and to push during vaginal delivery
Not interfere with the progress of labor and possibly improve the course of a dysfunctional labor
Be technically straightforward and devoid of unacceptable side effects, complications, or risks
Not imperil the fetus in utero
Provide satisfactory conditions for delivery
Permit early mother-baby interaction
Have no short- or long-term impact on neonatal outcome
Not unexpectedly, no single technique of analgesia or anesthesia meets all of these criteria. The challenge for
the anesthesiologist is to balance the needs of the mother and fetus while being flexible enough to modify or
change the approach as circumstances dictate.
ACKNOWLEDGEMENT
My thanks to the staff at the Neil John Maclean Health Sciences Library, University of Manitoba; to online search
system developers everywhere for helping to simplify the process of identifying and retrieving journal articles; to
the inventors of systematic reviews and metanalyses; and to Mr. Hiscoke, my grade 7 science teacher, for
introducing me to the thesaurus.
REFERENCES
1. Caton D. The history of obstetric anesthesia. In: Chestnut DH, ed. Obstetric Anesthesia, Principles and
Practice, 2nd ed. New York: Mosby, 1999:1-13.
2. Apgar V. A proposal for a new method of evaluation of the newborn infant. Curr Res Anesth Analg
1953;32:260-267.
3. Committee on Fetus and Newborn, American Academy of Pediatrics, and Committee on Obstetric Practice,
American College of Obstetricians and Gynecologists. Use and abuse of the Apgar score. Pediatrics
1996;98:141-142.
4. Papile LA. The Apgar score in the 21st century. N Engl J Med 2001;344:519-520.
5. Thorp JA, Rushing RS. Umbilical cord blood gas analysis. Obstet Gynecol Clin North Am 1999;26:695-709.
6. Helwig JT, Parer JT, Kilpatrick SJ, et al. Umbilical cord blood acid-base state: what is normal? Am J Obstet
Gynecol 1996;174:1807- 1812.
7. Ross MG, Gala R. Use of umbilical artery base excess: Algorithm for the timing of hypoxic injury. Am J
Obstet Gynecol 2002;187: 1-9.
8. National Committee for Clinical Laboratory Standards. Blood gas preanalytical considerations: specimen
collection, calibration and controls (approved guideline C27-A). Wayne, PA: NCCLS; 1993.
9. Yoon BH, Kim SW. The effect of labor on the normal values of umbilical blood acid-base status. Acta Obstet
Gynecol Scand 1994; 73:555-561.
10. Reynolds F, Sharma SK, Seed PT. Analgesia in labour and fetal acid-base balance: a meta-analysis
comparing epidural with systemic opioid analgesia. BJOG 2002;109:1344-1353.
11. Greene KR. Scalp blood gas analysis. Obstet Gynecol Clin North Am 1999;26:641-656.
12. Als H, Tronick E, Lester BM, et al. The Brazelton Neonatal Behavioral Assessment Scale (BNBAS). J
Abnorm Child Psychol 1977;5:215-231.
13. Scanlon JW, Brown WU Jr, Weiss JB, et al. Neurobehavioral responses of newborn infants after maternal
epidural anesthesia. Anesthesiology 1974;40:121-128.
P.279
14. Amiel-Tison C, Barrier G, Shnider SM, et al. A new neurologic and adaptive capacity scoring system for
evaluating obstetric medications in full-term newborns. Anesthesiology 1982;56:340- 350.
15. Brockhurst NJ, Littleford JA, Halpern SH. The neurologic and adaptive capacity score: A systematic review
of its use in obstetric anesthesia research. Anesthesiology 2000;92:237-246.
16. Camann W, Brazelton TB. Use and abuse of neonatal neurobehavioral testing. Anesthesiology 2000;92:35.
17. Halpern SH, Littleford JA, Brockhurst NJ. The neurologic and adaptive capacity score is not a reliable
method of newborn evaluation. Anesthesiology 2001;94:958-962.
18. Sepkoski CM, Lester BM, Ostheimer GW, et al. The effects of maternal epidural anesthesia on neonatal
behavior during the first month. Dev Med Child Neurol 1992;34:1072-1080.
19. Richardson MG. Regional anesthesia for obstetrics. Anesthesiol Clin North America 2000;18:383-406.
20. Liston R, Crane J. Fetal health surveillance in labour, Part 1. SOGC Clinical Practice Guidelines No. 112,
March 2002. Available at: http://www.sogc.org/SOGCnet/sogc_docs/common/guide/library_e.
shtml#obstetrics Accessed May 6, 2003.
21. Liston R, Crane J. Fetal health surveillance in labour, Part 2. SOGC Clinical Practice Guidelines No. 112,
April 2002. Available at: http://www.sogc.org/SOGCnet/sogc_docs/common/guide/library_e.shtml#obstetrics
Accessed May 6, 2003.
22. Thacker SB, Stroup D, Chang M. Continuous electronic heart rate monitoring for fetal assessment during
labor. Cochrane Database Syst Rev 2001;2:CD000063.
23. Dawes G, Meir YJ, Mandruzzato GP. Computerized evaluation of fetal heart-rate patterns. J Perinat Med
1994;22:491-499.
24. Neilson JP, Mistry RT. Fetal electrocardiogram plus heart rate recording for fetal monitoring during labour.
(Cochrane Database Syst Rev 2000;2:CD000116.
25. Norn H, Ameer-Whlin I, Hagberg H, et al. Fetal electrocardiography in labor and neonatal outcome: data
from the Swedish randomized controlled trial on intrapartum fetal monitoring. Am J Obstet Gynecol
2003;188:183-192.
26. Solt I, Ganadry S, Weiner Z. The effect of meperidine and promethazine on fetal heart rate indices during
the active phase of labor. Isr Med Assoc J 2002;4:178-180.
27. Hoffman CT 3rd, Guzman ER, Richardson MJ, et al. Effects of narcotic and non-narcotic continuous
epidural anesthesia on intrapartum fetal heart rate tracings as measured by computer analysis. J Matern Fetal
Med 1997;6:200-205.
28. St Amant MS, Koffel B, Malinow AM. The effects of epidural opioids on fetal heart rate variability when
coadministered with 0.25% bupivacaine for labor analgesia. Am J Perinatol 1998;15: 351-356.
29. Manning FA. Fetal biophysical profile: a critical appraisal. Clin Obstet Gynecol 2002;45:975-985.
30. Manning FA, Snijders R, Harman CR, et al. Fetal biophysical profile score. VI. Correlation with antepartum
umbilical venous fetal pH. Am J Obstet Gynecol 1993;169:755-763.
31. Farrell T, Owen P, Harrold A. Fetal movements following intrapartum maternal opiate administration. Clin
Exp Obstet Gynecol 1996;23:144-146.
32. Kopecky EA, Ryan ML, Barrett JF, et al. Fetal response to maternally administered morphine. Am J Obstet
Gynecol 2000;183: 424-430.
33. Smith CV, Rayburn WF, Allen KV, et al. Influence of intravenous fentanyl on fetal biophysical parameters
during labor. J Matern Fetal Med 1996;5:89-92
34. Dildy GA. Fetal pulse oximetry: current issues. J Perinat Med 2001;29:5-13.
35. East CE, Colditz PB, Begg LM, et al. Update on intrapartum fetal pulse oximetry. Aust N Z J Obstet
Gynaecol 2002;42:119-124.
36. Carbonne B, Langer B, Goffinet F, et al. Multicenter study on the clinical value of fetal pulse oximetry. II.
Compared predictive values of pulse oximetry and fetal blood analysis. Am J Obstet Gynecol 1997;177:593598.
37. Paternoster DM, Micaglio M, Tambuscio B, et al. The effects of epidural analgesia and uterine contractions
on fetal oxygen saturation during the first stage of labour. Int J Obstet Anesth 2001;10:103-107.
38. East CE, Colditz PB. Effect of maternal epidural analgesia on fetal intrapartum oxygen saturation. Am J
Perinatol 2002;19: 119-126.
39. Harman CR, Baschat AA. Comprehensive assessment of fetal wellbeing: which Doppler tests should be
performed? Curr Opin Obstet Gynecol 2003;15:147-157.
40. Ramos-Santos E, Devoe LD, Wakefield ML, et al. The effects of epidural anesthesia on the Doppler
velocimetry of umbilical and uterine arteries in normal and hypertensive patients during active term labor.
41. Erkinjuntti M, Vaahtoranta K, Alihanka J, et al. Use of the SCSB method for monitoring of respiration, body
movements and ballistocardiogram in infants. Early Hum Dev 1984;9:119-126.
42. Nikkola EM, Kirjavainen TT, Ekblad UU, et al. Postnatal adaptation after caesarean section or vaginal
delivery, studied with the static-charge-sensitive bed. Acta Paediatr 2002;91:927-933.
43. Nikkola EM, Jahnukainen TJ, Ekblad UU, et al. Neonatal monitoring after maternal fentanyl analgesia in
labor. J Clin Monit Comput 2000;16:597-608.
44. May AE, Elton CD. The effects of pain and its management on mother and fetus. Baillieres Clin Obstet
Gynaecol 1998;12:423- 441.
45. Brownridge P. The nature and consequences of childbirth pain. Eur J Obstet Gynecol Reprod Biol 1995;59
Suppl:S9-S15.
46. Neumark J, Hammerle AF, Biegelmayer C. Effects of epidural analgesia on plasma catecholamines and
cortisol in parturition. Acta Anaesthesiol Scand 1985;29:555-559.
47. Lowe NK. The nature of labor pain. Am J Obstet Gynecol 2002;186(5 Suppl):S16-S24.
48. Kubli M, Scrutton MJ, Seed PT, et al. An evaluation of isotonic sport drinks during labor. Anesth Analg
2002;94:404-408.
49. Bhutta AT, Garg S, Rovnaghi CR. Fetal response to intra-uterine needling: is it pain? Does it matter?
50. Glover V, Fisk NM. Fetal pain: implications for research and practice. Br J Obstet Gynaecol 1999;106:881886.
51. Gitau R, Fisk NM, Teixeira JM, et al. Fetal hypothalamic-pituitary-adrenal stress responses to invasive
procedures are independent of maternal responses. J Clin Endocrinol Metab 2001;86:104- 109.
52. Fisk NM, Gitau R, Teixeira JM. Effect of direct fetal opioid analgesia on fetal hormonal and hemodynamic
stress response to intrauterine needling. Anesthesiology 2001;95:828-835.
53. Anand KJ, Maze M. Fetuses, fentanyl, and the stress response: signals from the beginnings of pain?
Anesthesiology 2001;95:823- 825.
54. Rowlands S, Permezel M. Physiology of pain in labour. Baillieres Clin Obstet Gynaecol 1998;12:347-362.
55. Kitzinger S. Natural childbirth is inappropriate in a modern world. Int J Obstet Anesth 2002;11:30-32.
56. American Society of Anesthesiologists Task Force on Obstetrical Anesthesia. Practice guidelines for
obstetrical anesthesia: a report. Anesthesiology 1999;90:600-611.
57. American College of Obstetrics and Gynecology. Obstetric analgesia and anesthesia. Number 36, July
2002. Int J Gynaecol Obstet 2002;78:321-335.
58. Marmor TR, Krol DM. Labor pain management in the United States: understanding patterns and the issue
of choice. Am J Obstet Gynecol 2002;186(5 Suppl):S173-S180.
59. Simkin PP, O'Hara M. Nonpharmacologic relief of pain during labor: systematic reviews of five methods.
Am J Obstet Gynecol 2002;186:S1
Chapter 17
Cardiorespiratory Adjustments at Birth
Ruben E. Alvaro
Henrique Rigatto
Respiratory physiologists and physicians have long been interested in the respiratory and cardiovascular events
that occur at birth. However, apart from occasional references to the pulmonary circulation, the fetal circulation
only received serious consideration in the middle of the twentieth century, when it was recognized that dramatic
changes in blood flow through the lungs occurred after birth.
The first detailed description of circulation in the mammalian fetus was provided by Harvey in 1628 (1).
Although he correctly described blood flow from the inferior vena cava through the foramen ovale, he thought
that the blood had to enter the pulmonary veins before returning to the left atrium. He was also perplexed by
how the fetus survives in utero without the aid of respiration. The answer to the last question came in 1799,
when Scheel noted light red blood in the umbilical vein and dark red blood in the umbilical artery in the fetal
sheep, as well as darkening of that color when the pregnant ewe was asphyxiated (2). However, it was Zweifel,
in 1876, who categorically stated that the placenta was the lung of the fetus, describing the presence of
oxyhemoglobin in the umbilical blood before any breathing had occurred (3).
The conventional belief during the nineteenth century was that the fetal pulmonary blood flow progressively
increased over gestation and that it was relatively higher in the fetus than after birth (4,5). It was not until the
first part of the twentieth century that the right ventricular pressure was demonstrated to fall and pulmonary
blood flow to increase after the establishment of breathing (6,7). It was only 50 years ago that Dawes and
colleagues (8) demonstrated by direct measurements in fetal lambs that pulmonary blood flow increased when
the lungs were ventilated with air. Over these past 50 years, the developmental changes in the pulmonary
circulation and in its responses to stresses of hypoxia, and increases in pulmonary arterial pressure and blood
flow, have become subjects of intense investigation (9).
It is well known now that during fetal life the placenta and not the lungs serves as the organ for gas exchange.
Because of this, the normal fetal circulatory pattern is arranged very differently from that observed after birth
and is quite satisfactory for survival in the womb. The placental circulation is in a parallel arrangement; that is,
it receives blood from the descending aorta and drains blood to the systemic venous circulation. To accomplish
this, the fetal circulation depends on a series of intra- and extracardiac shunts that allows the oxygenated blood
to flow from the placenta to the systemic organs and for the deoxygenated blood to return to the placenta. This
blood flow distribution allows the delivery of blood with the highest oxygen content (from the placenta) to the
heart and brain and blood with lower oxygen content to the lower body and placenta. Because the lungs are not
required for gas exchange, pulmonary blood flow is low (approximately 10% of the combined ventricular
output), yet adequate for lung growth and development. Unique only to fetal life, the blood exiting the lungs has
a lower saturation than does blood entering the pulmonary circulation, and although the lungs do not participate
in gas exchange in utero, they are metabolically active, secreting liquid into the potential air spaces and
synthesizing surfactant, a substance that is vital to achieve adequate ventilation at birth. The fetal lungs are
also physically active in that they simulate breathing movements.
The transition from the placenta to the lungs at birth is accomplished by three main cardiopulmonary processes:
(a) onset of breathing, resulting in lung expansion with concomitant decrease in pulmonary vascular resistance
and increase pulmonary blood flow; (b) increase in blood oxygen content that further decreases pulmonary
vascular resistance; and (c) loss of the placental circulation with resultant increase in systemic vascular
resistance leading to the closure of the fetal cardiovascular shunts and transition from fetal to neonatal
circulation. Thus, to establish the lungs as the site of gas exchange after birth, significant
P.285
changes in the cardiac and pulmonary circulation as well as the initiation of pulmonary ventilation must occur.
Many abnormal maternal, placental and fetal conditions may interfere with this physiologic transition and
compromise the newborn infant.
The establishment of effective pulmonary ventilation at birth requires that the lungs develop to a stage where
the alveoli can be inflated to provide adequate gas exchange. It also requires the lowering of the pulmonary
vascular resistance to allow for the increase in pulmonary blood flow to accommodate the entire cardiac output.
The successful transition also requires that the lung liquid volume be removed from the alveolar spaces and that
surfactant material be secreted into the acinus to allow for satisfactory physical expansion of the lungs after the
initial postnatal breaths. Adequate neurologic drive to generate and maintain spontaneous continuous breathing
is essential to maintain ventilation postnatally.
The change to pulmonary ventilation at birth increases pulmonary venous oxygenation that serves to suppress
active vasoconstriction of pulmonary vessels. The result is a tenfold increase in pulmonary blood flow and a
rapid decrease in pulmonary vascular resistance. The removal of the placenta after constriction of the umbilical
vessels in response to increased oxygenation, contributes to a rise in systemic vascular resistance. The decline
in pulmonary vascular resistance below systemic values contributes to the closure of the foramen ovale and
ductus arteriosus, establishing the adult circulatory patterns.
This chapter reviews some of the most important cardiorespiratory adjustments that occur at the time of
delivery allowing the fetus to achieve a successful extrauterine transition.
PULMONARY ADAPTATION
Fetal Lung Fluid
During fetal life, the internal volume of the lungs is maintained by the secretion of liquid into the pulmonary
lumen. This liquid expansion of potential air spaces is essential for the growth and the development of normal
lung structure before birth, which, in turn, may influence lung function after birth (10).
TABLE 17-1 COMPOSITION OF LUNG LUMINAL LIQUID, AMNIOTIC LIQUID, AND PLASMA OF FETAL
LAMBS
Lung Liquid
Amniotic Liquid
Plasma
6.27 0.01
300 6
7.07 0.22
257 14
7.34 0.04
291 8
Na+ (mEq/L)
150 1
113 6
150 1
CI- (mEq/L)
157 4
87 5
107 1
31
19 3
24 1
pH
Osmolality
HCO-3 (mEq/L)
Protein (g/dL)
0.03 0.01
0.1 0.1
4.1 0.3
Values are mean SEM (standard error of mean) and are taken from the work of Adamson et al.,
Adams et al., and Humphreys et al.
The fluid in the fetal lung was for many years assumed to be aspirated amniotic fluid as a result of fetal
breathing movements (11). In 1941, Potter and Bohlender (12) observed alveolar fluid in two human fetuses
with malformations of the respiratory tract which blocked the entrance of amniotic fluid, thus establishing that
the lung fluid was secreted, not inhaled. Experiments performed in other species confirmed that fetal pulmonary
fluid was indeed generated within the lungs (13,14,15).
We know now that the fetal lung fluid is neither a mere ultrafiltrate of plasma nor aspirated amniotic fluid.
Compared to plasma this lung fluid is rich in chloride and potassium, is significantly lower in bicarbonate and has
similar sodium concentration. It is also quite different from amniotic fluid having much higher osmolality, Na+
and Cl- concentrations, and significantly lower K+, protein and urea concentration (Table 17-1) (10,16,17,18).
This distinctive composition of the lung liquid changes very little during gestation (17,19). The high Cl- and the
low protein content characteristics of the lung fluid result from active Cl- secretion and tight junctions between
epithelial cells respectively.
It is not known exactly when this secretory activity begins, but already during the glandular stage of lung
development, at about 3 months of gestation, the lung epithelium actively secretes fluid (20,21). By this time
the epithelium has developed tight junctions, which are evident by morphologic examination and also by the low
protein concentration present in the lung liquid in relationship to the plasma. In fetal lambs, the volume of lung
liquid increases from about 5 mL/kg of body weight at midgestation (18) to about 30 to 50 mL/kg at term
(18,22,23,24,25). The secretion rate increases from about 2 mL/ kg body weight at midgestation (18) to about
5 mL/kg at term (26,27). More recently, Pfister and associates (28) demonstrated that the lung liquid volume
exhibited a plateau level in the near-term fetal sheep before it began to decline toward birth. They also
observed that the rate of lung liquid secretion declined in two linear phases that commenced earlier than the
changes in lung liquid volume (28). In fetal lambs, the amount of lung liquid in relation to lung weight remains
P.286
relatively constant at approximately 90% through much of the pregnancy (29,30). The lung liquid secretion
decreases with increased luminal hydrostatic pressure induced by a prolonged obstruction of the fetal trachea
and increases when luminal pressure falls below amniotic fluid pressure and when fetal breathing movements
are abolished (24).
This liquid secreted by the fetal lungs flows intermittently up the trachea with fetal breathing movements. Some
of this fluid is swallowed and the remainder contributes directly to the formation of amniotic fluid production,
accounting for approximately 25% to 50% of the amniotic fluid turnover in the sheep fetus, with the rest being
formed by the fetal urine (10). The mechanism by which amniotic fluid is not aspirated into the lungs was
demonstrated by Brown and associates (31) in 1983, when they showed that the larynx acts as a one-way valve
allowing only liquid outflow under normal circumstances. The continuous secretion of liquid by the lungs
confronted with a flow impediment produced by the larynx and the amniotic fluid pressure creates a small but
important positive intrapulmonary pressure, which is essential for normal growth and for the structural and
biochemical maturation of the developing lung (10,29,31,32). Thus, in fetal sheep, unimpeded leakage of
tracheal liquid decreases lung size by arresting pulmonary tissue growth, whereas prolonged obstruction of
tracheal outflow leads to lung hyperplasia (32,33). Nardo and associates (34) showed that lung hypoplasia in
fetal sheep can be considerably improved by short-term obstruction at the tracheal level. Conversely, pulmonary
hypoplasia in humans can be observed in pathologic conditions such as diaphragmatic hernia, pleural effusion,
or severe oligohydramnios (Potter syndrome) as a result of the compression of the fetal lungs and the decrease
in their internal volume (35,36).
Congenital high airway obstruction syndrome (CHAOS) is a clinical condition caused by complete or nearcomplete obstruction of the fetal airway that results in elevated intratracheal pressure, distention of the
tracheobronchial tree, and lung hyperplasia. The enlarged lungs may cause cardiac and caval compression
leading to in utero heart failure manifested by ascites, hydrops fetalis, and placentomegaly (37,38).
The production of fetal lung liquid depends on a system of active ion transport across the alveolar type II cells of
the pulmonary epithelium (10,39,40). Olver and Strang (41) demonstrated that lung liquid secretion is coupled
with active transport of Cl- toward the pulmonary lumen, generating an electrical potential difference of -5 mV
(lumen negative). This chloride secretion generates an osmotic gradient that causes liquid to flow from the
microcirculation through the interstitium into the potential air spaces. This chloride secretion occurs through
chloride channels in the apical membrane (alveolar side) and depends largely on chloride influx at a bumetanidesensitive Na+-K+-2Cl- (NKCC) cotransporter system in the basolateral membrane (interstitial side) (42). Thus,
Cl- enters the cell on a cotransporter linked with K+ and Na+ down the electrochemical potential gradient for Na
+
generated by Na+-K+-adenosine triphosphatase (ATPase) in the basolateral membrane of the cell.
Consequently, Cl- concentration increases inside the cell above its equilibrium potential, which provides an
electrochemical gradient for Cl- exit across the luminal membrane of the epithelial cell through Cl- permeant ion
channels (Fig. 17-1) (10). Addition of the loop diuretic bumetanide or furosemide (specific NKCC inhibitors) into
the fetal lung liquid decreases fluid secretion by decreasing Cl- entry into the epithelial cell through the
basolateral membrane.
Figure 17-1 Lung liquid secretion. The present model for chloride secretion emphasizes the central role of
basolateral Na-K-ATPase in maintaining high intracellular K- and low NA+ concentrations. Na- re-enters the cell
down its electrochemical gradient via the Na/K/2CL cotransporter on the basolateral membrane. K+ may leave
the cell through basolateral channels. The high intracellular Cl- concentration and the negative intracellular
membrane potential facilitate the rise of the Cl- concentration to tally its electrochemical equilibrium. A further
rise in intracellular Cl- concentration will lead to opening of the apical Cl- channels and flow of this anion into
the lung lumen. Na+ and water will follow Cl- through a paracellular route. (Adapted from O'Brodovich HM.
Immature epithelial Na+ channel expression is one of the pathogenetic mechanisms leading to human neonatal
respiratory distress syndrome. Proc Assoc Am Physicians 1996;108:345-355.)
Clearance of Fluid at Birth

To allow adequate postnatal gas exchange, pulmonary fluid, which is essential to fetal lung development, must
be rapidly removed at birth in order (43,44,45). Thus the transition from intra- to extrauterine life requires the
effective clearance of lung liquid to support air breathing and the conversion of the pulmonary epithelium in the
distal air spaces from fluid secretion to fluid absorption. Disruption of this process has been implicated in several
disease states, including transient tachypnea of the newborn (TTN) and hyaline membrane disease
(44,46,47,48). Preterm delivery and cesarean section without prior labor result in excessive retention of lung
fluid and may contribute to respiratory compromise in the newborn infant (47,49,50,51,52,53).
A complete understanding of the mechanisms by which fetal lungs are able to clear themselves of fluid is still
lacking. It is clear now that the traditional explanations for fluid reabsorption which relied on Starling forces,
lymphatic uptake and vaginal squeeze at the time of birth, can
P.287
only account for a very small fraction of the lung fluid clearance (54). More recently, several studies have
strongly suggested that the primary mechanism of lung liquid reabsorption is the change in ion transport
induced by catecholamines during labor (10,31,44,55,56,57). Although the mechanisms responsible for lung
liquid clearance at birth develop gradually during the last part of pregnancy, the removal of lung liquid and the
switch from fluid secretion to fluid absorption, triggered by events at birth, are probably quite rapid because in
the human the functional residual capacity of the lungs rises to a near normal value of 25 to 30 mL/kg within 15
minutes after normal delivery (58).
Although there is disagreement about whether clearance begins prior to labor, or occurs entirely within labor, it
is now clear that the vast bulk (>75%) of this liquid leaves the lung some time before normal term birth (59). In
studies done in full-term fetal sheep, Pfister and associates (28) showed that the lung liquid volume began a
gradual decline well before the onset of labor and it was mainly a result of a decrease in the secretion rate
rather than active reabsorption. They also showed that this gradual decline was followed by a much more rapid
decline within labor that was mainly a result of active reabsorption of fluid driven by active Na+ transport
through the pulmonary epithelium. This last process is stimulated by the catecholamine surge that occurs just
before the onset of labor (31). Other authors, however, suggest that lung liquid volume continues to rise until
the day before labor (24,25,60).
As explained in the previous section, chloride secretion across the distal lung epithelium results in the
production of lung liquid which is necessary for proper lung development in fetal life. In contrast, sodium
absorption allows for fluid reabsorption and is critical for efficient oxygenation in the newborn.
The movement of sodium across the pulmonary epithelium from the alveolar lumen to the interstitium with
subsequent absorption into the vasculature can be considered a two-step process. In the first step, sodium
passively enters the apical membrane of the alveolar type II cell through amiloride-sensitive epithelial Na+
channels (ENaCs). Thus, intraluminal instillation of amiloride delays lung fluid clearance in fetal life
(61,62,63,64,65,66,67) and induces severe respiratory distress and a persistence of fetal lung fluid postnatally
(68). In the second step, sodium is actively pumped out of the cell into the interstitium through the basolateral
membrane by the ouabain inhibitable Na+-K+-ATPase. Thus, Na+-K+-ATPase pump inhibition with ouabain
consistently reduces liquid clearance in various species (62,66,69,70,71,72). To equilibrate the osmotic
pressure, generated by the movement of Na+, water diffuses from the alveolar to the interstitial space either
through specific water channels (aquaporins) or through the paracellular junctions (Fig. 17-2) (44,73,74).
Although recent data demonstrate that aquaporin-4 messenger RNA (mRNA) in the perinatal epithelium is
developmentally regulated and peaks at birth (74,75) there is no evidence that these channels regulate fluid
clearance (Fig. 17-2) (10,44,55,72,73,76,77,78). The mechanisms of Na+ absorption through the pulmonary
epithelium have been confirmed by isolation and culture of fetal alveolar type II cells (68,79,80,81,82).
Figure 17-2 Lung liquid absorption. At birth, to offset Cl- secretion Na+ channels on the epithelial cell are
activated, establishing conduit for the osmotically active Na+ ion. Basolateral membrane Na+-K+-ATPase
generates the gradient for apical Na+ entry into the epithelial cell and switches the distal lung epithelium from
a secreting to an absorbing organ. The net movement of Na+ to the basolateral direction induces a parallel
movement of Cl- and fluid through the paracellular route or the transcellular pathway, using specific water
channels, that is, aquaporins. Amil insens, amiloride insesitivity. (Adapted from O'Brodovich HM. Immature
epithelial Na+ channel expression is one of the pathogenetic mechanisms leading to human neonatal
respiratory distress syndrome. Proc Assoc Am Physicians 1996;108:345-355.)
The gene for the amiloride-sensitive epithelial sodium channel has been cloned; it consists of three homologous
subunits called -, -, and -ENaC. The subunit is a prerequisite unit for any channel activity. The expression
of the three subunits is necessary for maximum ion transport (83). The -ENaC (84) and the - and -ENaC
subunits (85) were detected in early midtrimester on the apical domain of human airway cells suggesting that
sodium absorption might begin significantly before birth, even if secretion is still dominant. The three subunits
increased sharply around the time of birth, were highest shortly after birth, and declined in parallel with
endogenous plasma epinephrine concentration during the first week of life. The importance of the ENaC was
confirmed by studies showing that -ENaC knockout mice were unable to clear lung liquid from the alveolar
spaces after birth resulting in death from respiratory failure (86). Mice with - or -ENaC null mutations had
delayed liquid clearance but had near-normal lung water content 12 hours after birth (86,87).
It is well known now that net alveolar fluid clearance occurs at a rapid rate late in gestation and that this
clearance is driven by elevations of endogenous epinephrine. The critical link between -adrenergic stimulation
and lung fluid clearance was made in 1978 by Walters and Olver who found that intravenously infused
epinephrine caused rapid absorption of lung fluid in near-term fetal lambs and that this response could be
inhibited by prior treatment with propanolol (88). The intravenous epinephrine caused an
P.288
immediate and reversible increase in luminal electronegativity by stimulating the active transport of Na+ out of
the lung lumen (55). In the absence of adrenergic stimulation the amiloride-sensitive Na+ channels remain
closed (secretory state). Thus, the opened or closed state of the Na+ channels determines whether the fetal
lungs, at any particular time, are secretory or reabsorptive (10). The epinephrine stimulation of amiloridesensitive alveolar fluid clearance is mediated by cyclic adenosine monophosphate (cAMP) and Ca2+ likely acting
as a intracellular second messenger (89,90).
The induced reabsorption with epinephrine increases strikingly with advancing gestational age and can be
induced by pretreatment of fetal sheep with the combination of corticosteroids and triiodothyronine (91). These
two hormones are required to switch the effect of -adrenergic stimulation from net chloride and liquid secretion
to net sodium and liquid absorption (30,56,57,92,93). Consistent with in vivo observations, -ENaC mRNA is
increased by glucocorticoids in the primary cultures of alveolar type II cells FDLE, while the expression of - and
-ENaC mRNA is unaffected by glucocorticoid (94).
Another factor that appears to be particularly important in the switch of transepithelial liquid flow from secretion
to absorption is the sharp increase in alveolar pressure of oxygen (PAO2) that occurs at birth. It was recently
shown that fetal PO2 favored the development of fluid filled cyst-like structures while the rise of PO2 to postnatal
levels reduced fluid in the cysts (95). The effect of oxygen seems to be the result of an increase in Na+-K+ATPase activity and the response is also enhanced by glucocorticoids and thyroid hormones. Thus, at birth,
epinephrine, oxygen, glucocorticoid and thyroid hormones interact to produce a permanent switch from
secretion to absorption in the distal epithelium.
Although the switch from Cl- secretion to sodium absorption is the critical mechanism for lung fluid clearance at
the time of delivery, other passive factors play a role in clearing the residual liquid present in potential air
spaces at birth. The first is transient and peaks early and relates to enlarged transepithelial pores produced by
spontaneous ventilation (22,29,47,51). Thus air inflation, by passive reabsorption down the transpulmonary
pressure gradient associated with ventilation, shifts residual liquid from the lung lumen into the interstitium
around distensible perivascular spaces of large pulmonary blood vessels and airways. These perivascular cuffs
progressively diminish in size as the fluid is removed by small pulmonary blood vessels and lymphatics. Bland
and associates (96,97) showed that the pulmonary circulation absorbs most of the residual liquid present in
potential air spaces at birth and that elevated left atrial pressure or reduction of plasma protein concentration
slows the rate of liquid clearance in mature animals. Although there is a small and transient increase in lymph
flow after postnatal breathing begins, the amount of excess liquid drained postnatally by pulmonary lymphatics
is only approximately 11% of the residual liquid in the lungs at birth (98). The second mechanism is the
transvascular protein gradient that facilitates the movement of fluid from the essentially protein-free lung fluid
into the interstitium, followed by passage of liquid into the bloodstream.
Fetal Breathing
The discovery of fetal breathing in the late 1960s immediately stimulated interest in the factors that control
breathing in utero (99,100,101). Although of unknown purpose, because no gas exchange is involved, fetal
breathing may represent preparation in utero for a vital function important in life. Shortly after its discovery, the
Oxford group (102) showed that, although fetal breathing was influenced by fetal behavior, occurring essentially
in rapid eye movement (REM) sleep, it was clearly regulated by other chemical factors, such as carbon dioxide
and oxygen concentration. Subsequent work confirmed and expanded these findings by recording the electrical
activity of the diaphragm and clearly demonstrating the central origin of the respiratory output in utero
(103,104,105,106,107). Using ultrasonographic technology, breathing movements were also identified in the
human fetus, being present approximately 40% of the time during late pregnancy, a figure similar to that in
sheep (108,109,110,111).
The discovery of fetal breathing not only stimulated the development of the area of fetal assessment but it also
brought a new dimension to the events occurring at birth. What has been traditionally called the initiation of
breathing at birth must now be called the establishment of continuous breathing at birth. Breathing begins
long before birth. The question is not what determines the appearance of breathing at birth, but what makes it
continuous. From another angle, what makes fetal breathing episodic in late gestation and present only during
low-voltage electrocortical activity? The answer to this question remains unknown.
Fetal breathing in sheep is mostly continuous in early gestation (90 to 115 days) but becomes episodic in late
gestation, primarily occurring during periods of low-voltage electrocortical activity (99,101,107,112,113). During
high-voltage electrocortical activity there is no established breathing present, but occasional breaths may
surface after episodic, generalized, tonic muscular discharges associated with body movements (Fig. 17-3)
(107). During low-voltage electrocortical activity, breathing is irregular, the diaphragmatic electromyelogram
(EMG) being characterized by an abrupt beginning and end. The physiologic mechanism responsible for the
occurrence of fetal breathing only during low-voltage electrocortical activity is unknown.
Modulation of Fetal Breathing by CO2, O2, and Neurochemicals

Many studies clearly show that the fetal breathing apparatus is capable of responding well to chemical stimuli
and other agents known to modify breathing postnatally. Thus it
P.289
became clear that the fetus responds to an increase in arterial carbon dioxide pressure (PaCO2) with an increase
in breathing (102,108,114,115,116,117,118). The increased breathing activity is prolonged into the transitional
high voltage electrocochleography but does not continue into the established (Fig. 17-4) (118). There is much
evidence suggesting that the actions of CO2 are central.
Figure 17-3 Fetal breathing in a fetal lamb at 134 days of gestation. The deflections in tracheal pressure and
diaphragmatic activity occur during periods of raid eye movement (REM) in low-voltage electrocortical activity
only. In high-voltage electrocortical activity (quiet sleep) breathing is absent. (From Rigatto H. Regulation of
fetal breathing. Reprod Fertil Dev 1996;8:23-33, with permission.)
Administration of low oxygen to the fetus by having the ewe breathe hypoxic mixtures abolished fetal breathing;
this was associated with a decrease in body movements and in the amplitude of the ECoG (102,119,120).
Transection of the brain at the upper level of the pons prevents the inhibitory action of hypoxia and induces
continuous breathing. Conversely, increase in arterial PO2 to levels above 200 mm Hg through the
administration of 100% O2 to the fetus via an endotracheal tube stimulated breathing and induced continuous
breathing in 35% of the experiments in fetal sheep (121). These findings suggest that low partial tension of O2
in the fetus at rest may be a normal mechanism inhibiting breathing in utero.
A comprehensive review of the effects of various neurochemicals agents on fetal breathing has been published
(122). In general, fetal breathing is inhibited by muscimol (123), a -aminobutyric (GABA) agonist,
pentobarbiton (124), and diazepam (125) acting at the GABAA receptor complex (126). Other endogenous
inhibitors include adenosine, prostaglandin E2 (PGE2) (see next subsection), and endorphins, which may be
present at high levels in the fetal circulation. Agents that stimulate fetal breathing movements include
prostaglandin inhibitors such as indomethacin, meclofenamate, morphine, caffeine, pilocarpine, and 5-hydroxy-ltryptophan (5-HTP). Fetal breathing is inhibited by maternal hypoglycemia whereas it is stimulated after
maternal meals and hyperglycemia.
Figure 17-4 Fetal breathing during control and during CO2 rebreathing. Note the increase in tracheal pressure
and diaphragmatic activity during CO2 rebreathing. Fetal breathing was prolonged into the transitional low- to
high-voltage ECoG, but stopped in established high-voltage ECoG. (From Rigatto H. Regulation of fetal
breathing. Reprod Fertil Dev 1996;8:23-33, with permission.)
Establishment of Continuous Breathing at Birth

The traditional view has been that labor and delivery produce a transient fetal asphyxia that stimulates the
peripheral chemoreceptors to induce the first breath. Breathing would then be maintained through the input of
other stimuli, such as cold or touch (127,128). More recent observations have questioned this general view.
First, the denervation of the carotid and aortic chemoreceptors does not alter fetal breathing or the initiation of
continuous breathing at birth (Fig. 17-5) (129,130). Second, continuous breathing can be established in utero,
with manifestations of arousal, by raising fetal PO2 and occluding the
P.290
umbilical cord (Fig. 17-6) (121). These observations during administration of high O2 or cord occlusion suggest
that the fetus can be made to resemble a neonate in utero without the transient hypoxemia to stimulate the
peripheral chemoreceptors and without any of the sensory stimuli, such as cold, once thought to be important
for the establishment of continuous breathing at birth.
Figure 17-5 Delay in minutes from opening of the window to () the appearance of the first breath and to
() sustained breathing. There were no significant differences between the sham-operated and the
chemodenervated groups. (From Rigatto H. Regulation of fetal breathing. Reprod Fertil Dev 1996;8:23-33, with
permission.)
Figure 17-6 Representative tracing showing the effect of fetal fraction of inspired oxygen (FiO2) on fetal
breathing and electrocortical activity. A: Control cycle showing little breathing in fetus in early labor at 143
days of gestation. B: Lung distention (mean airway pressure 30 cm H2O) and inspired N2 does not affect
baseline tracing. C: Seventeen percent O2 also does not alter breathing. D: One hundred percent O2 induces
continuous breathing. E: Occlusion on two occasions induces more forceful breathing than that observed with
O2 alone. Note that continuous breathing was elicited despite preventing the rise of PaCO2 by ventilating the
fetus with high-frequency ventilation (15 Hz, stroke = 7 cm H2O). ECoG, electrocochleography; EMGdi,
electromyogram of the diaphragm. (From Rigatto H. Regulation of fetal breathing. Reprod Fertil Dev 1996;8:2333, with permission.)
Thus, the physiologic mechanism responsible for the inhibition of fetal breathing and the establishment of
continuous breathing at birth remains unknown. It has been debated whether the key factors in inducing these
changes are intrinsic to the fetal brain or are in the placenta. Because placental separation at birth is associated
with the onset of continuous breathing, we, together with others, have hypothesized that placental factors might
be responsible for the inhibition of fetal breathing (131,132,133,134,135). This line of thinking is based on the
assumption that the release of a factor by the placenta into the fetal circulation prevents fetal breathing from
being continuous, with inhibition during high-voltage ECoG, and present only during periods of reticular
activation as it occurs during low-voltage ECoG. In the absence of this factor from the placenta at birth, after
cord clamping, the state-related inhibition observed during high-voltage ECoG is insufficient to disrupt
continuous breathing. Teleologically, it is interesting that nature may have delegated to the placenta the
important role of providing the fetus with gas exchange and nutrients and it is conceivable that it may also have
endowed the placenta with some form of chemoreceptor
P.291
activity regulating fetal breathing and behavior by the secretion of chemical substances into the fetal circulation.
More direct evidence for a placental role has been present since Dawes (127) and Harned and Ferreiro (128)
showed that only after clamping the umbilical cord does the newborn lamb start breathing and behaving like a
neonate. Subsequently, Adamson and associates (133) induced breathing in the fetus with umbilical cord
occlusion and supply of O2 via an endotracheal catheter. On release of the cord, breathing ceased immediately,
before any change in blood gases or pH, suggesting that a factor from the placenta might be involved. In our
laboratory, we were able to induce continuous breathing and wakefulness in fetal sheep by occluding the
umbilical cord, as long as we provided a gas exchange area for the fetus via an endotracheal tube
(132,136,137,138). These experiments suggest the origin in the placenta of a compound that inhibits fetal
breathing and fetal activity.
In trying to prove the hypothesis that a factor is released by the placenta, we injected the fetal sheep with a
placental extract (juice of cotyledons acutely dissected, sliced, and immersed in Krebs solution) after continuous
breathing was induced by cord occlusion (Fig. 17-7) (132). In all experiments the placental extract decreased or
abolished breathing. The infusion of the placental extract into the fetal circulation also inhibited spontaneous
fetal breathing present during low-voltage electrocortical activity without inducing significant changes in blood
gas tensions, pH, heart rate, and blood pressure (139). This factor appeared specific to the placenta because the
breathing response was absent with extracts from other tissues, such as liver, muscle, or blood. We have
demonstrated that this factor in the placental extract is likely a prostaglandin, because treatment of the extract
with indomethacin/acetylsalicylic acid (ASA), which significantly reduced the concentration of prostaglandins,
eliminated the activity of the extract (Fig. 17-8) (140).
Indirect evidence that placental prostaglandins, especially PGE2, are the mediators responsible for the inhibition
of breathing in fetal life has been provided by Kitterman and associates (141) and Wallen and associates (142)
who showed that infusion of PGE2 into the circulation of the fetal sheep induced a prompt and complete
cessation of breathing movements. In addition, the incidence of fetal breathing movements was inversely
correlated with both the PGE2 dose and the mean PGE2 concentration. Conversely, intravenous infusion of
prostaglandin synthetase inhibitors, such as indomethacin or meclofenamate, induces continuous breathing for
many hours in the fetus (141,143,144). Thus, the rate of placental prostaglandin production, plays a significant
role in setting the level of fetal breathing activity by producing a sleep-related inhibition in the fetal brainstem.
Figure 17-7 Representative tracing showing the effects of different placental infusates on fetal breathing and
ECoG. The whole placental extract and the subfraction 3-5 to 10 kDa decreased or abolished breathing in all
cases, and when given in low-voltage ECoG, this was associated with a switch to high voltage. No significant
effects were seen with infusates having molecular mass greater than 10 kDa or less than 1 kDa. ECoG,
electrocochleography; EMGdi, electromyogram of the diaphragm. (From Rigatto H. Regulation of fetal
breathing. Reprod Fertil Dev 1996;8:23-33, with permission.)
It is unlikely, however, that prostaglandins are involved in the inhibition of fetal breathing observed during
hypoxia, because this inhibition persists after the administration of prostaglandin inhibitors. Several studies
show that adenosine is the likely mediator of the respiratory depression observed during hypoxia because
intravascular administration of adenosine inhibits fetal breathing and eye movements (145) and the infusion of
adenosine receptor antagonists blunts this inhibition (145,146). Also, brain disruptions that eliminate hypoxic
inhibition of breathing also abolish the depressant effects of adenosine (147). Koos and associates (148) have
shown that hypoxia inhibits fetal breathing through activation of central adenosine receptors, specially the A(2A)
subtype.
Mechanics of the First Postnatal Breaths

Although the neural and chemical control mechanisms responsible for establishment of continuous breathing at
birth are not completely understood, the mechanical processes responsible for inflating the lungs at birth are in
some detail accepted.
It is well known that a sufficient force (opening pressure) must be generated across the lungs with the first
inspiration to overcome the viscosity of the fluid in the
P.292
airways, as well as the forces of surface tension and tissue resistance (149). As we described in the subsection
Clearance of Fluid at Birth in preparation for lung ventilation at birth, the volume of pulmonary fluid in the
fetus decreases during late gestation, especially during initiation of labor, as a result of reduction in the rate of
secretion and an increase in reabsorption of fluid in response to catecholamines. Pfister and associates (28)
found large negative intrapulmonary pressure in fetal sheep during labor, which implied that the lung volume
was less than functional residual capacity. The findings of a negative intrapulmonary pressure near the end of
labor may in part explain reports that the first inspiration in human infants does not require diaphragmatic
contraction. The elastic recoil of the chest wall after delivery would tend to rebound to the resting position
causing a small passive inspiration of air.
Figure 17-8 Incidence of fetal breathing movements during the infusion of placental extracts. The regular
placental extracts () induced a profound inhibition of fetal breathing that progressively recovered upon
discontinuation of the infusions. This effect disappeared when the extracts were treated with indo-methacin/
acetylsalicylic acid (ASA) (). Values are mean standard error (SE). * p- <0.05.
Although most of the lung fluid is reabsorbed during labor and delivery, a small but significant amount of fluid is
still present in the lungs at the time the newborn infant is ready to take the first postnatal breath. A pressure of
about 60 cm H2O is required to make this fluid flow through the airways with the first inspiration (150).
However, a much higher opening pressure would be needed to overcome the high surface tension forces if the
airways were not partially distended with this fluid (Fig. 17-9) (151). According to the Laplace equation for a
cylinder state, the pressure required to overcome surface tension is directly proportional to the surface tension
and indirectly proportional to the radius of curvature (P = t/r). If the airways were not partially distended by
liquid, the opening pressure in the terminal airway would be very large because of the small radius of curvature.
Thus, the normal fluid content of the lung at birth facilitates the first breath by lowering the opening pressure,
and ensuring a more homogeneous filling of the lung with air. A significant reduction in the volume of fetal
pulmonary fluid, as seen sometimes in postterm deliveries, may not be beneficial at birth. First, greater
pressures would be required to inflate the air sacs in fluid-free lungs, and second, the distribution of inspired air
during the first breath may not be as uniform. Faridy (151) showed that the highest opening pressure is seen in
fluid-free lungs and the lowest in lungs containing fluid of approximately 25% of maximum lung volume.
The opening pressure of the lungs at birth also depends on the compliance of the alveolar tissue and the surface
forces at air-fluid interface. During labor and birth, a massive release of surfactant in pulmonary fluid facilitates
lung opening by lowering the opening pressure through the decrease in surface forces and the improvement of
lung compliance (151,152,153,154). Thus, as seen in Fig. 17-10, the first postnatal breath (I) begins with no air
volume in the lungs
P.293
and no transpulmonary pressure gradient. As the chest wall expands, the transpulmonary pressure increases
until it overcomes the surface tension of small airways and alveoli. At this point, actively inspired air begins to
enter the lungs and, according to the Laplace equation, as the radius increases, the distending pressure required
to open up those units decreases.
Figure 17-9 Pressure-volume curves after air versus liquid expansion of the lung. (From Radford EP. In:
Remington JW, ed. Tissue elasticity. Washington, DC: American Physiological Society, 1957, with permission.)
Figure 17-10 Pressure-volume curves of the first three extrauterine breaths. (From Smith CA, Nelson NM.
Physiology of the newborn infant, 4th ed. Springfield, IL: Charles C. Thomas, 1976:125, with permission.)
Although the first inspiratory effort is extremely important for lung opening, the creation of functional residual
capacity (FRC) at the end of the first expiratory effort is essential for the normal pulmonary adaptation at birth.
It is obvious that if all the air that entered the lung were to leave the lung, every breath would necessarily
resemble the first breath (155). This FRC can only be created if the pulmonary surfactant is present allowing for
the stabilization of the peripheral air spaces. The near-zero surface tension and the bubble formation produced
by surfactant allows for retention of large volumes of air at the end of the first expiration. When surfactant is
deficient, the consequences are a tendency to airlessness with each expiration and the application of high
inspiratory pressures to maintain respiration. This leads to the marked retractions so commonly associated with
atelectasis and hyaline membrane disease (HMD) as seen in preterm infants with surfactant deficiency.
Mortola and associates (156) showed that in healthy term infants born by cesarean section, the amount of air
exhaled after the first breath was less than the inhaled volume representing the formation of the FRC. This FRC
continued to rise after the first breath in a very irregular fashion from a mean of about 10 mL/kg at birth to 30
mL/kg by the second day of life (157). The first breaths are actively exhaled by the high negative
transpulmonary pressures (see Fig. 17-10). These expiratory breaths are also associated with interruptions in
the expiratory flow (braking of the expiration), likely as a result of closure in the pharyngeal/laryngeal region,
as indicated by the radiographic studies of Bosma and associates (158). The resulting positive pressure in the
airway generated by this respiratory pattern would not only facilitate liquid absorption, but it would also improve
air retention at the end of expiration, as well as increase lung compliance. Boon and associates (159) showed
that the formation of FRC in asphyxiated neonates born by cesarean section was associated with a stepwise
increase in tidal volume presumably indicating that the physical characteristics of the lungs had changed
sufficiently to allow the lungs to remain inflated. Although vaginally delivered infants have not been extensively
studied, it is likely that they require similar pressures for the first inflation as cesarean-delivered infants. In both
vaginal- and cesarean-delivered infants, the major sources of resistance to inflation (surface forces caused by
air-liquid interface and the frictional forces caused by the movement of the column of liquid in the airway) may
be very similar and yield comparable inspiratory volume and FRC of the first breath (156).
Fetal Circulation
A combination of preferential flow and streaming through structural shunts in the liver (ductus venosus) and
heart (foramen ovale and ductus arteriosus), allows the highest oxygen content blood coming from the placenta
to be delivered to the heart, brain, and upper torso (Fig. 17-11). This relative parallel flow contrasts with the
flow in series and without shunts of the adult circulation. Thus, the volume of blood in the fetal heart ventricles
is not equal. In fact, the right ventricle ejects approximately two-thirds of total fetal cardiac output (300 mL/kg
per minute), whereas the left ventricle ejects only a little more than one third (150 mL/kg per minute) (160).
Placenta blood is delivered to the fetus through the umbilical vein. Approximately 50% of this umbilical blood
flow passes through the ductus venosus directly into the inferior vena cava and mixes with the systemic venous
drainage from the lower body. The other 50% of the umbilical blood flow joins the hepatic portal venous system
and passes through the hepatic vasculature (161). Preferential streaming allows the well-oxygenated blood
derived from the ductus venosus to travel through the dorsal and leftward wall of the inferior vena cava (162). A
tissue flap called the eustachian valve, located at the junction of the inferior vena cava and right atrium, serves
to direct the highly oxygenated blood from the ductus venosus across the foramen ovale into the left atrium and
then the left ventricle and ascending aorta (162,163,164). The less-oxygenated anterior stream (mainly blood
from the lower body and the hepatic circulation) joins the oxygen-poor blood from the superior vena cava
(which drains the head and upper body) and the coronary sinus (which delivers venous return from the
myocardium) at the right atrium,
P.294
directing the blood through the tricuspid valve into the right ventricle (Fig. 17-11).
Figure 17-11 Fetal blood circulation. (From Bloom RS. Delivery room resuscitation of the newborn. In:
Fanaroff AA, Martin RJ, eds. Neonatal and perinatal medicine: diseases of the fetus and infant, 5th ed. St.
Louis: Mosby-Yearbook, 1992:302, with permission.)
Because the placenta is responsible for gas exchange in utero, very little blood flow is sent to the lungs. The
pulmonary circulation is a high-resistance, low-flow circuit that receives less than 10% of the ventricular output.
Instead of entering the pulmonary arteries, most of the right ventricular blood is diverted away from the lungs
through the widely patent ductus arteriosus to the descending aorta, reaching the placenta for oxygenation
through the umbilical arteries.
The well-oxygenated blood coming across the foramen ovale joins the small amount of blood returning from the
lungs via the pulmonary veins in the left atrium and traverses the mitral valve into the left ventricle. This blood
is then ejected across the aortic valve into the ascending aorta bringing well-oxygenated blood to the
myocardium, brain, head, and upper torso (see Fig. 17-11).
Regulation of Fetal Pulmonary Vascular Resistance

There is a close association between pulmonary arterial development and airway development. As gestation
progresses, synchronization of airway and vessel branching occurs, suggesting they may be regulated by
common mediators or exchange messenger molecules (165,166). The embryonic endothelial channels later
acquire a smooth-muscle layer and thus the ability to regulate vascular tone and blood flow.
During early gestation, pulmonary blood flow is limited by the paucity of pulmonary vessels. However, although
the number of small blood vessels per unit of lung volume increases tenfold during the last trimester, pulmonary
blood flow remains low because of a high pulmonary vascular resistance (Fig. 17-12). In addition to the
mechanical compression of pulmonary vessels by the fluid-filled, atelectatic lungs and the lack of rhythmic
distention, this high-resistance low-flow state is probably maintained in part by vasoconstriction because the
fetal pulmonary vascular bed responds readily to vasodilators.
Sympathetic tone does not appear to be responsible for the high pulmonary vascular resistance of the fetus,
P.295
because neither sympathectomy nor -adrenergic blockade decreases the resistance (167,168,169). Although
arachidonic acid and its cyclooxygenase products, prostaglandin F2 and thromboxane A2 are vasoconstrictors in
the fetal and newborn lung, their blockade does not decrease vascular resistance (170,171,172).
Figure 17-12 Representative changes in the pulmonary hemodynamics during transition from the late-term
fetal circulation to the neonatal circulation. P pulmonary vascular resistance (PVR) decreases progressively
during later gestation as a consequence of lung growth and increased cross-sectional area for flow. PVR
decreases dramatically at birth as a consequence of the vasodilating effect of lung aeration. PVR continues to
fall more gradually over the first 6 to 8 weeks of life. Pulmonary blood flow remains at relatively low levels
during fetal growth, then increases abruptly with lung expansion and the rapid fall in PVR. Mean pulmonary
artery pressure falls rapidly immediately after birth because the pulmonary vasodilation causes PVR to fall more
than pulmonary blood flow increases. (Adapted from Rudolph AM. Fetal circulation and cardiovascular
adjustment after birth. In: Rudolph AM, Hoffman JIE, Rudolph CD, eds. Rudolph's pediatrics, 19th ed. Norwalk,
CT: Appleton & Lange, 1991:1309-1313, with permission.)
The discovery that the effects of some vasodilators agents, such as acetylcholine, bradykinin, and histamine,
were dependent on release of an endothelium-derived relaxing factor (EDRF), later shown to be nitric oxide
(NO) (173,174), led to the exploration of its possible role in the perinatal decrease in pulmonary vascular
resistance. Nitric oxide, an inorganic, gaseous free radical discovered in the late 1980s, is produced by the
endothelial cells from the terminal nitrogen of L-arginine by nitric oxide synthase (NOS). NOS can be stimulated
by pharmacologic agents such as acetylcholine or bradykinin, and by birth, shear stress, and oxygen. Nitric
oxide activates soluble guanylate cyclase, which produces smooth-muscle relaxation by activation of protein
kinase C (Fig. 17-13) (175). Hydrolysis of cyclic guanosine monophosphate (cGMP) is accomplished by
phosphodiesterases that control the intensity and duration of cGMP signal transduction (176). In fetal life, NO
production is also stimulated by activation of adenosine triphosphate (ATP)-dependent K+ channels. A
maturational increase in NO-mediated relaxation has been documented during the late fetal and early postnatal
period, which parallels the dramatic fall in pulmonary vascular resistance at birth (177,178,179,180,181). In
fetal lambs, inhibiting NO synthesis increases resting pulmonary vascular resistance and inhibits the ventilationinduced fall in pulmonary vascular resistance. Increased oxygen tension increases both basal and stimulated NO
release and inhibition of NO blocks virtually the entire increase in fetal pulmonary blood flow caused by
hyperbaric oxygenation without ventilation (182,183,184). Shear stress resulting from increased pulmonary
blood flow and rhythmic distention of the lung without changing oxygen tension also induces endothelial NOS
gene expression and contributes to pulmonary vasodilatation at birth (185,186).
Figure 17-13 The proposed mechanism of synthesis and action of nitric oxide. Nitric oxide (NO) is produced in
the endothelium from the terminal guanido nitrogen of L-arginine by NO synthase. NO synthase can be
stimulated by pharmacologic agents such as acetylcholine (Ach) or bradykinin, and by birth, shear stress, and
oxygen. Endothelial production of NO can be blocked by arginine analogues that have modifications of the
guanido nitrogen of the molecule. Nitric oxide activates soluble guanylate cyclase, increases cyclic guanosine
monophosphate (cGMP) concentrations in vascular smooth muscle, and initiates the cascade resulting in
smooth-muscle relaxation. The magnitude and duration of the effect of cGMP is controlled by its inactivation by
specific phosphodi-esterases. GMP, guanosine monophosphate; GTP, guanosine triphosphate; L-NA, nitro-Larginine; L-NMMA, NG-mono- methyl-L-arginine. (From Lakshminrusimha S, Steinhorn R. Pulmonary vascular
biology during neonatal transition. Clin Perinatol 1999;26(3):601-619, with permission.)
Endothelin-1 (ET-1), a 21-amino acid peptide also produced by vascular endothelial cells, has potent vasoactive
activities (187). Although ET-1 appears to play an important and active role in mediating pulmonary vascular
resistance, in vivo studies indicate that the effects of exogenous endothelin are complex and depend on the site,
developmental
P.296
age, and tone of the vascular bed (188,189,190). Consequently, exogenous ET-1 predominantly vasodilates the
fetal and newborn pulmonary circulations by acting on the ETB receptors located on the endothelial cells but
causes vasoconstriction in the adult pulmonary circulation by acting on the ETA receptors located in the smooth
muscle cells (191,192). Increasing data also suggest that endogenous NO and ET-1 participate in the regulation
of each other through an autocrine feedback loop. Thus, ET-1 stimulates the release of NO, and NO inhibits the
ET-1 system (193,194).
The low oxygen tension in fetal life is a physiologic stimulus for the pulmonary vasculature to be constricted.
This hypoxic pulmonary vasoconstriction develops over the period of gestation when the cross-sectional area of
the vascular bed is increasing rapidly. Decreasing oxygen tension in the fetus at 103 days of gestation does not
increase pulmonary vascular resistance, but in the 132- to 138-day fetus it doubles the resistance. Conversely,
increasing oxygen tension before 100 days of gestation does not decrease pulmonary vascular resistance, but by
135 days it decreases resistance markedly and increases pulmonary blood flow to normal newborn levels. Its
mechanism of action may be in part through regulation of activity and gene expression of voltage-gated K+
channels (195), nitric oxide synthase and/or endothelin (196,197).
Unlike the mature pulmonary circulation, the fetal pulmonary vasculature appears to regulate flow through a
myogenic response. The fetal pulmonary vasculature exhibits a time-limited vasodilation in response to dilating
stimuli, including shear stress, oxygen, and many pharmacologic vasodilators, with return to the constricted
resting state despite continued exposure to the vasodilating stimulus. This unique mechanism limits pulmonary
blood flow and preserves placental perfusion and gas exchange.
Transition to Extrauterine Circulation

At birth, a number of complex events must take place so that the fetal circulation that depends on the placenta
for gas exchange and intracardiac and extracardiac shunts to deliver oxygenated blood to the heart and brain,
switches to the neonatal circulation in which the gas exchange is transferred to the lungs and the fetal shunts
are eliminated. A rapid and sustained decrease in pulmonary vascular resistance during the first breaths
facilitate this adaptation. Although this normal pulmonary vascular transition occurs spontaneously and quickly
in most neonates, failure of the pulmonary circulation to undergo this changes results in persistent pulmonary
hypertension of the newborn (PPHN). This condition, which carries significant morbidity and mortality, develops
when the pulmonary vascular resistance fails to decrease adequately during the transition to extrauterine life.
Altered intrauterine environment producing structural changes in the pulmonary circulation or hypoxemia,
acidosis and/or hypercarbia secondary to meconium aspiration, surfactant deficiency, or pneumonia at birth
abnormally constricts the transitional pulmonary circulation. In this condition, right-to-left shunts at the atrial
and ductus arteriosus levels continue secondary to the high pulmonary vascular resistance, producing significant
hypoxemia, which, in turn, causes an increase in pulmonary vasoconstriction. Thus, two major hemodynamic
events must occur at delivery to allow for a normal transition from fetal to neonatal circulation.
Pulmonary Vasodilatation
At birth, pulmonary arterial blood flow increases eight- to tenfold and pulmonary vascular resistance (PVR)
decreases by 50% within the first 24 hours, as the lung assumes the function of gas exchange (Fig. 17-14)
(198,199,200,201,202). This decrease in PVR is brought about by active vasodilation which is regulated by a
complex and incompletely understood interaction among metabolic, hormonal, and mechanical factors, triggered
by a number of birth-related stimuli. Three main factors contribute to the increase in pulmonary blood flow
during this transition: (a) ventilation of the lungs, (b) increased oxygenation, and (c) hemodynamic forces such
as increased shear stress. The effects of these factors on pulmonary circulation at birth appears to be mediated
primarily by the release of NO from the vascular endothelium which results in smooth muscle relaxation via
activation of the intracellular cGMP-dependent protein kinase (203). The initial partial increase in pulmonary
vasodilation may be independent of oxygenation and may be caused by physical expansion of the lungs and the
production of prostaglandins (204). The next component is maximal pulmonary vasodilation associated with
oxygenation, which may be mainly caused by NO synthesis. The increased shear forces related to the rise in
pulmonary blood flow may stimulate endothelial cells to produce NO, which helps maintain pulmonary
vasodilation (Fig. 17-14).
Figure 17-14 Pulmonary vascular conductance increases with the onset of ventilation. Separate curves depict
the contributions of gaseous inflation, increased PO2, and decreased PCO2. (Adapted from Strang LB. The lungs
at birth. Arch Dis Child 1965;40:575, with permission.)
P.297
Ventilation of the Lungs

It has long been known that the initiation of rhythmic breathing causes vasodilation, even in the absence of
increase in oxygen tension (see Fig. 17-14) (8,199). However, physical expansion of the lungs alone produces
only partial increase in pulmonary blood flow and a decrease in PVR. A small proportion of this relates to the
mechanical distention of the lungs and the establishment of an air-liquid interface, as alveolar fluid is replaced
by gas, which increases vessel radius by exerting a negative dilating pressure on the small pulmonary arteries
and veins (205,206). Ventilation of the lung also releases vasoactive substances such as prostacyclin (PGI2)
from vessel walls, which increases pulmonary blood flow and decreases PVR (207). Cyclooxygenase inhibition,
which blocks PGI2 production, prevents the normal decrease in PVR with lung expansion, but not the changes
that occur with oxygenation (207,208,209,210). Other prostaglandins, such as PGD2 and histamine released by
the mast cells during lung expansion, may also contribute to the initial postnatal pulmonary vasodilation (211).
NOS and Kca channel inhibition also blunt ventilation-induced pulmonary vasodilatation (212,213).
Increased Oxygenation
Oxygen is a potent stimulus for pulmonary vasodilation. Even in the absence of ventilation, increased oxygen
tension reduces PVR. Ventilation plus increased oxygen tension produces complete pulmonary vasodilation and
both are particularly crucial in the normal transition to postnatal circulation and pulmonary gas exchange (see
Fig. 17-14) (199). Although the mechanisms of oxygen-induced pulmonary vasodilation are not completely
understood, several factors appear to contribute to this response. It is known that K+ channels are the major
regulators of resting membrane potential in pulmonary arterial smooth muscle cells and that hypoxia decreases
K+ current by repressing K+ channel gene expression (214,215). Gosche and associates (216) have found that
Kvoltage but not Kca or KATP channel activation contributes to the oxygen-induced vasodilation in isolated thirdgeneration pulmonary arterioles from term fetal rats. Whether K+ channels function as sensors of, or effectors
for, oxygen-induced changes in pulmonary vascular tone is unknown. Thus, changes in K+ channel activity may
occur via a direct effect of oxygen tension on the K+ channel (217) or, alternatively, K+ channels may open in
response to changes in the concentration of a second messenger substance like NO (218,219). Many studies
show that the effect of oxygen on perinatal pulmonary circulation appears to be mediated primarily by the
effects of NO on K+ channels in arterial smooth muscle cells either directly or through a cGMP-sensitive kinase
(220,221,222,223). Endothelium-derived nitric oxide modulates pulmonary vascular tone under basal conditions
in the fetus and during transition of the pulmonary circulation at birth. Several studies show that the in utero
increase in pulmonary blood flow observed in response to either ventilation with oxygen or maternal hyperbaric
oxygen exposure can be markedly attenuated with inhibition of NO synthesis (219). It has also been shown that
the maturational rise in NO production seen from late gestation to 4 weeks postnatally is modulated by oxygen.
Thus, the sudden increase in oxygen tension that occurs at birth appears to enhance NO synthesis and NOS
inhibition blunts the oxygen-induced decrease in PVR. Although it is assumed that oxygen may act directly on
endothelial cells to increase NO production, it is possible that it stimulates release of another agent, such as
bradykinin, calcitonin gene-related peptide, or adrenomedullin, which, in turn, stimulates NO production
(224,225).
Oxygen could also increase plasma and red blood cell ATP, which releases NO from endothelial cells and is a
potent fetal pulmonary vasodilator (226). It is known that the release of NO from cultured human vascular
endothelial cells is stimulated by ATP and that the inhibition of NOS by N-nitro-L-arginine attenuates the
vasodilation caused by ATP and its metabolites in the fetal circulation. Konduri and associates (227) suggested
that an increase in oxidative phosphorylation and a release of ATP can also mediate the endothelium-dependent
pulmonary vasodilation that occurs in response to oxygen exposure.
Other studies show that NO release and pulmonary vasorelaxation can also be mediated by endothelial 2adrenoreceptors activation (228,229,230). Magnenant and associates suggested that 2-adrenoreceptors are
involved in the control of basal pulmonary vascular tone and in the pulmonary vasodilator effect of
norepinephrine during fetal life through activation of the NO-dependent pulmonary vasodilation (231).
Recent studies also support the hypothesis that hypoxia causes ETA-mediated inhibition of a K+ channels, which
leads to vessel depolarization and calcium influx, resulting in vasoconstriction (232,233,234,235). Goirand and
associates (236) showed that ETA receptor blockade opposes hypoxic pulmonary vasoconstriction in the rat
isolated perfused lung through the suppression of the inhibition of K+ channels by endogenous ET-1. Thus, the
increase in oxygen, together with the perinatal decrease in ETA receptor message, probably contributes also to
decreased hypoxic pulmonary vasoconstriction observed at birth.
Increased Shear Stress

Partial or complete compression of the ductus arteriosus in utero increases pulmonary blood flow and causes a
progressive fall in pulmonary vascular resistance in response to flow- or shear stress-induced pulmonary
vasodilation (237,238). The initial increase in pulmonary blood flow that occurs with lung expansion and
oxygenation at birth raises shear stress in the pulmonary vascular endothelium, causing further pulmonary
vasodilation (239). This blood flow-dependent pulmonary vasodilation is mainly caused by nitric oxide, which is
produced by an increase in shear stress of vascular endothelium (240,241,242). The released of NO from
arterial and venous endothelial cells contributes to the maintenance of the physiologically low basal pulmonary
vascular tone. Stepwise elevations in flow
P.298
increased exhaled NO and nonselective inhibition of NOS completely blocks the shear stress-induced pulmonary
vasodilation (240,242). This NO shear stress-induced vasodilation is mediated by activation of both voltage- and
Ca2+-dependent K+ channels (241,242,243,244). Ralevic and associates (245) showed that vascular endothelial
cells also release ATP when they are exposed to shear stress. Inhibition of receptors for ATP attenuates the
pulmonary vasodilation that occurs in response to stepwise increases in flow (246).
Loss of the Placental Circulation and Closure of the Fetal Shunts

At birth, the umbilical circulation is abolished as the vascular smooth muscle of the umbilical cord go into spasm
in response to the abrupt longitudinal stretching and the increased oxygenation (247). The removal of the lowresistance bed of the placenta produces much of the increase in systemic vascular resistance at birth. Without
umbilical venous flow, the ductus venosus receives little blood flow and a passive, as well as active, constriction
begins. Complete closure of the ductus venosus usually occurs by the end of the first week of life. The reasons
for closure of the foramen ovale after birth are twofold. First, the occlusion of the umbilical cord decreases the
volume of blood flowing up the inferior vena cava decreasing the right-left atrial pressure difference. Second,
the increase in pulmonary venous return to the left atrium in response to the marked increase in pulmonary
blood flow increases left atrial pressure. The reverse pressure difference across the foramen ovale pushes the
foramen ovale flap against the atrial septum, closing the shunt. The foramen ovale remains patent, with no flow
through it for weeks or months. The closure of the third fetal shunt, the ductus arteriosus, depends
predominantly on the net balance between vasodilators, mainly oxygen, prostaglandins, and other vasoactive
substances (248,249). Increased oxygen tension is known to be a potent stimulant of ductal smooth muscle
constriction after birth (249). Prostaglandins, mainly PGE2, are major mediators of ductal relaxation pre- and
postnatally. The rapid fall in circulating prostaglandins at birth, caused by increased lung metabolism (from
increased pulmonary blood flow) and the loss of placental prostaglandins, facilitates the constrictive effects of
oxygen on the ductal tissue. Although functional closure of the ductus arteriosus occurs in the first 72 hours of
life (250), permanent or anatomical closure takes several days and weeks and is achieved by endothelial
destruction, subintimal proliferation, and connective-tissue formation (251). When the ductus arteriosus fails to
close or reopens in the first days of life, as is frequently observed in small preterm infants, a significant left-toright shunt may occur. The incidence of reopening is inversely related to birth weight. The closure of the patent
ductus arteriosus effectively separates the pulmonary and systemic circulations and establishes the normal
postnatal circulatory pattern.
CONCLUSION
The transition from fetal to neonatal life represents one of the most dynamic and difficult periods in human life
cycle. Dramatic neurohormonal, metabolic, and cardiorespiratory adjustments must occur over hours to days
around the time of delivery to insure the smooth and successful transition to extrauterine life. These changes
are invoked by a variety of processes, including perinatal surges in hormones, labor, delivery, gaseous
ventilation and oxygenation of the lungs, cord occlusion, and decrease in environmental temperature. Intensive
research over the last century has significantly improved our understanding of the normal development of the
cardiorespiratory system that allows the fetus to rapidly and efficiently adapt to air breathing at birth. This
transitional period is characterized by removal of the lung liquid volume from the alveolar spaces and by the
secretion of surfactant material into the acinus for satisfactory physical expansion of the lungs after the initial
postnatal breaths. To maintain adequate ventilation and oxygenation the newborn infant must also switch from
intermittent fetal breathing to continuous breathing at birth, a process that it is still not completely understood.
The switch from placental to pulmonary gas exchange also requires the elimination of the fetal shunts and a
rapid and sustained decrease in pulmonary vascular resistance to allow a significant increase in pulmonary blood
flow. Thus, the circulation changes from one characterized by a relatively low combined ventricular output, right
ventricular dominance, and pulmonary vasoconstriction, to a circulation in series with a high cardiac output
equally divided between the two ventricles, and a greatly dilated pulmonary vascular bed. Many factors can
disrupt this physiologic process causing significant morbidity and mortality.
REFERENCES
1. Harvey W. Exercitatio anatomica de motu cordis et sanguinis in animalibus. London: 1628. Surrey,
England, 1847. Barnes WR, translator.
2. Scheel P. Comentatio de Liquoris amnii aspiras arteriae foetuum humanorum natura et usu. HFNIAE,
1799:86.
3. Zweifel P. Die Respiration des Ftus. Arch Gynakol 1876;9:291.
4. Bichat X. Anatomie gnrale, pplique la physiologie et la mdecine. Paris: Brosson, Gabon, 1801.
5. Kilian HF. Ueber den Kreislauf des Blutes im Kinde, welches noch nicht geathmet hat. Karlsruhe: Muller,
1826.
6. Hamilton WF, Woodbury RA, Woods EB. The relation between systemic and pulmonary blood pressures in
the fetus. Am J Physiol 1937;119:206-212.
7. Barclay AE, Barcroft J, Barron DH, et al. A radiographic demonstration of the circulation through the heart
in the adult and in the foetus, and the identification of the ductus arteriosus. Br J Radiol 1939;12:505.
8. Dawes GS, Mott JC, Widdicombe JG, et al. Changes in the lungs of the newborn lamb. J Physiol
1953;121:141-162.
9. Rudolph AM. In: Weir IK, Archer SL, Reeves JT, eds. The development of concepts of the ontogency of the
pulmonary circulation. Armonk, NY: Futura Publishing, 1999;3-18.
10. Strang LB. Fetal lung liquid: secretion and reabsorption. Physiol Rev 1991;71:991-1016.
11. Preyer W. Specielle Physiologic des Embryo. Leipzig: Greeben Verlag (L. Fernau), 1885:149.
P.299
12. Potter EL, Bohlender GP. Intrauterine respiration in relation to development of the fetal lung. Am J Obstet
Gynecol 1941;42:14-22.
13. Jost A, Policard A. Contribution experimental a l'etude du development prenatal du poumon chez le lapin.
Arch Anat Microsc 1948;37:323-332.
14. Reynolds SRM. A source of amniotic fluid in the lamb nasopharyngeal and buccal cavities. Nature
1953;175:307.
15. Dawes GS, Mott JC, Widdicombe JG. The foetal circulation in the lamb. J Physiol 1954;126(3):563-587.
16. Adams, FH, Moss AJ, Fagan L. The tracheal fluid in the fetal lamb. Biol Neonate 1963;5:151-158.
17. Adamson TM, Boyd RDH, Platt HS, et al. Composition of alveolar liquid in the fetal lamb. J Physiol
1969;204:159-168.
18. Olver RE, Schneeberger EE, Walters DV. Epithelial solute permeability, ion transport and tight junction
morphology in the developing lung of the fetal lamb. J Physiol 1981;315:395-412.
19. Mescher EJ, Platzker ACG, Ballard PL, et al. Ontogeny of tracheal fluid, pulmonary surfactant, and plasma
corticoids in the fetal lamb. J Appl Physiol 1975;39:1017-1021.
20. Burri PH. Fetal and postnatal development of the lung. Annu Rev Physiol 1984;46:617-628.
21. Adamson IYR. Development of lung structure. In: Crystal RG, West JB, Barnes PJ, Cherniack NS, Weibel
ER, eds. The lung: scientific foundations. New York: Raven Press, 1991:663-670.
22. Humphreys PW, Normand ICS, Reynolds EOR, et al. Pulmonary lymph flow and the uptake of liquid from
the lungs of the lamb at the start of breathing. J Physiol 1967;193:1-29.
23. Normand ICS, Olver RE, Reynolds OR, et al. Permeability of lung capillaries and alveoli to non-electrolytes
in the fetal lamb. J Physiol 1971;219:303-330.
24. Harding R, Hooper S. Regulation of lung expansion and lung growth before birth. J Appl Physiol
1996;81:809-224.
25. Hooper SB, Harding R. Fetal lung liquid; a major determinant of the growth and functional development of
the fetal lung. Clin Exp Pharmacol Physiol 1995;22:235-247.
26. Mescher EJ, Platzker ACG, Ballard PL, et al. Ontogeny of tracheal fluid, pulmonary surfactant, and plasma
corticoids in the fetal lamb. J Appl Physiol 1975;39:1017-1021.
27. Adamson TM, Brodecky V, Lambert TF, et al. Lung liquid production and composition in the in utero'
foetal lamb. Aust J Exp Biol Med Sci 1975;53:65-75.
28. Pfister RE, Ramsden CA, Neil HL, et al. Volume and secretion rate of lung liquid in the final days of
gestation and labour in the fetal sheep. J Physiol 2001;535;3:889-899.
29. Orzalesi MM, Motoyama EK, Jacobson HN, et al. The development of the lungs of lambs. Pediatrics
1965;35:373-381.
30. Walters DV. Fetal lung liquid: secretion and absorption. In: Hanson MA, Spencer JAD, Rodeck CH, Walters
D, eds. Fetus and neonate: physiology and clinical application, vol. 2: breathing. Cambridge: Cambridge
University Press, 1994:43-62.
31. Brown MJ, Olver RE, Ramsden CA, et al. Effects of adrenaline and of spontaneous labour on the secretion
and absorption of lung liquid in the fetal lamb. J Physiol 1983;344:137- 152.
32. Alcorn D, Adamson TM, Lambert TF, et al. Morphological effects of chronic tracheal ligation and drainage
in the fetal lamb lung. J Anat 1977;123:649-660.
33. Fewell JE, Johnson P. Upper airway dynamics during breathing and during apnea in fetal lambs. J Physiol
1983;339:495-504.
34. Nardo I, Hopper SB, Harding R. Lung hypoplasia can be reversed by short-term obstruction of the trachea
in fetal sheep. Pediatr Res 1995;38:690-696.
35. Scurry JP, Adamson TM, Cussen LJ. Fetal lung growth in laryngeal atresia and tracheal agenesis. Aust
Paediatr J 1989;25:47-51.
36. Souza P, O'Brodovich H, Post M. Lung fluid restriction affects growth, but not airway branching of
embryonic rat lung. Int J Dev Biol 1995;39:629-637.
37. Crombleholme TM, Albanese CT. The fetus with airway obstruction. In: Harrison MR, Evans MI, Adzick NS,
et al., eds. The unborn patient: the art and science of fetal therapy, 3rd ed. Philadelphia: WB Saunders,
2001:357-371.
38. Lim F, Crombleholme M, Hedrick HL, et al. Congenital high airway obstruction syndrome: natural history
and management. J Pediatr Surg 2003;38:940-945.
39. Matalon S. Mechanisms and regulation of ion transport in adult mammalian alveolar type II pneumocytes.
Am J Physiol 1991;261:C727-C738.
40. Saumon G, Basset G. Electrolyte and fluid transport across the mature alveolar epithelium. J Appl Physiol
1993;74:1-15.
41. Olver RE, Strang LB. Ion fluxes across the pulmonary epithelium and the secretion of the lung liquor in
the fetal lamb. J Physiol 1974;241:327-357.
42. Frizzell RA, Field M, Schultz SG. Sodium-coupled chloride transport by epithelial tissues. Am J Physiol
1979;236:FI-F8.
43. Bland RD, Nielson DW. Developmental changes in lung epithelial ion transport and liquid movement. Annu
Rev Physiol 1992;54:373-394.
44. O'Brodovich HM. Immature epithelial Na+ channel expression is one of the pathogenetic mechanisms
leading to human neonatal respiratory distress syndrome. Proc Assoc Am Physicians 1996;108:345-355.
45. Adams FH, Yanagisawa M, Kuzela D, et al. The disappearance of fetal lung fluid following birth. J Pediatr
1971;78:837-843.
46. O'Brodovich HM, Hannam V. Exogenous surfactant rapidly increases PaO2 in mature rabbits with lungs
that contain large amounts of saline. Am Rev Respir Dis 1993;147:1087-1090.
47. Egan EA, Dillon WP, Zorn S. Fetal lung liquid absorption and alveolar epithelial solute permeability in
surfactant deficient, breathing fetal lambs. Pediatr Res 1984;18:566-570.
48. Barker PM, Gowen CW, Lawson EE, Knowles MR. Decreased sodium ion absorption across nasal epithelium
of very premature infants with respiratory distress syndrome. J Pediatr 1997;130:373-377.
49. Aherne W, Dawkins MJR. The removal of fluid from the pulmonary airways after birth in the rabbit, and
the effect on this of prematurity and pre-natal hypoxia. Biol Neonate 1964;7:214.
50. Bland RD, Carlton DP, Scheerer RG, et al. Lung fluid balance in lambs before and after premature birth. J
Clin Invest 1989:84; 568-576.
51. Bland RD, McMillan DD, Bressack MA, et al. Clearance of liquid from lungs of newborn rabbits. J Appl
Physiol 1980;49:171-177.
52. Sundell HW, Brighman KL, Harris TR, et al. Lung water and vascular permeability-surface area in newborn
lambs delivered by cesarean section compared with the 3-5 day old lamb and adult sheep. J Dev Physiol
1980;2:191-204.
53. Sundell HW, Harris TR, Cannon JR, et al. Lung water and vascular permeability-surface area in premature
newborn lambs with hyaline membrane disease. Circ Res 1987;60:923-932.
54. Jain L. Alveolar fluid clearance in developing lungs and its role in neonatal transition. Clin Perinatol
1999;26(3):585-599.
55. Olver RE, Ramsden CA, Strang LB, et al. The role of amiloride-blockable sodium transport in adrenalineinduced lung liquor reabsorption in the fetal lamb. J Physiol 1986;376:321-340.
56. Walters DV, Ramsden CA, Olver RE. Dibutyryl cAMP induces a gestation-dependent absorption of fetal
lung liquid. J Appl Physiol 1990;68:2054-2059.
57. Chapman DL, Carlton DP, Cummings JJ, et al. Intrapulmonary terbutaline and aminophylline decrease
lung liquid in fetal lambs. Pediatr Res 1991;29:357-361.
58. Barker PM, Olver RE. Clearance of lung liquid during the perinatal period. J Appl Physiol 2002;93:15421548.
59. Berger PJ, Kyriakides MA, Smolich JJ, et al. Massive decline in lung liquid before vaginal delivery at term
in the fetal lamb. Am J Obstet Gynecol 1998;178:223-227.
60. Lines A, Hopper SB, Harding R. Lung liquid production rates and volumes do not decrease before labour in
healthy fetal sheep. J Appl Physiol 1997;82:927-932.
61. Crandall ED, Heming TH, Palombo RL, et al. Effect of terbutaline on sodium transport in isolated perfused
rat lung. J Appl Physiol 1986;60:289-294.
62. Basset G, Crone C, Saumon G. Significance of active ion transport in transalveolar water absorption: a
study on isolated rat lung. J Physiol 1987;384:311-324.
63. Berthiaume Y, Boraddus VC, Gropper MA, et al. Alveolar liquid and protein clearance from normal dog
lungs. J Appl Physiol 1988;65:585-593.
64. O'Brodovich H, Hanman V, Seear M, et al. Amiloride impairs lung liquid clearance in newborn guinea pigs.
J Appl Physiol 1990;68:1758-1762.
65. Smedira N, Gates L, Hastings R, et al. Alveolar and lung liquid clearance in anesthetized rabbits. J Appl
Physiol 1991;70:1827-1835.
66. Sakuma T, Okaniwa G, Nakada T, et al. Alveolar fluid clearance in the resected human lung. Am J Respir
Crit Care Med 1994;150: 305-310.
P.300
67. Yue G, Matalon S. Mechanisms and sequelae of increased alveolar fluid clearance in hyperoxic rats. Am J
Physiol Lung Cell Mol Physiol 1997;272:L407-L412.
68. O'Brodovich H, Hannam V, Rafii B. Sodium channel but neither Na+-H+ nor Na-glucose symport inhibitors
slow neonatal lung water clearance. Am J Respir Cell Mol Biol 1991;5:377-384.
69. Sakuma T, Pittet JF, Jayr C, et al. Alveolar liquid and protein clearance in the absence of blood flow or
ventilation in sheep. J Appl Physiol 1993;74:176-185.
70. Jayr C, Garat C, Meignan M, et al. Alveolar liquid and protein clearance in anesthetized ventilated rats. J
Appl Physiol 1994;76: 2636-2642.
71. Icard P, Saumon G. Alveolar sodium and liquid transport in mice. Am J Physiol Lung Cell Mol Physiol
1999;277:L1232-L1238.
72. Matthay MA, Folkesson HG, Clerici C. Lung epithelial fluid transport and the resolution of pulmonary
edema. Physiol Rev 2002;82:569-600.
73. Walters DV. Fetal lung liquid: secretion and absorption. In: Hanson MA, Spencer JAD, Rodeck CH, et al.,
eds. Fetus and neonate: physiology and clinical application. vol. 2: breathing. Cambridge: Cambridge
University Press, 1994:43-62.
74. Umenishi F, Carter EP, Yang B, et al. Sharp increase in rat lung water channel expression in the perinatal
period. Am J Respir Cell Mol Biol 1996;15:673-679.
75. Ruddy MK, Drazen JM, Pitkanen OM, et al. Aquaporin-4 is expressed in rat fetal distal lung epithelial cells
(FDLE) where it may function in Na+ mediated water reabsorption during the perinatal period. Am J Respir
Crit Care Med 1996;153:A2321(abst).
76. Fyfe GK, Kemp PJ, Cragoe EJ Jr, et al. Conductive cation transport in apical membrane vesicles prepared
from fetal lung. Biochim Biophys Acta 1994;1224:355-364.
77. Matthay MA, Folkesson HG, Verkman AS. Salt and water transport across alveolar and distal airway
epithelium in the adult lung. Am J Physiol Lung Cell Mol Physiol 1996;270:L487-L503.
78. Ma T, Fukuda N, Song Y, et al. Lung fluid transport in aquaporin-5 knockout mice. J Clin Invest
2000;105:93-100.
79. Cott GR. Modulation of bioelectric properties across alveolar type II cells by substratum. Am J Physiol
1989;257:C678-C688.
80. Orser BA, Bertlik M, Fedorko L, et al. Cation selective channel in fetal alveolar type II epithelium. Biochim
Biophys Acta 1991;1094: 19-26.
81. Rao AK, Cott GR. Ontogeny of ion transport across fetal pulmonary epithelial cells in monolayer culture.
Am J Physiol 1991;261:L178-L187.
82. Barker PM, Boucher RC, Yankaskas JR. Bioelectric properties of cultured monolayers from epithelium of
distal human fetal lung. Am J Physiol 1995;268:1270-1277.
83. Canessa CM, Schild L, Buell G, et al. Amiloride-sensitive epithelial Na+ channel is made of three
homologous subunits. Nature 1994;367:412-413.
84. Smith DE, Otulakowski G, Yeger H, et al. Epithelial Na(+) channel (ENaC) expression in the developing
normal and abnormal human perinatal lung. Am J Respir Crit Care Med 2000;161: 1322-1231.
85. Gaillard D, Hinnrasky J, Coscoy S, et al. Early expression of - and -subunits of epithelial sodium channel
during human airway development. Am J Physiol Lung Cell Mol Physiol 2000;278:L177-L184.
86. Hummler E, Barker P, Gatzy J, et al. Early death due to defective neonatal lung liquid clearance in ENaCdeficient mice. Nat Genet 1996;12:325-328.
87. Barker PM, Nguyen MS, Gatzy JT, et al. Role of -ENaC subunit in lung liquid clearance and electrolyte
balance in newborn mice. Insights into perinatal adaptation and pseudohypoaldosteronism. J Clin Invest
1998;102:1634-1640.
88. Walters DV, Olver RE. The role of catecholamines in lung liquid absorption at birth. Pediatr Res
1978;12:239-242.
89. Niisato N, Ito Y, Marunaka Y. cAMP stimulates Na+ transport in rat fetal pneumocyte: involvement of a
PTK- but not a PKA-dependent pathway. Am J Physiol Lung Cell Mol Physiol 1999;277:L727-L736.
90. Norlin, Andreas, Folkensson Hans G. Ca2+-dependent stimulation of alveolar fluid clearance in near-term
fetal guinea pigs. Am J Physiol Lung Cell Mol Physiol 2002;282:L642-L649.
91. Brown MJ, Olver RE, Ramsden CA, et al. Effects of adrenaline and spontaneous labour on the secretion
and absorption of lung liquid in the fetal lamb. J Physiol 1983;344:137-142.
92. Krochmal-Mokrzan EM, Barker PM, Gatzy JT. Effects of hormones on potential difference and liquid
balance across explants from proximal and distal fetal rate lung. J Physiol 1993;463:647-665.
93. Cott GR, Rao AK. Hydrocortisone promotes the maturation of the Na+ dependent ion transport across the
fetal pulmonary epithelium. Am J Respir Cell Mol Biol 1993;9:166-171.
94. Tehepichev S, Ueda J, Canessa C, et al. Lung epithelial Na channel subunits are differentially regulated
during development and by steroids. Am J Physiol 1995;269:C805-C812.
95. Barker PM, Gatzy JT. Effects of gas composition on liquid secretion by explants of distal lung of fetal rat in
submersion culture. Am J Physiol Lung Cell Mol Physiol 1993;265:L512.
96. Raj JU, Bland RD. Lung luminal liquid clearance in newborn lambs. Effect of pulmonary microvascular
pressure elevation. Am Rev Respir Dis 1986;134:305.
97. Cummings JJ, Carlton DP, Poulain FR, et al. Hypoproteinemia slows lung liquid clearance in young lambs. J
Appl Physiol 1993; 74:153-160.
98. Bland RD, Hansen TN, Haberkern CM, et al. Lung fluid balance in lambs before and after birth. J Appl
Physiol 1982;53:992-1004.
99. Dawes GS, Fox HE, Leduc BM, et al. Respiratory movements and paradoxical sleep in the fetal lamb. J
Physiol 1970;210:47P.
100. Merlet C, Hoerter J, Devilleneuve C, et al. Mise en evidence de mouvements respiratoires chez le foetus
d'agneau au cours du dernier mois de la gestation. C R Acad Sci Ser D 1970;270: 2462-2464.
101. Dawes GS, Fox HE, Leduc MB, et al. Respiratory movements and rapid eye movement sleep in the fetal
lamb. J Physiol 1972;220: 119-143.
102. Boddy K, Dawes GS, Fisher R, et al. Fetal respiratory movements, electrocortical and cardiovascular
responses to hypoxaemia and hypercapnia in sheep. J Physiol 1974;243:599-618.
103. Maloney JE, Adamson TM, Brodecky V, et al. Modification of respiratory center output in the
unanesthetized fetal sheep in utero. J Appl Physiol 1975;39:552-558.
104. Maloney JE, Bowes G, Wilkinson M. Fetal breathing and the development of patterns of respiration
before birth. Sleep 1980;3:299-306.
105. Ioffe S, Jansen AH, Russell BJ, et al. Respiratory response to somatic stimulation in fetal lambs during
sleep and wakefulness. Pfluegers Arch 1980;388:143-148.
106. Ioffe S, Jansen AH, Russell BJ, et al. Sleep, wakefulness and the monosynaptic reflex in fetal and
newborn lambs. Pfluegers Arch 1980;388:149-157.
107. Rigatto H, Moore M, Cates D. Fetal breathing and behavior measured through a double-wall Plexiglas
window in sheep. J Appl Physiol 1986;61:160-164.
108. Boddy K, Dawes GS. Fetal breathing. Br Med Bull 1975;31:3-7.
109. Dawes GS, Fox HE, Leduc MB, et al. Respiratory movements and rapid eye movement sleep in the fetal
lamb. J Physiol 1972;220:119-143.
110. Patrick J, Campbell K, Carmichael L, et al. A definition of human fetal apnea and the distribution of fetal
apneic intervals during the last ten weeks of pregnancy. Am J Obstet Gynecol 1980;136:471-477.
111. Patrick J, Campbell K, Carmichael L, et al. Patterns of human fetal breathing during the last 10 weeks of
pregnancy. Obstet Gynecol 1980;56:24-30.
112. Dawes GS. Breathing before birth in animals and man. N Engl J Med 1974;290:557-559.
113. Kitterman JA, Liggins GC, Clements JA, Tooley WH. Stimulation of breathing movements in fetal sheep
by inhibitors of prostaglandin synthesis. J Dev Physiol 1979;1:453-466.
114. Dawes GS, Gardner WN, Johnston BM, et al. Effects of hypercapnia on tracheal pressure, diaphragm and
intercostal electromyograms in unanesthetized fetal lambs. J Physiol 1982;326: 461-474.
115. Jansen AH, Ioffe S, Russell BJ, et al. Influence of sleep state on the response to hypercapnia in fetal
lambs. Respir Physiol 1982;48: 125-142.
116. Moss IR, Scarpelli EM. Generation and regulation of breathing in utero: fetal CO2 response test. J Appl
Physiol 1979;47:527- 531.
117. Rigatto H. A new window on the chronic fetal sheep model. In: Nathanielsz PW, ed. Animal models in
fetal medicine. Ithaca, NY: Perinatology Press, 1984:57-67.
P.301
118. Rigatto H, Hasan SU, Jansen A, et al. The effect of total peripheral chemodenervation on fetal breathing
and on the establishment of breathing at birth in sheep. In: Jones CT, ed. Fetal and neonatal development.
Ithaca, NY: Perinatology Press, 1988:613-621.
119. Clewlow F, Dawes GS, Johnston BM, et al. Changes in breathing, electrocortical and muscle activity in
unanesthetized fetal lambs with age. J Physiol 1983;341:463-476.
120. Koos BJ, Sameshima H, Power GG. Fetal breathing, sleep state, and cardiovascular responses to graded
hypoxia in sheep. J Appl Physiol 1987;62:1033-1039.
121. Baier RJ, Hasan SU, Cates DB, et al. Effects of various concentrations of O2 and umbilical cord occlusion
on fetal breathing and behavior. J Appl Physiol 1990;68:1597-1604.
122. Moss IR, Inman JG. Neurochemicals and respiratory control during development. J Appl Physiol
1989;67:1-13.
123. Johnston BM, Gluckman PD. GABA mediated inhibition of breathing in the late gestation sheep fetus. J
Dev Physiol 1983;5:353-360.
124. Boddy K, Dawes GS, Fisher R, et al. The effects of pentobarbitone and pethidine on foetal breathing
movements in sheep. Br J Pharmacol 1976;57:311-317.
125. Piercy WN, Day MA, Nims AH, et al. Alteration of ovine fetal respiratory-like activity by diazepam,
caffeine and doxapram. Am J Obstet Gynecol 1977;127:43-49.
126. Paul SM, Maraugos PJ, Skolnick P. The benzodiazepine-GABA-chloride ionophore receptor complex:
common site of minor tranquilizer action. Biol Psychol 1981;16:213-229.
127. Dawes GS. The establishment of pulmonary respiration. In: Foetal and neonatal physiology. Chicago:
Year Book, 1968:125-159.
128. Harned H, Ferreiro J. Initiation of breathing by cold stimulation: effects of change in ambient
temperature on respiratory activity of the full-term fetal lambs. J Pediatr 1973;88:663-669.
129. Jansen AH, Ioffe S, Russell BJ, et al. Effect of carotid chemoreceptor denervation on breathing in utero
and after birth. J Appl Physiol 1981;51:630-633.
130. Rigatto H, Lee D, Davi M, et al. Effect of increased arterial CO2 on fetal breathing and behavior in sheep.
J Appl Physiol 1988; 64:982-987.
131. Alvaro R, Weintraub Z, Alvarez J, et al. The effects of 21 or 30% O2 plus umbilical cord occlusion on fetal
breathing and behavior. J Dev Physiol 1992;18:237-242.
132. Alvaro R, deAlmeida V, Al-Alaiyan S, et al. A placental extract inhibits breathing induced by umbilical
cord occlusion in fetal sheep. J Dev Physiol 1993;19:23-28.
133. Adamson SL, Richardson BS, Homan J. Initiation of pulmonary gas exchange by fetal sheep in utero. J
Appl Physiol 1987;62: 989-998.
134. Adamson SL, Kuiper IM, Olson DM. Umbilical cord occlusion stimulates breathing independent of blood
gases and pH. J Appl Physiol 1991;70:1796-1809.
135. Thorburn GD. The placenta and the control of fetal breathing movements. Reprod Fertil Dev 1995;7:577594.
136. Alvarez JE, Baier RJ, Fajardo CA, et al. The effect of 10% O2 on the continuous breathing induced by O2
or O2 plus cord occlusion in the fetal sheep. J Dev Physiol 1992;17:227-232.
137. Baier, RJ, Fajardo CA, Alvarez J, et al. The effects of gestational age and labour on the breathing and
behavior response to oxygen and umbilical cord occlusion in the fetal sheep. J Dev Physiol 1992;18:93-98.
138. Baier, RJ, Hasan SU, Cates DB, et al. Hyperoxemia profoundly alters breathing pattern and arouses the
fetal sheep. J Dev Physiol 1992;18:143-150.
139. Alvaro RE, Robertson M, Lemke R, et al. Effects of a prolonged infusion of a placental extract on
breathing and electrocortical activity in the fetal sheep. Pediatr Res 1997;41:300A.
140. Alvaro RE, Hasan S, Chemtob S, et al. The inhibition of breathing observed with a placental extract in
fetal sheep is due to prostaglandin. Pediatr Res 2002;51:332A.
141. Kitterman J, Liggins GC, Fewell JE, et al. Inhibition of breathing movements in fetal sheep by
prostaglandins. J Appl Physiol 1983;54:687-692.
142. Wallen LD, Mural DT, Clyman RI, et al. Regulation of breathing movements in fetal sheep by
prostaglandin E2. J Appl Physiol 1986;60:526-531.
143. Koos BJ. Central stimulation of breathing movements in fetal lambs by prostaglandin synthetase
inhibitors. J Physiol 1985;362: 455-456.
144. Kitterman J. Arachidonic acid metabolites and control of breathing in the fetus and newborn. Semin
Perinatol 1987;11:43-52.
145. Koos BJ, Maeda T. Fetal breathing, sleep state and cardiovascular response to adenosine in sheep. J Appl
Physiol 1990;68:489-495.
146. Bissonette JM, Hohimer AR, Knopps SJ. The effect of centrally administered adenosine on fetal breathing
movements. Respir Physiol 1991;84:273-285.
147. Koos BJ, Maeda T, Jan C. Adenosine A1 and A2A receptors modulate sleep state and breathing in fetal
sheep. J Appl Physiol 2001;91:343-350.
148. Koos BJ, Phil D, Takatsugu M, et al. Adenosine A2A receptors mediate hypoxic inhibition of fetal breathing
in sheep. Am J Obstet Gynecol 2002;186:663-668.
149. Chernick V. Mechanics of the first inspiration. Semin Perinatol 1977;1:4:347-355.
150. Agostoni E, Talietti A, Agostoni AF, et
Chapter 18
Delivery Room Management
Virender K. Rehan
Roderic H. Phibbs
Although more than 90% of neonates undergo smooth feto/neonatal transition, of the 10% who
require some resuscitative assistance during delivery, a small minority requires extensive
resuscitative efforts (1,2). More importantly, because the need for resuscitation can come as a
complete surprise, at every delivery there should be at least one person whose primary
responsibility is the management of the newly born and who is capable of initiating resuscitation.
Either this person or someone else who is immediately available should be skilled in all aspects of
neonatal resuscitation. However, if the need for resuscitation is anticipated, additional skilled
personnel should be called on before delivery. Furthermore, because many high-risk deliveries
occur in nonteaching and smaller hospitals, all personnel involved in delivery room care of the
newborn should be trained adequately in all aspects of neonatal resuscitation. All necessary
resuscitation equipment should be checked and in working order before each delivery.
There is a high risk of asphyxia, defined as a combination of hypoxemia, hypercapnia, and
acidosis, during labor, delivery, and in the first minutes after birth. This is because the newborn
infant must successfully inflate his or her lungs and make adaptations to the circulation
immediately after birth. Failure of either to occur leads to asphyxia. The key changes during feto/
neonatal transition are the establishment of effective ventilation and perfusion of the lungs to raise
partial pressure of arterial oxygen (PaO2) from the normal low levels of the fetus to the normal,
relatively higher level of the neonate, together with the shutting down of the fetal circulatory
pathways, which include the right-to-left shunts through the foramen ovale and the ductus
arteriosus. Skillful resuscitation of infants with impaired transition can prevent brain damage and
minimize subsequent morbidity and mortality. An understanding of the physiologic changes in the
respiratory and circulatory systems that occur normally as the newborn infant adapts to
extrauterine life is essential for a rational and effective approach to resuscitation.
At birth, the lungs must transition rapidly to become the site for gas exchange or else hypoxia and
cyanosis (see Color Plate) will rapidly develop. For the lungs to exchange gas adequately after
birth, the airways and the alveoli must be cleared of fetal lung fluid, and an increase in pulmonary
blood flow must occur. In utero, fetal pulmonary vascular resistance is high and the fetal systemic
vascular resistance is low; most of the cardiac output is shunted away from the lungs and is
directed to the placenta where fetoplacental gas exchange occurs. Within minutes of delivery, the
newborn's pulmonary vascular resistance may decrease eight- to tenfold, causing a corresponding
increase in neonatal pulmonary blood flow. Effective transition requires that the lung fluid be
expelled or quickly absorbed to allow effective gas exchange. In fact, the decrease in lung fluid
begins during labor (3). As a consequence of increased catecholamine levels, during labor, there is
also an increase in lymphatic drainage. The absence of these physiologic events accounts for the
increased incidence of transient tachypnea of the newborn after a cesarean section without labor.
The first breath must generate a high transpulmonary pressure to overcome the viscosity of the
lung fluid and the intraalveolar surface tension. It also helps to drive the alveolar fluid across the
alveolar epithelium. Lung expansion and aeration also stimulate surfactant release with the
resultant establishment of an air-fluid interface and development of functional residual capacity
(FRC) (4). Normally, 80% to 90% of FRC is established within the first hour of birth in the term
neonate with spontaneous respirations.
At birth, the clamping of the umbilical cord increases the systemic vascular resistance with a
resultant increase in left ventricular and aortic pressures. Lung aeration and subsequent
P.305
gas exchange result in increased PaO2 and pH, which result in pulmonary vasodilation. These
physiologic changes increase flow of blood to the left atrium via the pulmonary veins, so that left
atrial pressure exceeds right atrial pressure, resulting in functional closure of foramen ovale. When
pulmonary vascular resistance decreases to a level lower than the systemic vascular resistance,
the ductus arteriosus closes functionally. As a result of the cessation of umbilical venous return,
clamping of the umbilical cord also leads to the closure of the ductus venosus.
Asphyxia has the potential to set in motion a series of responses that can not only impair the
normal feto/neonatal transition, but may, in fact, reverse this process and lead to the persistence
of the fetal circulatory state. Hypoxia keeps the ductus arteriosus open and causes pulmonary
vasoconstriction leading to right-to-left flow across the ductus. Tissue hypoxia causes metabolic
acidosis worsening the pulmonary vasoconstriction. The pulmonary hypertension leads to tricuspid
insufficiency, raising right atrial pressure, which leads to right-to-left shunting of blood through the
foramen ovale, causing further tissue hypoxia. However, in the majority of instances, with timely
and appropriate resuscitation, the changes leading to a persistent fetal circulation like state can be
reversed quickly.
PATHOPHYSIOLOGY OF INTRAPARTUM ASPHYXIA AND

RESUSCITATION
Asphyxia occurs when the organ of gas exchange fails. When this happens, arterial carbon dioxide
partial pressure (PaCO2) rises, and PaO2 and pH fall. Despite the low PaO2, tissues continue to
consume O2, although at a lower rate in some organs. When the PaO2 is very low, anaerobic
metabolism sets in, producing large quantities of metabolic acids. These are buffered partly by the
bicarbonate in the blood (5).
The human infant is particularly vulnerable to asphyxia in the perinatal period. During normal
labor, transient hypoxemia occurs with uterine contractions, but the healthy fetus tolerates this
well. There are five basic events that lead to asphyxia during labor and delivery:
Interruption of umbilical blood flow (e.g., cord compression);
Failure of gas exchange across the placenta (e.g., placental abruption);
Inadequate perfusion of the maternal side of the placenta (e.g., severe maternal
hypotension);
An otherwise compromised fetus who cannot further tolerate the transient, intermittent
hypoxia of normal labor (e.g., the anemic or growth-retarded fetus); and
Failure to inflate the lungs and complete the changes in ventilation and lung perfusion that
must occur at birth. This failure may occur because of airway obstruction, excessive fluid in
the lungs, or weak respiratory effort. Alternatively, it may occur as a result of fetal asphyxia
from one of the other four events, because fetal asphyxia often results in an infant who is
acidotic and apneic at birth.
The umbilical cord blood pH, partial pressure of oxygen (PO2), partial pressure of carbon dioxide
(PCO2), and calculated base excess are standard measures of fetal asphyxia (6,7,8). With fetal
acidosis, the pH can vary over a wide range. Consequently, it is important to remember that pH is
a logarithmic function of hydrogen ion concentration. A decrease of 0.3 pH units from 7.40 to 7.10
indicates only a 40 nmol/L increase in hydrogen ion (i.e., from 40 to 80 nmol), whereas a 0.3
decrease from 7.10 to 6.80 indicates an increase of 80 nmol/L (i.e., from 80 to 160 nmol). The
gradient in blood gas tensions between umbilical artery and vein gives some indication of placental
perfusion at the time of birth. The slower the flow of fetal blood through the placenta, the more
complete the equilibration of gas tensions between fetal and maternal blood. For example, an
arterial PO2 of 25 mm Hg with a venous PO2 of 32 mm Hg suggests good placental blood flow. An
arterial PO2 of 12 mm Hg with a venous PO2 of 45 mm Hg suggests very slow flow. Metabolic
acidosis suggests asphyxia, although some of the increased lactic acid in the blood may be a result
of reduced uptake of lactate by the asphyxiated liver rather than increased lactate production from
anaerobic metabolism (6,9). If asphyxia occurred just before birth, there may be lactic acid in the
tissues that has not yet reached the central circulation. This will be detected only by blood gas
measurements a few minutes after birth. If the fetus was asphyxiated an hour before delivery and
recovered, that event may not be reflected in the umbilical cord blood gases at birth. Other
indicators of asphyxia include plasma hypoxanthine, which increases because of lack of aerobic
metabolism, plasma erythropoietin, which increases in response to fetal hypoxia, increased plasma
levels of several lipid mediators, such as platelet-activating factor, and increased cerebrospinal
fluid levels of several proinflammatory cytokines such as interleukin (IL)-1, IL-6, and IL-8
(10,11,12).
Asphyxia in the fetus or newborn infant (including a preterm infant) is a progressive and reversible
process. The speed and extent of progression are highly variable. Sudden, severe asphyxia can be
lethal in less than 10 minutes. Mild asphyxia may progressively worsen over 30 minutes or more.
Repeated episodes of brief, mild asphyxia may reverse spontaneously but produce a cumulative
effect of progressive asphyxia. In the early stages, asphyxia usually reverses spontaneously if its
cause is removed. Once asphyxia is severe, spontaneous reversal is unlikely because of the
circulatory and neurologic changes that accompany it. Other sources provide a more detailed
review of these phenomena (13,14).
Figure 18-1 schematically represents the sequence of pathophysiologic changes that accompany
asphyxia. Although there are some quantitative differences between the changes that occur in the
fetus and those in the newborn infant, the scheme generally applies to both. It is useful to
consider the changes in both fetus and newborn infant together, because many cases of neonatal
asphyxia
P.306
begin in the fetus and continue after birth. Cardiac output is maintained early in asphyxia, but its
distribution changes radically. Selective regional vasoconstriction reduces blood flow to less vital
organs and tissues such as gut, kidneys, muscle, and skin (15). Blood flow to the brain and
myocardium increases, thereby maintaining adequate oxygen delivery despite reduced oxygen
content of the arterial blood. Other organs and tissues must depend on increased oxygen
extraction to maintain oxygen consumption (16,17). Pulmonary blood flow is low in the fetus. It is
decreased further by hypoxia and acidosis (18). As a consequence of these adaptations, fetal
oxygen consumption decreases (19).
Figure 18-1 The sequence of cardiopulmonary changes with asphyxia and resuscitation. If there
is complete interruption of respiratory gas exchange, the entire process of asphyxia could occur in
about 10 minutes. It could take much longer with an asphyxiating process that only partly
interrupts gas exchange or one that does so completely but only for repeated brief periods. With
resuscitation, the process reverses, beginning at the point to which the asphyxia has proceeded.
(Adapted from Dawes G. Fetal and Neonatal Physiology. Chicago: Year-Book, 1968, with
permission.)
Early in asphyxia, newborns make vigorous attempts to inflate their lungs. If successful, the lungs
become adequately ventilated and perfused, but the mere presence of gasping does not ensure
that this will happen. As asphyxia becomes more severe, the respiratory center is depressed, and
the chances of an infant spontaneously establishing effective ventilation and pulmonary perfusion
diminish.
If asphyxia progresses to the severe stage, oxygen delivery to the brain and heart decreases. The
myocardium then uses its stored reserve of glycogen for energy. Eventually, the glycogen reserve
is consumed and the myocardium is exposed simultaneously to progressively lower values of PO2
and pH. The combined effects of hypoxia and acidosis lead to decreased myocardial function and
decreased blood flow to the vital organs (20,21). Brain injury begins late during this phase (13,14).
This sequence of cardiovascular events is manifested by changes in heart rate and aortic and
central venous pressures (see Fig. 18-1), all of which are measured easily in the newborn
immediately after birth. The early bradycardia and hypertension are caused by the reflexes that
shunt blood away from nonvital organs. Early in asphyxia, central venous (i.e., right atrial)
pressure may rise slightly, owing to pulmonary hypertension and constriction of systemic
capacitance vessels. As the myocardium fails, central venous pressure rises further, aortic
pressure decreases, and heart rate is reduced further.
The initial adaptations of the systemic circulation to asphyxia are mediated by various physiologic
reflexes (22). There also are major hormonal responses to asphyxia, including elevations in
plasma corticotrophin, glucocorticoids, catecholamines, arginine vasopressin, renin, and atrial
natriuretic factor, and a decrease in insulin (23,24). Some of these are important in maintaining
the circulatory adaptations to asphyxia. Catecholamines, which come mainly from the adrenal
medulla, maintain myocardial function in the presence of asphyxia, thereby increasing survival
(25,26). Arginine vasopressin helps maintain the hypertension, bradycardia, and redistribution of
systemic flow (27). Increased hepatic glycogenolysis helps maintain plasma glucose concentrations
(9).
The physiology of resuscitation is essentially a reversal of the pathophysiology of asphyxia. In Fig.
18-1, which illustrates both processes, asphyxia proceeds from left to right, and resuscitation from
right to left. It is crucial to determine where the infant is in this sequence of pathophysiologic
events when resuscitation is started. If asphyxia has proceeded to myocardial failure, resuscitation
must include restoration of cardiac output as well as establishment of effective ventilation and
perfusion of the lungs. Generally, myocardial failure does not occur until both pH and PaO2 are
extremely low, approximately 6.9 and 20 mm Hg, respectively. Cardiac output is reestablished
through rapid correction of the severe hypoxia and acidosis. Until this is done, output must be
maintained by cardiac massage. As soon as pH is raised to approximately 7.1 and PaO2 to 50 mm
Hg, the myocardium responds rapidly, heart rate rises, aortic pressures rise, and pulse pressure
widens, whereas central venous pressure falls. These changes indicate that cardiac massage can
be stopped. At this point, the infant may be hypertensive because the vasoconstriction in nonvital
organs is still present. This vasoconstriction is relieved only by continued adequate oxygenation
and correction of acidosis. Pressures then will fall toward normal. The vasoconstriction also is
manifested by intense pallor of the skin. As the vasoconstriction is relieved, the skin becomes pink
and well perfused, with rapid capillary refilling (i.e., less than 2 seconds) when blanched by
pressure. As peripheral flow improves, lactic
P.307
acid sequestered in these tissues enters the central circulation and a large base deficit, which may
have been corrected earlier, now reappears.
If asphyxia is only moderately severe, resuscitation begins in the middle of the sequence depicted
in Fig. 18-1. There is hypertension, indicating that the myocardium has not yet failed. Effective
ventilation of the lungs with a high oxygen concentration may correct acidosis by lowering the
PaCO2, oxygenating the blood, and adequately dilating the pulmonary vascular bed. If significant
acidosis persists after alleviation of the hypercarbia, alkali might be needed to correct the
metabolic component of the acidosis, relieve pulmonary vasoconstriction, and establish good
pulmonary perfusion. Generally, raising pH to 7.25 is sufficient for this purpose, but there are
some important exceptions (discussed below) in which a higher pH is needed to dilate the
pulmonary vascular bed.
When the effects of asphyxia are alleviated, spontaneous respiratory efforts return. The duration
between the onset of resuscitation and reappearance of spontaneous respiratory efforts is directly
proportional to the amount of brain injury that has occurred (13).
Onset of spontaneous respiratory efforts is not necessarily an indication to withdraw assisted
ventilation. Often, there is residual atelectasis and the infant does not have strong, regular
respiratory efforts. PaCO2 may be normal, and PaO2 may rise to a high level with assisted
ventilation. But when assisted ventilation is withdrawn, effective ventilation may decrease and the
whole process of asphyxia recurs. Therefore, assisted ventilation and supplemental oxygen should
be only gradually withdrawn.
The blood volume of the asphyxiated infant may be abnormal. Asphyxia during labor usually shifts
blood from the placenta to the fetus. There are certain situations, however, in which the infant's
blood volume may be reduced. The most obvious of these is hemorrhage from the fetoplacental
unit, which is manifested by vaginal bleeding. Three other conditions that shift blood volume from
the fetus to the placenta are compression of the umbilical cord, in which umbilical venous flow is
reduced selectively more than arterial flow; severe hypotension in the mother; and asphyxia
occurring only at the end of labor (28).
Initially, it may be difficult to determine whether or not blood volume is adequate in the
asphyxiated newborn. There are two reasons for this. First, many of the circulatory responses to
asphyxia are similar to those associated with loss of blood volume. Either asphyxia or hypovolemia
may cause bradycardia, metabolic acidosis, poor peripheral perfusion indicated by pallor and slow
capillary filling, and a large difference between core and skin temperature. A low aortic pressure
could be a result of either the end stage of asphyxia or to shock. Only changes in central venous
pressure are in the opposite direction, and even here the coexistence of the two processes can
have offsetting effects. Second, the circulatory changes during asphyxia and resuscitation may
determine the adequacy or inadequacy of the circulating blood volume. If an infant is moderately
asphyxiated and has systemic and pulmonary vasoconstriction (see Fig. 18-1, center) and a small
blood volume, aortic and central venous pressures will be nearly normal. Administration of a blood
volume expander at this point would only overload the circulation. The effects of volume expansion
would be even worse if the asphyxia were more severe and myocardial failure were present.
Correction of asphyxia (see Fig. 18-1, going from right to left) relieves the vasoconstriction of
resistance and capacitance vessels, and the small blood volume now becomes inadequate to
support the circulation. Reperfusion of asphyxic and ischemic tissues also increases loss of
intravascular water from these capillary beds, leading to edema and reduced plasma volume.
During recovery from asphyxia, several metabolic abnormalities appear. There may be
hypoglycemia caused by depletion of carbohydrate reserves during the asphyxia. Hypoglycemia
must be prevented because it can cause myocardial failure in a heart recently subjected to
asphyxia (29). Hyperglycemia caused by excessive glucose administration similarly is dangerous
during asphyxia because it worsens the acidosis by increasing lactic acid production (30).
Hypocalcemia also develops, possibly as a result of increased calcitonin release during asphyxia
(31), and can lead to myocardial failure.
Hyperkalemia occurs during asphyxia, when, in the process of buffering acidosis, H+ enters the
erythrocytes and K+ is displaced from them. Although this increases plasma K+ while the patient is
asphyxiated, total body K+ decreases as some of the K+ is excreted by the kidney. On relief of
asphyxia, the buffering processes are reversed and K+ leaves the plasma and reenters the
erythrocytes, leading to hypokalemia.
HIGH-RISK PREGNANCIES
Certain situations during pregnancy, labor, or delivery carry an increased risk of intrapartum
asphyxia. If these high-risk deliveries are identified before birth, their progress during labor and
delivery should be closely monitored and resuscitation can be initiated at birth. Tables 18-1 and 182 list some of the factors that alert the physician to a high-risk delivery. Optimal management of
these cases requires good communication between the obstetrician, anesthesiologist, and
pediatrician.
RESUSCITATION OF THE ASPHYXIATED INFANT

If a severely asphyxiated infant is expected, a resuscitation team must be present at delivery. In
the majority of instances, good communication between the obstetrician and pediatrician will
provide timely notice of the impending delivery of an asphyxiated infant. The actual extent of
resuscitation needed can be determined only after someone with considerable clinical experience
evaluates the infant's condition. It is helpful to assign the responsibility of each member of the
resuscitation team before delivery.
P.308
The following is an example of outlining the responsibilities of each member of the resuscitation
team.
TABLE 18-1 SOME FACTORS THAT PLACE THE NEWBORN AT HIGH RISK FOR ASPHYXIA
Labor and Delivery
Maternal Conditions
Diabetes mellitus
Preeclampsia, hypertension,
chronic renal disease
Anemia (i.e., hemoglobin <10
g/dL)
Blood type or group

alloimmunization
Abruptio placentae, placenta
previa, or other antepartum
hemorrhage
Narcotic, barbiturate,
tranquilizer, psychedelic drug
use or alcohol intoxication
History of previous perinatal

loss
Prolonged rupture of
membranes
Lupus
Maternal heart disease
Maternal fever or other
evidence of amnionitis
Abnormal umbilical artery
Doppler velocity
Conditions
Fetal Conditions
Forceps delivery other

Premature delivery
than low-elective or
Postmature delivery
vacuum-extraction delivery
Breech or other abnormal Acidosis determined by fetal
presentation and delivery scalp capillary blood
Cephalopelvic
Abnormal heart rate pattern or
disproportion: shoulder
dysrhythmia
dystocia, prolonged
second stage
Cesarean section
Meconium-stained amniotic
fluid
Prolapsed umbilical cord
Oligohydramnios
Polyhydramnios
Cord compression (e.g.,
nuchal cord, cord knot,
compression by
aftercoming head in
breech delivery)
Maternal hypotension or
hemorrhage
Decreased rate of growth:

uterine size or fetal size
determined by ultrasonography
Macrosomia
Immaturity of pulmonary
surfactant system
Fetal malformations
determined by sonography
Hydrops fetalis
Low biophysical profile
Multiple birth; in particular,
discordant, stuck, or
monamniotic
Member A:
Assess infant.
Manage airway and intubate the trachea, if needed.
Provide positive-pressure ventilation.
Secure endotracheal tube (ET).
Member B:
Listen for heart rate and, if needed, give cardiac massage.
Auscultate chest to be sure ET is in proper position and gas exchange is good.
Catheterize umbilical vessel or vessels and maintain patency of catheters.
Measure intravascular pressures, assess perfusion, sample blood for pH, PO2, and PCO2, and
draw blood cultures.
Administer fluids and drugs.
Continue assessment of infant.
TABLE 18-2 FETAL HEART RATE PATTERNS ASSOCIATED WITH FETAL AND NEONATAL
DISTRESS
Heart Rate Pattern
Severe (i.e., <80 beats/min), sustained

bradycardia, with loss of variability
Sustained tachycardia, uncomplicated by other
abnormal patterns
Late decelerations with loss of variability
Severe, recurrent variable decelerations, with loss
of variability
Sinusoidal
Member C:
Fetal or Neonatal Problems
Fetal hemorrhage, fetal asphyxia

Infection, often with apnea
Asphyxia
Asphyxia and possible hypovolemia
Severe anemia with asphyxia
Blot baby dry; apply electrocardiograph (ECG) monitor leads, radiant monitor servocontrol,
and transcutaneous oxygen sensor.
Keep timed, written record of resuscitation and vital signs and assign the Apgar scores at 1
and 5 minutes and every 5 minutes thereafter until the score is 7 or greater; time and
record the rate and volume of infusions such as alkali and blood volume expanders.
Assist member A by providing ET suction, adjusting the fraction of oxygen inspired (FiO2),
and helping to secure ET.
Help member B by providing medications and blood volume expanders in sterile syringes; B
is working in a sterile field early in resuscitation.
Monitor baby's temperature and capillary blood glucose.
Resuscitation Equipment and Supplies for Optimal

Resuscitation (Modified from Reference 1)
All resuscitation equipment (listed below) should be checked before each delivery and should be
fully operational.
Resuscitation table with heat source
Suction equipment
Bulb syringe
P.309
Mechanical suction and tubing
Suction catheters: 5 Fr, 6 Fr, 8 Fr, 10 Fr, and 12 Fr
8 Fr feeding tube and 20 mL syringe
Meconium aspirator
Bag and mask equipment
Neonatal resuscitation bag with a pressure-release valve or pressure manometer (the bag
must be capable of delivering 90% to 100% oxygen)
Face masks, newborn and premature sizes (cushioned-rim masks preferred)
Oxygen source with flowmeter (flow rate up to 10 L/min) and tubing
Intubation equipment
Laryngoscope with straight blades: No. 00 (very-low-birth-weight infant); No. 0 (preterm

infant); and No. 1 (term infant)
Endotracheal tubes, 2.5-, 3.0-, 3.5-, 4.0-mm internal diameter (ID)
Stylet (optional)
Scissors
Tape or securing device for ET
Alcohol sponges
CO2 detector (optional)
Laryngeal mask airway (optional)
Medications
Epinephrine 1:10,000 (0.1 mg/mL)-3-mL or 10-mL ampules
Isotonic crystalloid (normal saline or Ringer lactate) for volume expansion: 100 or 250 mL
Sodium bicarbonate 4.2% (5 mEq/10 mL)-10-mL ampules
Naloxone hydrochloride 0.4 mg/mL-1-mL ampules or 1.0 mg/mL-2-mL ampules
Dextrose 10%, 250 mL
Normal saline for flushes
Feeding tube, 5 Fr (optional)
Umbilical vessel catheterization supplies
Sterile gloves
Scalpel or scissors
Povidone-iodine solution
Umbilical tape
Umbilical catheters, 3.5 Fr, 5 Fr
Three-way stopcock
Syringes: 1, 3, 5, 10, 20, and 50 mL
Needles: 25, 21, and 18 gauge, or puncture device for needleless system
Miscellaneous
Gloves and appropriate personal protection
Radiant warmer or other heat source
Firm, padded resuscitation surface
Clock (timer optional)
Warmed linens
Stethoscope (neonatal head preferred)
Tape: or inch
Cardiac monitor and electrode or pulse oximeter and probe
Oropharyngeal airways (0, 00, 000 sizes or 30-, 40-, and 50-mm lengths)
Arterial and venous pressure monitor with waveform displays; transducers can be connected
to the catheters beforehand so that aortic pressure is displayed as soon as the umbilical
artery catheter is inserted, and the venous waveform can be used to localize the catheter tip
in the thoracic inferior vena cava (32)
Indirect blood pressure monitor
Tube thoracostomy tray with instruments and catheters from 10 to 14 Fr
Blood gas electrodes with a trained operator of the blood gas machines close enough to the
resuscitation area so that results are available in less than 5 minutes
In selected situations (see the section Special Problems), it is useful to have present in the
delivery room a unit of whole blood or packed erythrocytes that have been cross-matched against
the mother; this blood can be kept in a cold pack and returned to the blood bank if not used.
Clinical Assessment of Severity of Asphyxia

The Apgar score was the first attempt at a systematic assessment of birth asphyxia (6,33). There
is a loose correlation between low Apgar scores and umbilical cord blood gases. However, some
infants with severe acidosis have normal Apgar scores, and some with normal blood gases and pH
have very low scores (34,35). Maternal anesthetics, sedatives, maternal drugs, fetal sepsis, and
central nervous system pathologic conditions can lower the Apgar score; extremely premature
infants often have low scores without any other evidence of asphyxia (36,37). Regard-less of the
cause, an Apgar score that remains low calls for action. The clinical significance of the Apgar score
increases with time. Scoring should continue every 5 minutes until the score increases to 7 or
above. The length of time it takes to reach a score of 7 is a rough indication of severity of
asphyxia. Umbilical cord blood gases, discussed previously, are useful measures of fetal asphyxia,
but this information will not be available until a few minutes after birth, and resuscitation must be
started before that. Thus, their main value is in guiding subsequent management of the infant.
It is important to point out that the term perinatal asphyxia should be used with caution because
an inaccurate use of this term may have medicolegal ramifications. It should be reserved in the
clinical context of damaging hypoxemia and metabolic acidosis. A neonate who suffers from
hypoxemia proximate to delivery that is severe enough to result in hypoxic encephalopathy will
show other evidence of hypoxic damage, including severe metabolic or mixed acidemia (pH <7.00)
on an umbilical cord arterial sample; persistence of an Apgar score of 0 to 3 for longer than 5
minutes; neonatal neurologic sequelae, for example, seizures, coma, hypotonia; and evidence of
multiorgan system dysfunction. However, because neonatal neurologic problems may be attributed
to factors other
P.310
than those involving labor and delivery, the term neonatal depression should be used when cause
is uncertain.
The following resuscitation approach is based mostly on the recommendations of the American
Academy of Pediatrics and the American Heart Association for neonatal resuscitation (1). However,
it must be pointed out that most of these recommendations are based on anecdotal experience
and a review of the available literature, but have not been rigorously tested. It is expected that in
the future some of these recommendations will undergo close scrutiny and possible change.
Initial Steps in Resuscitation

Initial assessment, performed within a few seconds of delivery, dictates the extent of resuscitation
needed. Initial assessment includes determining whether or not the infant is breathing, if there is
good muscle tone, if the color is pink, and whether the infant looks term or preterm. If the infant
is term, vigorous, without any known risk factors, and born through clear amniotic fluid, the infant
need not be separated from the mother to receive initial care. Thermal care can be given by
putting the infant on the mother's chest (direct skin-to-skin contact), drying the infant, and
covering the infant with dry linen. If the infant is apneic, gasping, has decreased muscle tone, or is
cyanotic, immediate resuscitation is needed. Place the infant under a radiant warmer; quickly
towel dry the baby; open the airway by laying the infant in the sniffing position; suction the mouth
first and then the nose; provide tactile stimulation (by gently slapping or flicking the soles of the
feet or by gently rubbing the back); and, if necessary, give oxygen. In the majority of instances,
with these initial steps, the infant will start breathing adequately and demonstrate a color change
to pinkish. If the infant starts breathing adequately, but continues to have central cyanosis,
provide free-flow 100% oxygen. However, if the infant does not start breathing adequately or has
a heart rate of less than 100 beats per minute, positive pressure ventilation (PPV) should be
instituted immediately. The entire process up to this point should not take more than 20 to 30
seconds.
Positive Pressure Ventilation

Ensuring adequate ventilation is the single most important step in the cardiopulmonary
resuscitation of the compromised newborn infant (38). Indications for initiating PPV include (a)
apnea or gasping breaths; (b) heart rate of less than 100 beats per minute (bpm); and (c)
persistent central cyanosis despite the provision of free-flow 100% oxygen. Initial PPV can be
provided by either a flow-inflating bag (also called an anesthesia bag) or a self-inflating bag. Each
type has its own advantages and disadvantages, and the resuscitator should be familiar with the
type of bag used in his or her institution. If PPV is needed for a term infant, provide bag-and-mask
ventilation at a rate of 40 to 60 breaths per minute with initial pressures of as high as 30 to 40 cm
H2O. Subsequently, however, one should ventilate with the lowest pressures required to produce a
gentle chest rise. Begin ventilation by slowly applying a pressure of 20 to 30 cm H2O. Maintain this
inflating pressure for 1 to 2 seconds, then ventilate at a rate of 40 to 60 breaths per minute, using
an inflation time of 0.3 to 0.5 seconds and enough pressure to provide a visible rise of the upper
portion of the chest. Repeat the application of the initial inflating pressure pattern three to four
times over the first 2 minutes. These prolonged breaths inflate regions of the lungs that were
gasless and create necessary FRC (39). Figure 18-2 illustrates this process. Care must be taken to
avoid rapid inflation or overdistension of the lungs.
It is important to realize that infants respond to initial lung inflation by eliciting a variety of
physiologic responses. The infant may respond by the rejection response, in which the infant
responds to PPV with a positive intraesophageal pressure to resist the inflation, that is, the infant
actively resists attempts to inflate the lungs by generating an active exhalation. This response acts
to not only reduce lung inflation, but also may cause high transient inflation pressures. Another
response is Head's paradoxical response in which the neonate responds to PPV with an
inspiratory effort, causing a negative intraesophageal pressure. This inspiratory effort, with the
resultant negative pressure produces a fall in inflation pressures but results in a transient increase
in tidal volume. Of course, the neonate may demonstrate no response to the inflation attempt,
that is, not generating any change in intraesophageal pressure during the positive pressure
inflation, and passive inflation subsequently results. It is important to recognize that these
physiologic responses to positive pressure inflation in the delivery room may cause large variability
in the tidal volume and intrapulmonary pressures, despite constant delivery of inflation pressure.
The very premature infant with a very small lung volume and a surfactant-deficient lung presents
special problems for initial ventilation that are discussed later (see section Special Problem).
Whether low tidal volume ventilation starting in the delivery room, in an attempt to avoid lung
injury and hypocapnia in premature infants, can be safely delivered, and whether it is as
efficacious as normal tidal volume ventilation remains to be seen.
In the majority of instances, provision of appropriate PPV is followed by an increase in heart rate,
improvement in color, and spontaneous breathing. Rate and pressures of PPV should be gradually
reduced before deciding to see if the infant will tolerate its discontinuation. Free-flow oxygen may
be continued as long as necessary to keep the infant pink. Some of the factors determining the
success of PPV using a bag and mask include choosing the correct size mask, proper positioning of
the infant, achieving tight seal between the face and the mask, and using adequate inspiratory
pressure. If despite correct bag-and-mask PPV the infant fails to improve or continues to
deteriorate, consider beginning chest compressions and bag-and-endotracheal-tube ventilation.
Figure 18-2 Initial inflation of the lungs by assisted ventilation in two asphyxiated infants. A: An
inflation pressure of 30 cm H2O is applied repeatedly for 1 to 2 seconds. For the first several
breaths, the volume entering and leaving the lungs is the same. Then the volume out is slightly
less than the volume in for several breaths, and this trapped gas begins to form the functional
residual capacity (FRC). B: The first inflation is with a pressure that is increased slowly up to 30
cm H2O over 8 seconds and held at that pressure for 2 seconds. During exhalation, less gas
leaves than what entered; therefore, some FRC has been generated with the first breath. Pair,
pressure applied to the airway; VT, tidal volume. (Adapted from Boon AW, Milner AD, Hopkin IE.
Lung expansion, tidal exhange, and formation of the functional residual capacity during
resuscitation of asphyxiated neonates. J Pediatr 1979;95: 1031; and Vyas H, Milner AD, Hopkin
IE, et al. Physiologic responses to prolonged and slow-rise inflation in the resuscitation of the
asphyxiated newborn infant. J Pediatr 1981:99:635.)
P.311
Chest Compressions
Chest compressions are infrequently needed during neonatal resuscitation. However, infants who
have a heart rate of less than 60 bpm, despite 30 seconds of effective PPV, need immediate chest
compressions (Fig. 18-3) Chest compressions can be provided by either the thumb technique or
the two-finger technique, the thumb technique being preferred. Chest compressions are performed
by placing thumbs or finger on the sternum immediately above the xiphoid and compressing at a
rate of 90 per minute with an accompanying breath rate of 30 per minute (chest-compressions-toventilation ratio = 3:1). To provide effective chest compressions, one should ensure that the depth
of compressions is one-third the depth of the chest, thumbs or fingers remain in contact with the
chest at all times, the duration of the downward stroke of the compression is shorter than that of
the release, compressions are well coordinated with ventilation, and there is adequate chest
movement during ventilation. Continue chest compressions until the heart rate is greater than 60
bpm
P.312
and ventilate until heart rate is greater than 100 bpm, at which point the infant's own respiratory
effort is assessed for adequacy. Ventilation can then be discontinued if heart rate remains above
100 bpm and the infant continues to breathe spontaneously. However, if the heart rate remains
below 60 bpm despite effective chest compressions and ventilation, ET intubation (if it has not
been already done) and administration of medications may be needed. However, it is important to
realize that the most important reason for continued bradycardia during resuscitation is failure to
establish effective postnatal ventilation rather than perinatal asphyxia. Therefore, it important to
make every effort to optimize ventilation before chest compressions and medication administration
are considered.
Figure 18-3 Resuscitation and cardiac massage were performed on a 2.1-kg infant who was
delivered by cesarean section because of signs of fetal asphyxia at 34 weeks of gestation. The
infant was intubated and ventilated with 60% oxygen beginning 30 seconds after birth. An
electrocardiogram was begun at 1 minute, and at 2.5 minutes an umbilical artery catheter
connected to a pressure transducer and a recorder was passed into the descending aorta. Note
persistent bradycardia despite assisted ventilation and low aortic pressure with narrow phasic
pressure. Cardiac massage raised heart rate and pressure. When briefly discontinued after 1
minute, pressure and heart rate fell. After another minute of massage and assisted ventilation,
good cardiac output had returned. This was manifested by a sustained higher heart rate and
higher blood pressure with wider phasic pressure when massage was discontinued a second time
5 minutes after birth. By 8.5 minutes, the infant was still acidotic, but there was adequate
oxygenation and aortic pressure continued to rise. PaCO2, arterial carbon dioxide partial pressure;
PaO2, arterial oxygen partial pressure; SaO2, saturation of arterial blood hemoglobin with oxygen.
Endotracheal Intubation
Endotracheal intubation is probably the most difficult step in neonatal resuscitation and the one
that requires constant practice to maintain sharp skills in this technique. Someone experienced in
endotracheal intubation should be available to assist at every delivery. Therefore, if the
resuscitator is not comfortable with endotracheal intubation, he/she should focus on providing
effective ventilation via bag and mask until somebody experienced in endotracheal intubation
arrives. Indications for endotracheal intubation include (a) to suction meconium; (b) to improve
ventilation when bag-and-mask ventilation is ineffective; (c) to coordinate ventilation and chest
compressions; (d) to administer medications such as epinephrine; (e) when prolonged ventilation
is needed, for example, extreme prematurity; (f) to administer surfactant; and (g) when
congenital diaphragmatic hernia is suspected (probably the only absolute indication for
endotracheal intubation).
The correct size laryngoscope blade is No. 1 for a term and No. 0 for a preterm infant. For a verylow-birth-weight (VLBW) infant, a size 00 blade may be used. The ET should be of uniform internal
diameter as tubes with shoulders may obstruct the line of vision during insertion and are more
likely to cause trauma to the vocal cords. The choice of ET is based on the infant's weight or
gestational age: for infants weighing less than 1000 g (<28 weeks' gestational age), use a 2.5-mm
internal diameter (ID) tube; for infants weighing 1000 to 2000 g (28 to 34 weeks' gestational
age), use a 3.0-mm ID tube; for infants weighing 2000 to 3000 g (34 to 38 weeks' gestational
age), use a 3.5-mm ID; and for larger infants, a 3.5- to 4.0-mm ID tube may be used. To
minimize the risk of vocal cord trauma and the subsequent development of laryngeal stenosis, one
should make sure that the inserted ET is not too tight a fit in the larynx. In general, gas should
leak from the space between the ET and the trachea when 15 to 30 cm H2O pressure is applied to
the airway. Cutting the tube at the 13- to 15-cm mark prior to insertion makes it easier to handle
during the procedure, decreases the chances of inserting the tube too far, and decreases the
resistance to airflow. The use of a stylet during ET insertion is optional. If a stylet is used to stiffen
the ET, it should be ensured that its tip is approximately 0.5 cm above from the tip of the tube and
does not protrude from the end or the side hole of the ET. In addition, the stylet should be well
secured so that it does not advance farther into the ET during intubation. If the stylet extends
beyond the tip of the ET, it could traumatize the airway.
For intubation, stabilize the infant's head in the sniffing position, slide the laryngoscope into the
mouth and advance it gently beyond the base of the tongue, lift the blade, and look for landmarks,
suctioning if necessary. Once the vocal cords are visualized (as an inverted letter V [] within the
glottis), insert the ET gently until the vocal cord guide is at the level of the vocal cords. Application
of downward pressure on the cricoid may facilitate intubation. Preoxygenating the infant before
attempting intubation, delivering free-flow oxygen during intubation, and limiting actual intubation
attempts to no longer than approximately 20 seconds minimizes hypoxia associated with
intubation. If intubation is unsuccessful in 20 seconds, ventilate the infant by bag and mask for at
least 1 minute before again attempting intubation. Other complications of intubation include apnea
and bradycardia; contusion or laceration of the tongue, gums, or airway; perforation of the
trachea or esophagus; infection; and pneumothorax. With all these complications possible,
intubate gently.
If the ET is inserted to aspirate meconium, immediately connect a meconium aspirator and apply
suction as the tube is gradually withdrawn. Repeat the procedure as necessary until little additional
meconium is suctioned or until the infant's heart rate indicates that PPV is required. If the ET is
inserted to ventilate the infant, after insertion, immediately ensure that the ET is in the trachea by
observing the chest rise with each breath, listening for good breath sounds with each breath, the
absence of breath sounds over the stomach, and observing vapor condensation on the inside of
the tube during exhalation. If there is still any doubt whether the tube is in the esophagus or
trachea, a color change on a CO2 detector attached to the ET may be helpful. However, be aware
that extremely premature infants and infants with very poor cardiac output may exhale insufficient
CO2 to be detected reliably by CO2 detectors. Correct placement of the ET within the trachea is
guided by the lip-to-tip distance in centimeters (6 + the infant's body weight in kilograms), which
places the ET tip midway between the vocal cords and the carina. Correct placement is further
suggested by listening to equal air entry on both sides of chest, in the axillae. If the breath sounds
are louder on the right side of chest, the ET is likely to be in the right main stem bronchus and
should be slowly pulled back until equal breath sounds are heard over both sides of chest before it
is taped in place. If the ET is to be left in place beyond initial resuscitation, final confirmation of
appropriate placement should be made through a chest radiograph.
Medications
More than 99% of infants requiring resuscitation improve with timely and skillful implementation of
initial steps of
P.313
resuscitation and establishment of effective ventilation. Only a small fraction of infants require
medications during resuscitation (2). Each delivery area should have a chart displaying the
appropriate dosages and concentrations of all medications used in neonatal resuscitation. All
medications needed for resuscitation can be administered via the ET or via an umbilical venous or
an arterial catheter.
Epinephrine is usually the first drug administered during resuscitation. It is indicated when the
heart rate remains less than 60 bpm after 30 seconds of adequate PPV and another 30 seconds of
chest compressions and PPV. It is administered either via an ET or umbilical vein (40).
Administration via the umbilical route is probably more effective, but is often delayed because of
the time required to establish an umbilical venous access. Therefore, during neonatal resuscitation,
administration via an ET is generally the most rapidly accessible and practical route for epinephrine
administration. The recommended dose is 0.1 to 0.3 mL/kg of a 1:10,000 solution, given rapidly.
In addition to its 1 effect on the heart, it also has a peripheral 1 effect that results in peripheral
vasoconstriction and increased blood flow to the heart and brain. The heart rate should increase
promptly after epinephrine has been administered; if it does not, epinephrine can be repeated
every 3 to 5 minutes. Epinephrine should not be given before establishing adequate ventilation,
because in the absence of available oxygen it may cause myocardial damage by increasing the
workload and oxygen consumption of the heart muscle. One area of controversy is the use of high-
dose epinephrine during neonatal resuscitation. Because of the theoretical risks of higher doses
coupled with lack of any human data proving its added advantage, higher doses are not currently
recommended for neonatal resuscitation.
If the infant looks pale and there is evidence of blood loss, administration of a volume expander is
indicated. Although it is most commonly given through the umbilical vein, the intraosseous route
can be used (41). The recommended solution for treating hypovolemia during neonatal
resuscitation is an isotonic crystalloid solution (normal saline or Ringer lactate at 10 mL/kg).
Whether albumin can be given safely in this situation, and whether it is as effective as normal
saline is not known. O-negative blood crossmatched with the mother's blood (if available) can be
transfused. This should be prepared before delivery if low fetal blood volume is suspected
antenatally. A volume expander can be repeated if the infant shows minimum improvement after
the first dose and evidence of hypovolemia persists (persistent pallor despite adequate
oxygenation, weak pulses, and poor response to adequate resuscitation). Although hypovolemia
should be corrected fairly quickly, there is concern that too rapid infusion of the volume expander
may result in intracranial hemorrhage, especially in the preterm infant. Therefore, each infusion
should be given over 5 to 10 minutes.
The use of sodium bicarbonate (NaHCO3) during resuscitation is controversial. The current
recommendation is to use NaHCO3 if the infant has undergone all other steps of resuscitation
appropriately and has failed to respond. Although there is a lack of data documenting its benefit, if
the pH is less than 7.05 as a consequence of a mixed acidosis, or the base deficit is 15 mEq/L or
more, correction of the metabolic component of the acidosis with an infusion of NaHCO3 may be
helpful. The immediate objectives are twofold: to reverse the myocardial failure and low cardiac
output that occurs from acute metabolic, but not respiratory, acidosis (22,42,43); and to relieve
the intense pulmonary vasoconstriction that occurs with severe acidosis, particularly in full-term
infants (19,44,45). The degree of pulmonary vasoconstriction is approximately the same for
metabolic and respiratory acidosis (46). The recommended dose is 2 mEq/kg [4 mL/kg of 4.2%
solution, osmolarity = 900 mOsm/L]) given slowly (no faster than 1 mEq/kg per minute). Some
studies suggest that there is an association between rapid infusions of large volumes of
concentrated NaHCO3 and intracranial hemorrhage in preterm infants. The hemorrhages might be
caused by transient hypernatremia from too rapid an infusion, by an acute rise in PaCO2 from
inadequate ventilation during the NaHCO3 infusion, which would cause cerebral vasodilation, or by
the asphyxia for which the NaHCO3 was given. However, infusion of NaHCO3 into the inferior vena
cava at the rate of 1 mEq/kg per minute, for a total dose of up to 5 mEq/kg, causes only a slight
transient increase in the arterial sodium concentration.
The ability of NaHCO3 buffer to raise pH depends on the ability of the lungs to eliminate the CO2
produced by the buffering process, as determined by the following equation:
Consequently, NaHCO3 should not be given unless ventilation is adequate and PaCO2 is low,
normal, or declining toward normal. Continue ventilation during bicarbonate therapy to eliminate
the excess CO2 produced. CO2 is highly diffusible, so even if ventilation is adequate, some of the
CO2 produced by buffering could enter cells and transiently increase acidosis. The implications of
this intracellular acidosis following sodium bicarbonate infusion are not known, and remain to be
studied.
Tris(hydroxymethyl)-aminomethane (THAM acetate), although generally not used during
resuscitation, may also be helpful in correcting metabolic acidosis. It has the dual advantage of
reducing PaCO2 and buffering metabolic acid. It is most useful for treating infants with severe
mixed metabolic and respiratory acidosis and for situations of severe asphyxia with suspected
extreme acidosis in which blood gas measurements are not available. It may cause respiratory
depression, and so should be used only in situations in which ventilation already is assisted. Also it
may cause hypoglycemia. Further, it should not be used in patients who are anuric or uremic. An
earlier preparation was very hyperosmolar, highly alkaline and tended to cause sclerosis of blood
vessels. These problems have been corrected by the use of the 0.3 mol/L preparation (can be used
without dilution), which also is adjusted to pH 8.6. Figure 18-4 illustrates correction of a severe
mixed acidosis.
P.314
Figure 18-5 shows how the cardiovascular effects of an infusion of alkali differ depending upon the
severity of the asphyxia.
Figure 18-4 The sequence of events following the onset of ventilation. (From Smith CA, Nelson
NM. Physiology of the newborn infant. 4th ed. Springfield, IL: Charles C. Thomas, 1976:131.)
If the infant's respiratory depression is suspected to be caused by maternal narcotics (e.g.,

morphine, meperidine, or butorphanol tartrate), administered within 4 hours prior to delivery,
administration of naloxone hydrochloride (Narcan), a narcotic antagonist should quickly reverse
the respiratory depression. However, prompt and adequate ventilatory support should always
precede the administration of naloxone and should be continued until the infant is breathing
normally. As the duration of action of narcotics often exceeds that of naloxone, continued close
observation is necessary and repeated doses of naloxone may be required. Because naloxone can
precipitate an acute withdrawal reaction, it should not be given to the infant of a mother who is
suspected of being addicted to narcotics or is on methadone maintenance. The recommended dose
of naloxone hydrochloride is 0.1 mg/kg (1.0 mg/mL solution). It should be given preferably
endotracheally or intravenously; however, intramuscular and subcutaneous routes are also
acceptable, although they have a delayed onset of action.

Infants who have had severe asphyxia and who do not respond promptly to adequate resuscitation
are likely to have myocardial depression with decreased cardiac output and hypotension. These
infants may respond to dopamine starting at 5 g/kg per minute and increasing to 15 to 20 g/ kg
per minute, as needed. Rarely, dobutamine might need to be added (starting at 5 g/kg per
minute and increasing to 15 to 20 g/kg per minute). During vigorous resuscitation it is useful to
monitor heart rate and oxygen saturation electronically and have ECG electrodes applied.
Figure 18-5 Changes occurred in heart rate, aortic blood pressure, and arterial blood gas
tensions during the first 45 minutes after birth of a 1.2-kg premature infant with severe asphyxia
complicated by bilateral pleural effusions. The child's trachea was intubated immediately after
birth, and he was manually ventilated with 100% oxygen throughout this time. Note the severe
mixed acidosis in the first blood gas measurement at 11 minutes after birth. Administration of
NaHCO3 at this point would have been inappropriate and ineffective because assisted ventilation
had not yet achieved adequate elimination of CO2. NaH CO3 was given only after adequate CO2
elimination was achieved. Note that there was no rise in the arterial carbon dioxide partial
pressure (PaCO2) after this, indicating that all the CO2 produced during the buffering process was
eliminated, and the only change was a reduction in base deficit form -14 to -5 mEq/L, which
raised the pH from 7.06 to 7.22. Note the high initial aortic pressure, which was due to the
vasoconstriction of asphyxia, indicating that myocardial failure had not yet developed. As
asphyxia was relieved, aortic pressure fell to normal. (BE, base excess; PaO2, arterial oxygen
partial pressure.)
Umbilical Vessel Catheterization

Catheterization of umbilical vessels is invaluable during neonatal resuscitation. The umbilical vein
is readily accessible in the newborn and can be quickly catheterized. It is most often used for the
administration of medications and volume expanders. For these purposes, the catheter should be
inserted only 3 to 5 cm, or until free flow of blood is established, to avoid placing the tip in the
portal circulation. The umbilical artery can be catheterized to obtain blood for arterial blood gas
and other samples as well as to monitor blood pressure. The umbilical arterial catheter can be
connected to a precalibrated pressure transducer, so that blood pressure can be monitored
continuously. Alternatively, it can be connected to the transducer ahead of time and the blood
pressure displayed as soon as the catheter is passed into the aorta. This also reduces the risk of
accidentally injecting air bubbles through the catheter into the infant's circulation while the
catheter is being connected to the transducer.
P.315
It is important to obtain quickly a measurement of pH and PaCO2 to determine that ventilation is
neither inadequate nor excessive and to detect metabolic acidosis. Although a venous blood gas
analysis is not as representative as an arterial one, it will suffice as an initial measurement to
detect severe acidosis and to determine whether or not PaCO2 is near the normal range. Venous
PaCO2 is about 6 mm Hg higher and pH is about 0.03 units lower than in arterial blood (32).
Other aspects of infant's care that need to be closely followed are discussed in the following
sections.
Hematology
Consumption of coagulation factors may complicate severe asphyxia. This is almost always a
transient process rather than continuing disseminated intravascular coagulation.
Thrombocytopenia is the most consistent finding. In extreme cases, clinical bleeding occurs and
requires replacement of platelets and plasma clotting factors. Other hematologic changes include a
transient rise in the number of granulocytes, including immature forms, and in erythroid
precursors in the peripheral blood. These can rise to very high levels and could be misleading
when considering diagnoses such as infection and hemolytic anemia. If the changes are secondary
to asphyxia, however, they will disappear in a day or so.
Glucose
When hypoxia and acidosis have been relieved, begin a continuous infusion of 10% dextrose in
water at 3 mL/kg per hour to maintain a normal concentration of blood glucose. This provides 5
mg of glucose per kg per minute. Begin screening for hyperglycemia and hypoglycemia with
repeated testing of capillary blood. Hypoglycemia can be corrected by a bolus infusion of 10%
dextrose (2 mL/kg over 2 to 3 minutes) and by temporarily increasing the infusion rate to 5 mL/kg
per hour (8.4 mg/kg per minute), which is sufficient in all but the most extreme cases of asphyxiainduced hypoglycemia. Rapid infusions of more concentrated solutions of dextrose can be
dangerous because of their hyperosmolarity and their tendency to produce serious vascular injury.
Fluid and Electrolytes

Asphyxia causes renal ischemia, which may persist after birth and resuscitation (47). Many
asphyxiated infants are oliguric for the first day of life and have a rising serum creatinine (48,49).
Very high serum creatinine concentrations in the first days of life, however, are not caused by
renal failure but are caused by tissue necrosis after severe asphyxia. When urine output increases,
there is transient hematuria and polyuria. With extreme asphyxia, corticomedullary hemorrhagic
necrosis occurs and the renal failure is more severe. As renal function returns, sodium, potassium,
and chloride can be added to parenteral fluids. As acidosis is corrected, extracellular potassium
shifts back into cells and postasphyxial infants with good renal function may become hypokalemic.
The usual maintenance amounts of electrolytes may be adequate to maintain serum electrolyte
concentrations, but in many cases renal losses of electrolytes are very high during this diuresis, so
it will be necessary to administer higher concentrations of electrolytes. When planning sodium
requirements, consider the sodium given as NaHCO3 during resuscitation and the NaCl in catheter-
flush solutions. Infants who are asphyxiated often become hypocalcemic by the first day after birth
(31). Serum ionized calcium should be measured and supplemental calcium given as needed.
Gastrointestinal Function
During asphyxia, blood flow to the small and large bowel is reduced. Severe asphyxia may cause
serious ischemic injury to these organs and gastrointestinal blood flow may remain abnormal for
up to 3 days after delivery and resuscitation (47). Because of this, it may be advisable to continue
intravenous fluids and delay enteral feedings for several days. Occasionally, acute necrotizing
enterocolitis occurs when severely asphyxiated infants are fed in the first day or two after birth.
This is particularly important in infants who also have suffered hypovolemic shock, because shock
severely compromises intestinal blood flow.
Central Nervous System

Severe asphyxia can lead to hypoxic ischemic encephalopathy. Because the homeostatic responses
to asphyxia tend to preserve oxygen delivery to the brain at the expense of other organs,
asphyxial encephalopathy tends to occur in infants with evidence of multiorgan injury (47,49).
However, there are situations in which encephalopathy appears in isolation. A significant portion of
the damage to the brain is reperfusion injury, which occurs some time after recovery from
asphyxia (50,51). This interval raises the potential for treatment that could prevent this portion of
the injury. Several forms of therapy that might provide such neuroprotection are under
investigation, but none has been shown to be effective (50). However, there are some lessons
from these studies that are applicable to clinical care. Many studies point to hypocarbia as a
contributor to neurologic injury (50,52). Closely monitoring PaCO2 and adjusting ventilation to
avoid hypocarbia has been discussed previously. There is much conflicting information about the
beneficial versus the harmful effects of hyperglycemia in the asphyxiated infant. So it seems
prudent to keep the blood glucose concentration in the physiologic range. Recent experimental
studies have shown that moderate cerebral hypothermia initiated as soon as possible in the latent
phase, before the onset of secondary injury, and continued for 48 hours or more, is associated
with potent, long-lasting neuroprotection. There are several recently published and ongoing
randomized clinical trials to test the safety and efficacy of whole-body hypothermia and selective
head
P.316
cooling (53,54). Current data suggest that whole-body cooling is feasible in term neonates, with
no life-threatening adverse events, but improvements are needed to maintain stable hypothermia
for prolonged periods, and its impact on long-term outcome remains to be seen (55).
DISCONTINUATION OF RESUSCITATION
Discontinuation of resuscitation in an infant with cardiorespiratory arrest should be strongly
considered, and may be appropriate, if despite all the steps of resuscitation, heartbeat remains
absent after 15 minutes. Current data support the position that after 10 minutes of asystole, a
newborn infant is very unlikely to survive, and if it survives, is likely to have severe neurologic
compromise (1,56,57,58). Of course, parents should have a major role in determining the care of
their newly born infant and the extent of resuscitation offered. Based on the data at hand, every
effort should be made to plan the resuscitation approach before delivery, with the provision that
the plan may change according to the infant's condition at delivery and the infant's response to
resuscitative efforts.
Situations When Noninitiation of Resuscitation is

Reasonable
In recent years there has been considerable debate over the concept of the lower limit of viability
and other circumstances when noninitiation of resuscitation can be recommended. Although there
are still no clear answers to such a sensitive subject, noninitiation of treatment may be appropriate
for infants with confirmed gestation of less than 23 weeks or a birthweight of less than 400 g,
anencephaly, or confirmed trisomy 13 or 18. Present data support the position that despite
resuscitation these infants are very unlikely to survive, or if they do survive, it will be with severe
disability. However, in situations of uncertain prognosis, including uncertain gestational age
(antenatal prediction of gestational age is variable; see Chapter 11), resuscitation options may
include a trial of resuscitation, noninitiation, or discontinuation of resuscitation following
assessment of the infant. If after initial assessment, there is still any doubt, full resuscitation and
continuation of life support can allow time to obtain more complete clinical information and the
family's viewpoint, and will allow for counseling to the family regarding continuation versus
withdrawal of support. Then, if agreed on, withdrawal of support may be appropriate. In general, it
is a bad idea to decide initially not to resuscitate and then, many minutes later, change the
decision to aggressively resuscitate.
Continuing Supportive Care After Resuscitation

After adequate ventilation and circulation have been established, the infant may still be at risk for
deterioration. Consequently, continuing close monitoring of heart rate, respiratory rate, blood
pressure, arterial oxygen saturation, blood gas analysis, blood glucose, and the administered
oxygen concentration may be indicated.
Monitoring Respiratory Status

During this time, the infant should be closely monitored for complications of assisted ventilation.
The ET may be dislodged from the trachea and advance into the esophagus. Alternatively, the tube
may advance into the right main bronchial stem, leading to nonventilation of the entire left lung
and the upper lobe of the right lung. This is the most common serious complication of
endotracheal intubation. Clinical suspicion may need to be confirmed by a repeat chest radiograph.
Examine the abdomen for distension that may elevate the diaphragm and interfere with
ventilation. This usually is a consequence of assisted ventilation via a face mask and can be
relieved by an orogastric tube.
Following successful resuscitation, pulmonary function may change rapidly, leading to several
complications. First, PaO2 will rise as ventilation and perfusion become better matched and may
reach levels that are dangerously high in the very premature infant. Hyperoxia is better managed
initially by reducing the inspired oxygen concentration rather than by withdrawing ventilatory
assistance. Second, improved ventilation may lead to acute hypocarbia, which will reduce cerebral
and myocardial blood flow (59,60,61). Correct the hypocarbia by reducing the rate of assisted
ventilation and inspiratory pressure. Figure 18-6 illustrates these changes. Third, as lung
compliance increases, the ventilatory pressure that was appropriate initially will become excessive.
If the excess pressure is mild, the result will be hyperventilation and hypocarbia. If it is extreme,
however, there will be tamponade of the pulmonary circulation and right-to-left shunting of blood
at the atrial and ductal levels and low systemic blood flow. This phenomenon is manifested by a
low aortic pressure, wide fluctuations in blood pressure in phase with the positive-pressure
ventilation, and arterial hypoxemia. This situation can be detected by briefly disconnecting the ET
from the ventilation system. Aortic pressure will rise within 5 seconds. If this occurs, restart
ventilation using lower airway pressures. The hypoxia caused by the shunting of blood from right
to left will decrease quickly as the lungs are ventilated with lower pressures.
Tension pneumothorax may occur during spontaneous or assisted ventilation of any infant. A
tension pneumothorax of small or moderate size may restrict ventilation and cause hypoxia and
hypercarbia. Pneumothorax must be suspected whenever PaO2 decreases despite a ventilation
system that is functioning properly. Sometimes the diagnosis of pneumothorax is difficult to make
by physical examination alone. Breath sounds may be unequal bilaterally, but often they are equal.
The upper portion of the affected side of the chest tends to lag behind the unaffected side during
inflation of the lungs. Transillumination with a cold fiberoptic light may cause the affected side to
glow brightly; however, the absence of this sign does not rule out pneumothorax, particularly in
the larger infant with a
P.317
thicker chest wall. The diagnosis of pneumothorax can be made best by a chest radiograph, but
this often is difficult to obtain quickly in the resuscitation area. The arterial and central venous
pressures may not change with small pneumothoraces. If hypoxia and hypercarbia become severe,
it may be necessary to perform a diagnostic thoracentesis with a small-gauge angiocatheter and
syringe before there is time to obtain a radiograph. The situation changes when a pneumothorax is
large and under tension. Venous return to the heart and cardiac output may fall precipitously to
extremely low levels. If blood pressure, PaO2, and PaCO2 are being measured, this critical situation
will be diagnosed easily because the onset of hypoxia and hypercarbia will be accompanied by
severe hypotension rather than the hypertension of asphyxia (see Fig. 18-1) (62). This situation
warrants intervention, which is as urgent as in cardiac arrest. One cannot wait for a confirmatory
radiograph. Figure 18-7 illustrates the diagnosis and successful treatment of such a case.
Satisfactory decompression of a tension pneumothorax usually requires insertion of a
thoracostomy tube and continuous suction applied to the tube through an underwater suction
system. Aspiration with a needle and syringe usually gives only very brief relief. While assembling
equipment for decompression of the pneumothorax, however, insert a 22-gauge Angiocath
connected to a three-way stopcock and a 30-mL syringe. This is a convenient and relatively safe
method for temporary decompression of the pneumothorax.
Figure 18-6 Changes in arterial blood gas tensions occurred during the resuscitation of two verylow-birth-weight infants. Each was intubated immediately after birth, and in each the umbilical
artery was catheterized before 10 minutes after birth to allow frequent measurements of blood
gasses. Both were hypoxic and hypercarbic at the first measurements at 6 and 7 minutes after
birth. As ventilation and oxygenation improved, ventilation pressures, rates, and inspired oxygen
concentration were reduced. In the baby in the upper panel, this led to normal blood gas
tensions. In the baby in the lower panel, adjustments were made too slowly, leading to hyperoxia
and extreme hypocarbia. PaCO2, arterial carbon dioxide partial pressure; PaO2, arterial oxygen
partial pressure.
Figure 18-7 Aortic blood pressure of a premature, 1.5-kg infant at 32 gestational weeks of age
during development of a tension pneumothorax. This is a continuous tracing at 2 hours of age.
Because the patient's condition was rapidly worsening, as shown by hypotension, a narrow pulse
pressure, and rapidly worsening, as shown hypotension, a narrow pulse pressure, and rapidly
increasing cyanosis despite assisted ventilation with 100% oxygen, thoracentesis of the right
pleural cavity was done (arrow) before radiologic confirmation of the pneumothorax was obtained.
About 50 mL of air escaped when the pleural cavity was opened. The patient's color improved,
and blood pressure promptly returned to normal.
Pulmonary function will improve rapidly in many infants as compliance improves with absorption of
lung water. Pulmonary perfusion will increase in response to a rising pH and PaO2. On the other
hand, if ischemia has caused more severe asphyxia with lung injury, there may be continued
respiratory distress that is indistinguishable from early hyaline membrane disease. This will require
continued ventilatory assistance. Unlike hyaline membrane disease, however, this form of
respiratory failure usually begins to improve within a few hours after birth (63), whereas hyaline
membrane disease as a result of immaturity of the surfactant system worsens over the first day
after birth. When a fetus with immature lungs has suffered significant intrapartum asphyxia, the
ensuing hyaline membrane disease generally will be more severe (64). Figure 18-8 illustrates
these divergent courses of respiratory distress.
Infants with early onset hyaline membrane disease should be given exogenous surfactant as soon
as the condition is apparent and the ET is in proper position with its tip above the carina. Early
treatment with surfactant is more effective than treatment that has been delayed several hours
(65). At this early stage in the infant's course, it often is impossible to distinguish between hyaline
membrane
P.318
disease, postasphyxial respiratory distress, and congenital pneumonia. This is not a reason to
withhold surfactant treatment, because it has virtually no adverse effects (under proper
monitoring) and may be of some benefit in some infants with other pulmonary diseases (66,67).
Figure 18-8 The course of respiratory distress is characterized by changes in mean airway
pressure in three groups of infants: those with severe perinatal asphyxia but no hyaline
membrane disease; those with perinatal asphyxia plus hyaline membrane disease; and those with
no perinatal asphyxia but hyaline membrane disease. In both groups with hyaline membrane
disease, the disease worsens, as indicated by the increased mean airway pressure required over
the first 24 hours. Those who also had asphyxia had more severe disease. Those with severe
asphyxia but no hyaline membrane disease had a completely different course, with progressive
improvement over the first 24 hours of life. ^, No respiratory distress syndromesevere acidosis;
^, respiratory distress syndromesevere acidosis; , respiratory distress syndromemild or no
acidosis. (From Thibeault DW, Hall FK, Sheehan MB, et al. Postasphyxial lung disease in newborn
infants with severe perinatal acidosis. Am J Obstet Gynecol 1984;150:393.)
Monitoring Circulatory Status

Asphyxia disrupts the newborn's ability to compensate for the loss of large amount of blood
volume (68). However, most asphyxiated infants have a normal or greater-than-normal blood
volume; only a few have a low blood volume (28). Consequently, hypovolemic shock does not
develop in most asphyxiated infants. Of the few infants in whom hypovolemic shock develops in
the first hours after birth, however, almost all have had intrapartum asphyxia (69). Blood volume
expansion is essential for the infant who is in hypovolemic shock, but may be harmful for the
asphyxiated infant who has a normal blood volume.
Because some circulatory changes caused by asphyxia may either mimic or mask hypovolemic
shock, it is impossible to identify those infants who need blood volume expansion until
resuscitation has produced adequate oxygenation of arterial blood and a normal PaCO2. Acute
hypocarbia causes systemic hypotension, and severe overventilation may reduce systemic blood
flow (see above). Neither of these states requires blood volume expansion.
Signs that suggest an inadequate blood volume include low aortic pressure and a narrow and
abnormal aortic waveform (Fig. 18-9 and Fig. 18-10), falling hematocrit, persistent metabolic
acidosis, low central venous PO2 (i.e., <30 mm Hg) after correction of arterial hypoxia, cold
extremities, and delayed (i.e., >3 seconds) filling of capillaries in the skin after they have been
blanched under pressure, provided that core temperature is normal. Tachycardia often is absent in
the early stages of severe shock and may be present from too many other causes to be a reliably
useful sign (70,71).
Figure 18-9 Mean aortic blood pressure was obtained from an umbilical artery catheter. The
dashed line is the average blood pressure at each birth weight, and the solid lines are the 95%
confidence limits of this relationship.Blood pressure values below the lower confidence line are
hypotensive. (From Versmold HT, Kitterman JA, Phibbs RH, et al. Aortic blood pressure during the
first twelve hours of life in infants with birth weights 610-4220 grams. Pediatrics 1981;67:607).
If some findings suggest shock but the diagnosis is uncertain it is useful to directly monitor the
central venous pressure through the umbilical venous catheter with the tip positioned in the
inferior vena cava or right atrium (32). Central venous pressure may be low or normal during
hypovolemic shock, but it will be high with circulatory tamponade from excessive PPV, tension
pneumothorax, or postasphyxial myocardiopathy. The findings of low or normal central venous
pressure in combination with signs of
P.319
poor systemic perfusion support a trial of volume expansion. Low central venous oxygen content is
a very sensitive, but nonspecific, indicator of increased oxygen extraction in the microcirculation in
response to inadequate oxygen delivery from any cause. It is one of the earliest changes during
hypovolemic shock (72).
Figure 18-10 Changes in aortic blood pressure during rapid hemorrhage in a newborn lamb.
Blood pressure falls and pulse pressure (i.e., systolic minus diastolic) narrows. Note the difference
in waveform before and after hemorrhage. Before the hemorrhage, pressure continues to
decrease after the dicrotic notch, Indicating continued systemic flow during diastole. This
disappears after hemorrhage, indicating little or no systemic flow during diastole. Heart rate has
not yet increased but will do so later.
Hypovolemic shock is best treated with repeated small infusions of whole blood that has been
crossmatched against the mother before delivery and is available in the resuscitation area at birth.
Group O Rh-negative blood given to newborns without crossmatching against the mother's serum
occasionally has produced fatal transfusion reactions caused by incompatibility in minor blood
groups and should not be used. If only packed erythrocytes are available, give equal volumes of
cells and a plasma substitute, such as isotonic saline. If no erythrocytes are available, use normal
saline for initial resuscitation, then give packed cells as soon as they are available. However, this is
less effective than giving blood initially. The objective of therapy is prompt restoration of adequate
tissue perfusion. This must be done rapidly enough to avoid the cumulatively harmful effects of
prolonged underperfusion of tissues. The latter can lead to the secondary effects of shock,
including increased capillary permeability and pulmonary disease, which make therapy more
difficult. Excessive speed in volume replacement is also dangerous, however. Some vascular beds,
such as that of the brain, vasodilate in response to systemic hypotension. If treatment produces
an abrupt rise in systemic pressure, there is not enough time for the vasculature to partially
constrict, and as a result the higher pressure is transmitted to the capillaries, where it may cause
capillary injury, edema, or hemorrhage. In some instances, with the initial correction of
hypovolemia, systemic vasoconstriction is relieved, and pressure falls again, requiring further
volume expansion provided other signs of poor perfusion persist. Occasionally, when there has
been massive hemorrhage, volume may have to be replaced more rapidly. In such a case, monitor
aortic pressure continuously to avoid abrupt rises in pressures. Figure 18-11 shows the course of
successful treatment during the first hour of life in an infant who lost approximately 50% of his
blood volume during delivery and also suffered asphyxia. Figure 18-12 shows the course in an
infant in whom hypovolemia did not become evident until assisted ventilation relieved asphyxia
and unmasked hypovolemia.
Figure 18-11 Heart rate, aortic blood pressure, and blood volume replacement were monitored
during the first hour after birth in a 2.8-kg infant who suffered massive blood loss when the
anteriorly placed placenta was incised deeply at cesarean section delivery. The shaded area shows
the cumulative volume of whole blood given expressed as mL/kg body weight. The final volume,
which produced a normal aortic blood pressure and relieved signs of poor perfusion, was 40 mL/
kg: approximately one-half the total blood volume for a normal newborn infant. The blood was
given as a series of small transfusions guided by the changes in blood pressure. Note that the
heart rate is not elevated at first, despite the extreme hypotension, and that subsequently heart
rate does not consistently change in the opposite direction of blood pressure changes.
Transient myocardial failure of 1 to 2 days of duration can occur after asphyxia (29,73). The
resulting circulatory failure is differentiated from that caused by hypovolemic shock by an elevated
central venous pressure. Postasphyxial cardiomyopathy responds to a continuous infusion of
dopamine with improved systemic perfusion and decreased central venous pressure. Start with a
dose of 5 g/kg per minute and increase the dose as needed to obtain adequate systemic
perfusion pressures. If possible, correct hypoxia and acidosis to improve myocardial function
before starting dopamine. If pulmonary vasoconstriction and hypertension coexist with systemic
hypotension caused by myocardial failure, venous-to-arterial shunting of blood through the
foramen ovale and ductus arteriosus often occurs. This leads
P.320
to additional systemic hypoxemia and worsening metabolic acidosis. When correcting acidosis in
this situation, it is important to avoid hyperventilation, because hypocarbia constricts the coronary
circulation and causes systemic hypotension (60,61). Some infants of poorly controlled diabetic
mothers have a particularly severe form of cardiomyopathy. The antecedents of this
cardiomyopathy are asphyxia plus hypoglycemia plus hypocalcemia, and all of these must be
corrected to improve myocardial performance.
Figure 18-12 Heart rate, aortic blood pressure (Pao), and therapy were monitored during the
first hour after the birth of a 1.5-kg second twin delivered by cesarean section. There had been a
large abruption of the placenta. Initially, the infant was hypoxic and acidotic, and aortic pressure
was normal. As blood gas tensions normalized, aortic pressure fell and the infant continued to
appear pale and poorly perfused. This probably is an example of the intense vasoconstriction of
asphyxia keeping blood pressure at a normal level despite a subnormal blood volume. Relief of
the asphyxia allowed sufficient vasodilation to unmask the hypovolemia.
Additional Measures
These may include treatment of hypoglycemia, suspected infection, or seizures. Complete
documentation of all observations, resuscitative actions, and timings of these actions is absolutely
essential not only for good clinical care but also for medicolegal purposes. Postresuscitation may
also be the first time to meet the family in situations where there has not been the opportunity to
meet the family before delivery. The family should be informed about the infant's condition at the
earliest opportunity, the resuscitative measures that have been taken, why they were necessary,
and, in fact, parents should be encouraged to interact with the newly born as soon as possible.
SPECIAL PROBLEMS
Meconium-Stained Amniotic Fluid
Meconium staining of amniotic fluid occurs in 10% to 15% of all deliveries (74,75). Mature fetuses
pass meconium in response to various stimuli, including asphyxia. Meconium staining diminishes
with decreasing gestational age and is rare before 34 weeks of gestation, whereas it is quite
common in postmature fetuses. The proportion of infants admitted to the neonatal intensive care
unit (NICU) is several-fold higher among infants born through meconium-stained amniotic fluid
(MSAF) than in those born through clear amniotic fluid (76,77). Furthermore, meconium-stained
infants are 100-fold more likely to develop substantial respiratory distress than those born through
clear amniotic fluid (78). Meconium can be aspirated into the airway by gasping, which may occur
in utero in response to a variety of stimuli, including hypoxia, or, by inhalation after delivery.
Aspiration of meconium can cause pulmonary disease both by plugging of the airways or by
producing a chemical pneumonitis. Clinical pulmonary disease is more likely if meconium staining
occurs before the second stage of labor, if meconium-stained fluid is thick with particulate matter,
if infant is depressed at delivery, and if there is meconium below the vocal cords (76). However, a
significant proportion of infants with meconium aspiration syndrome (MAS) are born through thin
consistency meconium staining of amniotic fluid or are vigorous at birth (79,80).
In many instances, infants with meconium aspiration demonstrate symptoms immediately after
birth. However, sometimes, infants are clinically well at birth and manifest symptoms of MAS
during the first few hours after birth. Severe disease nonetheless can develop in infants with this
more gradual onset of symptoms. Pulmonary air leaks are ten times more likely to develop in
infants with meconium aspiration than in infants without meconium staining; this air leak often
occurs during resuscitation. Infants with meconium aspiration are at a great risk of developing
hypoxemia, acidosis, hypercapnia, and the resultant persistent pulmonary hypertension of the
newborn (PPHN). Approximately one-third of infants with MAS have PPHN, and two-thirds of
infants with PPHN are associated with MAS (78,81).
When MSAF is identified before delivery, personnel trained in all aspects of neonatal resuscitation
must be present in the delivery room. On delivery of the head, the infant's nasopharynx and nares
be should be suctioned with a catheter or a bulb syringe before delivering the shoulders. Previous
recommendations included that after suctioning at the perineum, suctioning via ET intubation
should immediately follow if the amniotic fluid is stained with thick meconium versus thin
meconium, irrespective of infant's general condition. However, no study documents the efficacy of
this approach. In fact, a multicenter, prospective, randomized, controlled trial demonstrated that,
regardless of the consistency of meconium, there was no increase in the incidence of respiratory
distress or MAS when vigorous infants were not intubated and suctioned following delivery (80).
Consequently, present recommendations include ET intubation and suctioning soon after delivery
only if the infant is nonvigorous, that is, if the infant has depressed respirations, depressed muscle
tone, and/or heart rate less than 100 bpm. To avoid initiation of breathing, vigorous stimulation
and drying should be delayed until intubation and suctioning are performed in these infants.
Endotracheal intubation and suctioning should be repeated until little additional meconium can be
recovered or until the infant has significant bradycardia and requires positive pressure ventilation
or chest compressions. Therefore, during endotracheal suctioning, a second person should monitor
heart rate continuously. Furthermore, to avoid hypoxemia, free-flow oxygen should be provided
throughout the suctioning procedure. Other procedures to prevent meconium aspiration such as
squeezing baby's chest, chest physiotherapy, or inserting a finger in baby's mouth, which are
sometimes performed, have not been rigorously tested and may actually be harmful. These
procedures are not recommended, and should be strongly discouraged.
Preterm Infants
Preterm (<37 weeks' gestation) infants constitute a special group because of the special needs
and problems encountered in the resuscitation of these infants. As a group premature infants
make up the largest proportion of infants requiring resuscitation after delivery. The majority of
infants born at less than 28 weeks' gestation require some degree of resuscitation. The immature
respiratory system in these infants makes it extremely hard for them to spontaneously achieve
adequate lung inflation and create an adequate FRC. As lung inflation stimulates surfactant release
from the type II alveolar cells, the majority of extremely premature
P.321
infants do require some ventilatory assistance beginning in the delivery room. Further, there is a
large body of evidence suggesting that the pathogenesis of bronchopulmonary dysplasia begins
during the first few minutes of life, with aggressive ventilation that results in volutrauma and
hypocapnia (82). Therefore, initial ventilation requires special attention in the extremely preterm
infant with a small lung volume and surfactant deficiency. Ventilating such lungs with very large
tidal volumes even for a very brief period can do serious lung damage. In premature lambs, as few
as five sustained inflations of 35 to 40 mL/kg leads to the release of proinflammatory cytokines
and gross histopathologic changes (83). The large inflating breaths recommended earlier in this
chapter for resuscitation of larger infants may not be appropriate for extremely premature infants.
Therefore, it has been suggested that a gentler and noninvasive mode of supporting ventilation,
that is, starting continuous positive airway pressure (CPAP) in the delivery room may be an
effective strategy to prevent the development of bronchopulmonary dysplasia (84). However,
there are no convincing data as yet to support this notion. Whether low-tidal-volume ventilation
starting in the delivery room, in an attempt to avoid lung injury and hypocapnia in premature
infants, could be safely delivered and whether it is as efficacious as normal-tidal-volume
ventilation remains to be seen. As pulmonary function improves and these smallest infants are
weaned from assisted ventilation, it is important to remember that many of them cannot maintain
an adequate FRC, even in the absence of lung disease, and progressive atelectasis will gradually
develop unless end-expiratory distending pressure is applied to their lungs.
As with the prevention of volutrauma, special emphasis should also be given to prevent oxygeninduced lung injury by providing the lowest level of oxygen supplementation that maintains
adequate delivery of oxygen to tissues. This requires not only keeping vigilance on the oxygen
concentration provided but also optimizing cardiac output and hemoglobin concentration.
Continuous monitoring of oxygen saturation starting in the delivery room is likely to reduce overall
oxygen exposure (85). Further, recent data suggest that resuscitation with 100% oxygen may
generate oxygen free radicals that may cause tissue damage, particularly to the brain. Some
clinical studies suggest that resuscitation with room air may be both effective and possibly even
safer than with 100% oxygen, but no large, randomized, controlled trial has tested the efficacy of
using room air versus 100% oxygen for resuscitation (86,87,88,89). Such a trial is eagerly
awaited, but until then, the recommendation is to continue to use 100% oxygen for resuscitation.
However, if supplemental oxygen is unavailable, and PPV is indicated, use of room air may be
acceptable.
Although clinical trials have proven that surfactant therapy is effective in improving the clinical
outcome of premature newborns, it is unclear whether preventive (prophylactic or delivery room
administration) or therapeutic (selective or rescue administration) strategy is better. A recent
meta-analysis suggests that prophylactic surfactant administration to premature infants judged to
be at risk for developing respiratory distress syndrome, compared to selective use of surfactant in
infants with established respiratory distress syndrome (RDS), improves clinical outcome (90).
However, it remains unclear as to exactly which criteria should be used to judge at risk infants
who require prophylactic surfactant administration. A reasonable approach is to administer
surfactant to the infant delivered at less than 28 weeks' gestation born to a mother who did not
receive antenatal steroids. The disadvantages of this approach include that the infant has to be
intubated and a significant number of infants may not develop RDS, and would receive surfactant
unnecessarily. As more and more infants are being treated with CPAP in the delivery room, it has
been speculated that transient intubation of the infant solely for the purpose of surfactant
administration and continuation of CPAP may a more cost-effective way of managing RDS in
preterm infants (91).
Because of the relatively larger surface area-to-body mass ratio, thin permeable skin, decreased
amount of subcutaneous fat, and diminished metabolic response to cold stress, premature infants
are at great risk of developing hypothermia. Hypothermia, especially in extremely preterm infants,
may result in adverse outcomes; therefore, during the delivery of premature infants, especially
those of 28 weeks or less of gestation, preventing heat loss should be a top priority. One should
take all the necessary steps to reduce heat loss, even if these infants do not initially appear to
require resuscitation. The radiant heat required to maintain normal body temperature during
resuscitation evaporates water from the infant's skin. This can produce very high insensible water
losses from very premature infants. Once the ET and the umbilical catheter are in place and
secured and other emergency procedures are completed, cover the infant with a clear plastic wrap
to reduce insensible water loss. One study suggests that wrapping premature infants in
polyethylene immediately after delivery prevents heat loss (92). This practice needs to be further
evaluated to determine its potential value in changing the outcome of extremely preterm infants.
Furthermore, preterm infants have a very fragile network of capillaries in the germinal matrix of
their brains and, therefore, are at high risk of developing intraventricular bleeds and lifelong
neurodevelopmental problems. Both periventricular leukomalacia and intraventricular hemorrhage
are linked to birth asphyxia, need for resuscitation, hemodynamic instability, early significant
hypocarbia, and metabolic acidosis (93). To avoid the rupture blood vessels in the germinal
matrix, preterm infants should be handled very gently, and too rapid administration of blood and
volume expanders should be avoided.
Newborn with Airway Compromise

The management of a newborn with a compromised airway can be challenging and requires careful
identification of the source of the airway obstruction. In mild situations, appropriate positioning of
the infant's head and neck or provision of short-term positive pressure ventilation is all that may
be needed. However, in infants with severe airway
P.322
obstruction or with multiple problems, laryngeal mask airway, intubation (either oral or nasal)
requiring fiberoptic endoscopy, or emergent tracheostomy may be required. If the problem is
known prenatally, it may require a team approach involving the obstetrician, neonatologist,
anesthesiologist, and otolaryngologist. In some situations, an ex utero intrapartum tracheloplasty
(EXIT) following a planned cesarean section in which the maternal-fetal placental circulation can be
prolonged for several minutes following delivery can be lifesaving. Laryngeal mask airway may be
a useful temporizing device until a more definitive intervention for the compromised airway can be
arranged. It is particularly helpful when tracheal visualization is impaired because of glossoptosis,
macroglossia, hypoplastic mandible, or cervical immobility. However, its use requires considerable
practice (94).
Newborn with a Surgical Condition

The general principles of stabilization for any infant also apply to an infant requiring surgical
intervention during the initial care. These include establishment of airway patency and adequate
ventilation, effective circulation, correct fluid and electrolyte balance, paying careful attention to
thermoregulation, and providing appropriate antibiotic coverage. However, specific early
management will depend on the underlying surgical condition. In infants with gastroschisis or
omphalocele, avoidance of both direct and ischemic trauma to the eviscerated viscera by
appropriate infant positioning, keeping up with the infant's excessive fluid and heat losses from the
exposed viscera, and preventing infection of the exposed abdominal contents are the biggest
challenges. In an infant with a suspected diaphragmatic hernia, the greatest emphasis during the
initial management should be on avoiding bag and mask ventilation and immediate endotracheal
intubation. Insert an orogastric tube into the stomach and aspirate its contents immediately. This
reduces the risk of regurgitation and aspiration, which can occur despite the presence of an ET.
This also prevents the compression on the already hypoplastic lungs in these patients. If the
diagnosis is not known prenatally, a scaphoid abdomen and difficulty in achieving adequate
ventilation suggest a diaphragmatic hernia. Furthermore, because invariably there is
accompanying pulmonary hypertension, strategies such as sedation, paralysis, correction of
acidosis, avoidance of systemic hypotension, and selective pulmonary vasodilation might be
required from the outset.
Hydrops
Resuscitation and delivery room management of a newborn with hydrops fetalis pose a unique set
of problems for the neonatologist. Usually, the diagnosis is known before delivery. Every effort
should be made to establish the cause of hydrops before delivery to help with preparations for
resuscitation at delivery. With improvements in prenatal care, hydrops because of Rh
alloimmunization has considerably decreased and now hydrops because of nonimmune causes is
much more frequent. More personnel are needed for resuscitation of infants with hydrops than for
a routine resuscitation. Have whole blood or packed erythrocytes crossmatched against the mother
in the resuscitation area, even if the hydrops is not caused by Rh disease or other alloimmune
hemolysis, because many infants with nonimmune hydrops are also anemic at birth. In addition to
the usual supplies for resuscitation, supplies needed for a partial exchange transfusion,
thoracentesis, and paracentesis, plus tubes required for diagnostic studies must be on hand. Two
umbilical vessel catheters should be connected to pressure transducers and a recorder with one
channel calibrated for arterial pressures and the other for venous pressure. Blood should be
obtained from the umbilical cord at birth to measure hematocrit or hemoglobin immediately.
An ultrasonography examination should be done just before delivery to assess for the presence
and size of pleural effusions and ascites. On many occasions, drainage of large pleural effusions
and ascites may be needed before delivery to facilitate delivery as well as to facilitate postnatal
gas exchange (95). Similarly, postnatal drainage of pleural effusions and ascites is often required
in the delivery room (96). Lung inflation and ventilation often are difficult in hydropic infants
because the lungs are compressed by the diaphragm, which is elevated by ascites and by large
pleural effusions. There is also low compliance because of excessive lung water. Resuscitation
usually requires immediate endotracheal intubation and ventilation with oxygen at high pressures.
If paracentesis needs to be performed because of abdominal distension due to ascites, it should be
performed in the flank region to avoid puncturing a potentially large liver or spleen. Remove just
enough fluid so that the abdomen becomes soft and the diaphragm moves easily upon inflation. If
ventilation remains difficult and there are pleural effusions, these should be reduced to allow
ventilation. Even after fluid is removed from the abdomen and chest, many of these infants
continue to require high pressures to provide adequate ventilation because of excess lung water,
surfactant deficiency, and in some cases of long-standing hydrops, pulmonary hypoplasia.
While the infant is being ventilated and the effusions are being reduced, catheterize the umbilical
vein and artery and measure venous and arterial pressures, blood gas tensions and pH to assess
the state of circulation. Anemia compromises tissue oxygenation, and anemic hydropic infants
usually do not respond well to resuscitative measures until the hematocrit is at least 30% to 35%
(97). Transfused blood is virtually 100% hemoglobin A, which transports oxygen much more
efficiently after birth than hemoglobin F. How the anemia is corrected depends on the state of the
circulation. Most infants with hydrops caused by alloimmune disease have low or normal blood
volumes (98). The blood volume of infants with nonimmune hydrops of various causes is
unpredictable. If intravascular pressures indicate that blood volume is adequate, do a partial
exchange transfusion and keep the blood volume constant. If there is evidence
P.323
of hypovolemia, infuse more blood than is withdrawn until the intravascular pressures are normal.
Alternately, infuse a bolus of packed erythrocytes, as is done for the treatment of hypovolemic
shock. In most cases, evidence of hypovolemia does not appear until asphyxia is relieved (97).
The most common scenario is a partial exchange transfusion that keeps blood volume constant
while raising the hematocrit, and then repeated small infusions of packed erythrocytes or freshfrozen plasma to support the circulation. Fresh-frozen plasma may partially correct the coagulation
defects that are often present in these infants (99). Most infants with alloimmune hydrops have
very low concentrations of serum albumin and a low plasma oncotic pressure (97,100). About onehalf of those with nonimmune hydrops are also hypoalbuminemic. There is no evidence, however,
that albumin infusion or correction of hypoproteinemia is beneficial (96). Moreover, there is
concern that giving albumin and fresh frozen plasma to these infants during resuscitation may
raise plasma oncotic pressure enough to draw excessive volumes of fluid into the circulation and
worsen pulmonary edema. If this occurs, appropriate adjustments to ventilatory support are
needed. Pulmonary vasoconstriction is particularly common in infants with hydrops (96,101).
Anticipate and promptly correct metabolic derangements such as acidosis and hypoglycemia.
Surfactant deficiency and hypoplastic lungs may be associated with hydrops, and are managed
accordingly. Therefore, if pulmonary perfusion does not improve during resuscitation, correction of
metabolic acidosis may be required. After resuscitation is complete, there usually is pulmonary
disease that requires assisted ventilation. This may be transient respiratory distress, hyaline
membrane disease, pulmonary hypoplasia, or some combination of these.
Multiple Births
The three features of multiple births that complicate delivery room management are the following:
Increased incidence of preterm labor and delivery. This affects management only by
increasing the number of personnel needed for resuscitation. The risk of intrapartum
asphyxia is somewhat increased in the second-born twin.
Increased incidence of congenital anomalies in monozygotic multiple births.
Increased risk of intrauterine growth retardation because the placenta may be unevenly
shared among the fetuses. This results in the usual problems associated with intrauterine
growth retardation, for example, intrapartum asphyxia, polycythemia, hypoglycemia, and
pulmonary hemorrhage (see Chapter 11). Twin-to-twin syndrome with the most severe form
being stuck twin syndrome.
The twin-to-twin transfusion syndrome, which occurs in at least 5% of multiple pregnancies,

results from vascular anastomosis between the circulations of monozygotic twins, primarily with a
monochorionic placenta. Both the degree of transfusion from one fetus to the other and the time
course are quite variable and determine the clinical presentation at birth. Transfusion may have
been relatively recent or may have begun in the second trimester and been long-standing at the
time of birth. If the transfusion is primarily in one direction, the donor twin becomes anemic and
the recipient twin polycythemic. With ongoing transfusion, the recipient grows normally while the
donor becomes progressively smaller for gestational age. As the process becomes more severe,
polyhydramnios develops in the recipient fetus and oligohydramnios in the donor. Ultimately,
either twin may become hydropic, one from volume overload and the other from anemia. In
severe cases, the donor twin may die (102).
When twins suffering from this syndrome are delivered, management during resuscitation can be
extremely complicated. Findings in the mother should lead to close ultrasonographic surveillance of
the fetuses and prenatal recognition of the unidirectional twin-to-twin syndrome. The polycythemic
infant might need to have his or her hematocrit reduced. Management of the anemic donor is less
straightforward. Recent blood loss requires the same management as in any other hypovolemic
infant (see above). More often, the anemia is prolonged and severe, so the donor circulation may
be compromised and may not tolerate blood volume expansion. In this situation, proper therapy is
partial exchange transfusion with packed cells to raise the hematocrit to a normal level. Both
arterial and central venous pressures should be monitored beginning immediately after birth to
assess the circulatory status and make the correct adjustments in both hematocrit and
intravascular volume. Rapidly measure hematocrit or hemoglobin in each twin and begin
appropriate therapy.
In monochorionic twins, the vascular anastomoses may be multidirectional, so that the direction of
flow is determined at least in part by the difference in circulatory resistance between the twins.
Such twin-to-twin transfusions are not diagnosed so easily, nor are the hemoglobin measurements
necessarily different at birth even when the blood volumes are.
The stuck twin syndrome is a poorly understood phenomenon in monochorionic twins. There is
discordant growth with oligohydramnios in the growth restricted fetus and polyhydramnios in the
appropriate-for-gestational-age fetus. The growth-restricted fetus becomes impacted into a small
volume within the uterus. Lung growth is often restricted, which leads to pulmonary hypoplasia
that is lethal, if severe. If mild, it requires positive-pressure ventilation at high pressures and
rates. In some cases, there is marked myocardial dysfunction in one or both twins, which can be
detected antenatally by echocardiography. Preparations for delivery are the same as for any other
severe twin-to-twin transfusion syndrome.
Birth Injury
Severe birth injury and intrapartum asphyxia often occur together. The main problem for delivery
room management
P.324
in these infants is significant hemorrhage from the traumatized tissues, which complicates the
resuscitation. The blood loss almost always is internal and, therefore, not immediately evident.
Moderate blood loss can occur in fractured limbs or into the perineum in a difficult breech delivery.
Sites for major blood loss include intracranial, mediastinal, and intraabdominal (e.g., ruptured
spleen, hepatic subcapsular hematoma). A subgaleal hematoma can produce a massive loss of
blood volume because of the extremely large potential space. Any of these hematomas can contain
several hundred milliliters of blood and are particularly dangerous because they can consume large
quantities of coagulation factors and lead to generalized bleeding that perpetuates the
hypovolemia. In severe cases, only early and aggressive therapy can bring the situation under
control. Treatment includes replacement of the lost blood volume and erythrocyte mass and, if
there is depletion of clotting factors, treatment with fresh-frozen plasma, platelets, and,
occasionally, cryoprecipitate.
Early detection of internal hemorrhage from birth trauma is crucial. Abdominal distention and
discoloration suggest intraabdominal bleeding, which may need to be verified by abdominal
ultrasonography, computed tomographic scanning, magnetic resonance imaging, or needle
aspiration. Control of intraabdominal bleeding may require surgery. Intracranial hemorrhage
sufficient to cause hypovolemia usually is manifested by a bulging fontanelle and can be confirmed
by ultrasonography. Small intracranial hemorrhages also may cause circulatory instability through
their effects on the autonomic nervous system. A mediastinal hematoma does not declare itself by
the physical signs, but a chest radiograph often suggests its presence when the mediastinum is
widened. If suspected, it can be diagnosed quickly by an ultrasonographic examination of the
mediastinum. Early swelling of the back of the neck from a subgaleal hematoma may be hard to
recognize, but an expanding subgaleal hemorrhage pushes the ears laterally and forward. This
often is the earliest sign of this condition. Subgaleal hemorrhage, too, can be confirmed by
ultrasonography.
Overall, remarkable progress has been made since the first publication of the neonatal
resuscitation guidelines by the American Academy of Pediatrics and American Heart Association in
1985. However, many recommendations for neonatal resuscitation still are based on accepted
practice rather than research data. Many of these are being questioned, and are now being
rigorously tested. Undoubtedly, based on the analysis of new data, novel recommendations will
emerge. Fortunately, with skillful intervention, resuscitation of a newborn infant is usually
successful, in contrast to resuscitation attempts in an older child or an adult.
ACKNOWLEDGMENTS
We are extremely grateful to Nik Phou for his help in the preparation of the manuscript, and to M.
Vasudeva Kamath, MD, MPH, for helpful suggestions.
REFERENCES
1. Kattwinkel J, ed. American Academy of Pediatrics/American Heart Association: Textbook of
Neonatal Resusitation, 4th ed. Elk grove village, IL, American Academy of Pediatrics, American
Heart Association, 2000.
2. Perlman JM, Risser R. Cardiopulmonary resuscitation in the delivery room. Arch Pediatr
Adolesc Med 1995;149:20-25.
3. Brown MJ, Oliver RE, Ramsden CA, et al. Effects of adrenaline and spontaneous labour on the
secretion and absorption of lung fluid in the fetal lamb. J Physiol 1983:344:137-152.
4. Wirtz HR, Dobbs LG. The effects of mechanical forces on lung function. Respir Physiol
2000;119:1-17.
5. Torrance S, Wittnich C. The effect of varying arterial oxygen tension on neonatal acid-base
balance. Pediatr Res 1992;31: 112-116.
6. James LS, Weisbrot IM, Prince CE, et al. The acid-base status of human infants in relation to
birth asphyxia and onset of respiration. J Pediatr 1958;52:379-394.
7. Yeomans ER, Hauth JC, Gilstrap LC, et al. Umbilical cord pH, PCO2, and bicarbonate following
uncomplicated term vaginal deliveries. Am J Obstet Gynecol 1985;151:798-800.
8. Goodwin TM, Belai I, Hernandez P, et al. Asphyxial complications in the term newborn with
severe acidemia. Am J Obstet Gynecol 1992;167:1506-1512.
9. Rudolph CD, Roman C, Rudolph AM. Effect of acute umbilical cord compression on hepatic
carbohydrate metabolism in the fetal lamb. Pediatr Res 1989;25:228-233.
10. Sapirstein A, Bonventre JV. Phospholipases A2 in ischemic and toxic brain injury. Neurochem
Res 2000;25:745-753.
11. Savman K, Blennow M, Gustafson K, et al. Cytokine response in cerebrospinal fluid after
birth asphyxia. Pediatr Res 1998;43: 746-751.
12. Ruth V, Fyhrquist F, Clemons G, et al. Cord plasma vasopressin, erythropoietin, and
hypoxanthine as indices of asphyxia at birth. Pediatr Res 1988;24:490-494.
13. Dawes G. Fetal and neonatal physiology. Chicago: Year Book, 1968.
14. Volpe JJ. Neurology of the newborn, 4th ed. Philadelphia: WB Saunders, 2001.
15. Cohn HE, Sacks EJ, Heymann MA, et al. Cardiovascular responses to hypokalemia and
acidemia in fetal lambs. Am J Obstet Gynecol 1974;120:817-824.
16. Fisher DJ. Increased regional myocardial blood flows and oxygen deliveries during
hypoxemia in lambs. Pediatr Res 1984;18:602- 606.
17. Boyle DW, Hirst K, Zerbe GO, et al. Fetal hind limb oxygen consumption and blood flow
during acute graded hypoxia. Pediatr Res 1990;28:94-100.
18. Rudolph AM, Yuan S. Response of the pulmonary vasculature of hypoxia and H+ ion
concentration changes. J Clin Invest 1966;45:339-411.
19. Parer JT. The effect of acute maternal hypoxia on fetal oxygenation and the umbilical
circulation in the sheep. Eur J Obstet Gynecol Reprod Biol 1980;10:125-136.
20. Fisher DJ. Acidemia reduces cardiac output and left ventri-cular contractility in conscious
lambs. J Dev Physiol 1986;8: 23-31.
21. Lewinsky R, Szware R, Benson L, et al. The effects of hypoxic acidemia on left ventricular
end-diastolic pressure elastance in fetal sheep. Pediatr Res 1993;34:38-43.
22. Itskovitz J, LaGamma EF, Bristow J, et al. Cardiovascular responses to hypoxemia in

sinoaortic-denervated fetal sheep. Pediatr Res 1991;30:381-385.
23. Jones CT, Roebuck MM, Walker DW, et al. The role of the adrenal medulla and peripheral
sympathetic nerves in the physiological responses of the fetal sheep to hypoxia. J Dev Physiol
1988; 10:17-36.
24. Cheung CY, Brace RA. Fetal hypoxia elevates plasma atrial natriuretic factor concentration.
Am J Obstet Gynecol 1988;159:1263- 1268.
25. Fisher DJ. -Adrenergic influence on increased myocardial oxygen consumption during
hypoxemia in awake newborn lambs. Pediatr Res 1989;25:585-590.
26. Slotkin TA, Seidler FJ. Adrenomedullary catecholamine release in the fetus and newborn:
secretory mechanisms and their role in stress and survival. J Dev Physiol 1988;10:1-16.
P.325
27. Perez R, Espinoza M, Riquelme R, et al. Arginine vasopressin mediates cardiovascular
responses to hypoxia in fetal sheep. Am J Physiol 1989;256:R1011-1018.
28. Linderkamp O, Versmold HT, Messow-Zahn K, et al. The effects of intrapartum and
intrauterine asphyxia on placental transfusion in premature and full-term infants. Eur J Pediatr
1978; 127:91-99.
29. Bucciarelli RL, Nelson RM, Egan EA, et al. Transient tricuspid insufficiency of the newborn: a
form of myocardial dysfunction in stressed newborn. Pediatrics 1977;59:330-337.
30. D'Alecy LG, Lundy EF, Barton KJ, et al. Dextrose containing intravenous fluid impairs
outcome and increases death after eight minutes of cardiac arrest and resuscitation in dogs.
Surgery 1986;100:505-511.
31. Venkataraman PS, Tsang RC, Chen IW, et al. Pathogenesis of early neonatal hypocalcemia;
studies of serum gastrin and plasma glucagon. J Pediatr 1987;110:599-603.
32. Kitterman JA, Phibbs RH, Tooley WH. Catheterization of umbilical vessels in newborn infants.
Pediatr Clin North Am 1970; 17:895-912.
33. Apgar V. A proposal for a new method of evaluation of the newborn infant. Curr Res Anesth
Analg 1953;32:260-267.
34. Sykes GS, Molloy PM, Johnson P, et al. Do Apgar scores indicate asphyxia? Lancet
1982;1:494-496.
35. Marrin M, Paes BA. Birth asphyxia: does the Apgar score have diagnostic value? Obstet
Gynecol 1988;72:120-123.
36. Meyer BA, Dickinson JE, Chambers C, et al. The effect of fetal sepsis on umbilical cord blood
gases. Am J Obstet Gynecol 1992;166:612-617.
37. Catlin EA, Carpenter MW, Brann BS, et al. The Apgar score revisited: influence of gestational
age. J Pediatr 1986;109:865-868.
38. Milner A. The importance of ventilation to effective resuscitation in the term and preterm
infant. Semin Neonatol 2001;6:219-224.
39. Vyas H, Milner AD, Hopkin IE, et al. Physiologic responses to prolonged and slow-rise
inflation in the resuscitation of the asphyxiated newborn infant. J Pediatr 1981;99:635-639.
40. Lindemann R. Resuscitation of the newborn with endotracheal administration of epinephrine.

Acta Paediatr Scand 1984;73: 210-212.
41. Ellemunter H, Simma B, Trawoger R, et al. Intraosseous lines in preterm and full term
neonates. Arch Dis Child Fetal Neonatal Ed 1999;80:F74-F75.
42. Downing SE, Talner NS, Gardner TH. Influences of hypoxemia and acidemia on left
ventricular function. Am J Physiol 1966; 210:1327-1334.
43. Effron MB, Guarnieri T, Frederiksen JW, et al. Effect of tris (hydroxymethyl) aminomethane
on ischemic myocardium. Am J Physiol 1978;235:H167-H174.
44. Lewis AB, Heymann MA, Rudolph AM. Gestational changes in pulmonary vascular responses
in fetal lambs in utero. Circ Res 1976;39:536-541.
45. Schreiber MD, Heymann MA, Soifer SJ. Increased arterial pH, not decreased PaCO2,
attenuates hypoxia-induced pulmonary vaso-constriction in newborn lambs. Pediatr Res
1986;20:113-117.
46. Wiklund L, Oquist L, Skoog G, et al. Clinical buffering of metabolic acidosis: problems and a
solution. Resuscitation 1985;12:279-293.
47. Akinbi H, Abbasi S, Hilpert PL, et al. Gastrointestinal and renal blood flow velocity profile in
neonates with birth asphyxia. J Pediatr 1994;125:625-627.
48. Martin-Ancel A, Garcia Alix A, Gaya F, et al. Multiple organ involvement in perinatal
asphyxia. J Pediatr 1995;127:786-793.
49. Perlman JM, Tack ED. Renal injury in the asphyxiated newborn infant: relationship to
neurologic outcome. J Pediatr 1988;113: 875-879.
50. Vannucci RC, Perlman J. Interventions for perinatal hypoxic-ischemic encephalopathy.

Pediatrics 1997;100:1004-1014.
51. Fellman V, Raivio KO. Reperfusion injury as the mechanism of the brain damage after
perinatal asphyxia. Pediatr Res 1997; 41:599-606.
52. Vannucci RC, Brucklacher RM, Vannucci SJ. Effect of carbon dioxide on cerebral metabolism
during hypoxic-ischemia in the immature rat. Pediatr Res 1997;42:24-29.
53. Gunn AJ. Cerebral hypothermia for prevention of brain injury following perinatal asphyxia.
Curr Opin Pediatr 2000;12(2): 111-115.
54. Shankaran S, Laptook A, Wright LL, et al. Whole body hypothermia for neonatal
encephalopathy: animal observations as a basis for a randomized controlled pilot study in term
55. Debillon T, Daoud P, Durand P, et al. Whole-body cooling after perinatal asphyxia: a pilot
study in term neonates. Dev Med Child Neurol 2003;45:17-23.
56. Davis DJ. How aggressive should delivery room cardiopulmonary resuscitation be for
extremely low birth weight neonates? Pediatrics 1993;92:447-450.
57. Jain L, Ferre C, Vidyasagar D, et al. Cardiopulmonary resuscitation of apparently stillborn

infants: survival and long-term outcome. J Pediatr 1991;118:778-782.
58. Casalaz DM, Marlow N, Speidel BD. Outcome of resuscitation following unexpected apparent
stillbirth. Arch Dis Child Fetal Neonatal Ed 1998;78:F112-F115.
59. Patel J, Marks K, Roberts I, et al. Measurement of cerebral blood flow in infants using near
infrared spectroscopy with indocyanine green. Pediatr Res 1998;43:34-39.
60. Kruyswijk H, Jansen BH, Muller EJ. Hyperventilation-induced coronary artery spasm. Am
Heart J 1986;112:613-615.
61. Case RB, Felix A, Wachter M, et al. Relative effect of CO2 on canine coronary vascular
resistance. Circ Res 1978;42:410-418.
62. Ogata ES, Kitterman JA, Gregory GA, et al. Pneumothorax in the respiratory distress
syndrome: incidence and effect on vital signs, blood gases, and pH. Pediatrics 1976;58:177-183.
63. Desmond MM, Kay JL, Megarity AL. The phases of transitional distress occurring in neonates
in association with pro-longed postnatal umbilical cord pulsations. J Pediatr 1959;55:131-151.
64. Thibeault DW, Hall FK, Sheehan MB, et al. Post-asphyxial lung disease in newborn infants
with severe perinatal acidosis. Am J Obstet Gynecol 1984;150:393-393.
65. The OSIRIS Collaborative Group. Early versus delayed neonatal administration of a synthetic
surfactantthe judgment of OSIRIS. Lancet 1992;340:1363-1369.
66. Segerer H, Stevens P, Schadow B, et al. Surfactant substitution in ventilated very low birth
weight infants: factors related to response types. Pediatr Res 1991;30(6):591-596.
67. Robertson B. New targets for surfactant replacement therapy: experimental and clinical
aspects. Arch Dis Child Fetal Neonatal Ed 1996;75:F1-F3.
68. Simbruner G, Rudolph AM. Relationship between peripheral blood flow and blood
temperatures in lambs during hypoxemia and hemorrhage. Biol Neonate 1982;42(1-2):31-38.
69. Paxon CL Jr. Neonatal shock in the first postnatal day. Am J Dis Child 1978;132:509-514.
70. Sola A, Spitzer AR, Morin FC, et al. Effects of arterial carbon dioxide tension on the newborn
lamb's cardiovascular responses to rapid hemorrhage. Pediatr Res 1983;17:70-76.
71. Meyers RL, Paulick RP, Rudolph CD, et al. Cardiovascular responses to acute, severe
hemorrhage in fetal sheep. J Dev Physiol 1991;15:189-197.
72. Weil MH, Rackow EC, Trevino R, et al. Difference in acid-base state between venous and
arterial blood during cardiopulmonary resuscitation. N Engl J Med 1986;315:153-156.
73. Walther FJ, Siassi B, Ramadan NA, et al. Cardiac output in newborn infants with transient
myocardial dysfunction. J Pediatr 1985;107:781-785.
74. Gregory GA, Gooding C, Phibbs RH, et al. Meconium aspiration in infants: a prospective
study. J Pediatr 1974;85:848-852.
75. Wiswell TE, Tuggle JM, Turner BS. Meconium aspiration syndrome: have we made a
difference? Pediatrics 1990;85: 715-721.
76. Anyaegbunam A, Fleischer A, Whitty J, et al. Association between umbilical artery cord pH,
five-minute Apgar scores and neonatal outcome. Gynecol Obstet Invest 1991;32:220-223.
77. Nathan L, Leveno KJ, Carmody TJ, et al. Meconium: a 1990's perspective on an old obstetric
hazard. Obstet Gynecol 1994; 83:329-332.
78. Fleischer A, Anyaegbunam A, Guidetti D, et al. A persistent clinical problem: profile of the
term infant with significant respiratory complications. Obstet Gynecol 1992;79:185-190.
79. Cleary GM, Wiswell TE. Meconium-stained amniotic fluid and the meconium aspiration
syndrome: an update. Pediatr Clin North Am 1998;45:511-529.
80. Wiswell TE, Gannon CM, Jacob J, et al. Delivery room management of the apparently
vigorous meconium-stained neonate: results of the multicenter, international collaborative trial.
Pediatrics 2000;105:1-7.
P.326
81. Abu-Osba YK. Treatment of persistent pulmonary hypertension of the newborn: update. Arch
Dis Child 1991;66:74-77.
82. Auten RL, Vozzelli M, Clark RH. Volutrauma. What is it, and how do we avoid it? Clin
Perinatol 2001;28:505-515.
83. Bjorklund LJ, Ingimarsson J, Curstedt T, et al. Manual ventilation with a few large breaths at
birth compromises the therapeutic effect of subsequent surfactant replacement in immature
lambs. Pediatr Res 1997;42:348-355.
84. Narendran V, Donovan EF, Hoath SB, et al. Early bubble CPAP and outcomes in ELBW
preterm infants. J Perinatol 2003; 23:195-199.
85. Kopotic RJ, Lindner W. Assessing high-risk infants in the delivery room with pulse oximetry.
Anesth Analg 2002;94:S31-S36.
86. Saugstad OD, Rootwelt T, Aalen O. Resuscitation of asphyxiated newborn infants with room
air or oxygen: an international controlled trial: the Resair 2 study. Pediatrics 1998;102:e1.
87. Vento M, Asensi M, Sastre J, et al. Resuscitation with room air instead of 100% oxygen
prevents oxidative stress in moderately asphyxiated term neonates. Pediatrics 2001;107:642647.
88. Vento M, Asensi M, Sastre J, et al. Six years of experience with the use of room air for the
resuscitation of asphyxiated newly born term infants. Biol Neonate 2001;79:261-267.
89. Saugstad OD, Ramji S, Irani SF, et al. Resuscitation of newborn infants with 21% or 100%
oxygen: follow-up at 18 to 24 months. Pediatrics 2003;112:296-300.
90. Soll RF, Morley CJ. Prophylactic versus selective use of surfactant in preventing morbidity
and mortality in preterm infants. Cochrane Database Syst Rev 2001;2:CD000510.
91. D'Angio CT, Khalak R, Stevens TP, et al. Intratracheal surfactant administration by transient
intubation in infants 29-35 weeks gestation with RDS requiring CPAP decreases the likelihood of
later mechanical ventilation: a randomized controlled trial. Pediatr Res 2003;53:A2088.
92. Vohra S, Frent G, Campbell V, et al. Effect of polyethylene skin wrapping on heat loss in
very low birth weight infants at delivery: a randomized trial. J Pediatr 1999;134:547-551.
93. Ozdemir A, Brown MA, Morgan WJ. Markers and mediators of inflammation in neonatal lung
disease. Pediatr Pulmonol 1997;23:292-306.
94. Lopez-Gil M, Brimacombe J, Cebrian J, et al. Laryngeal mask airway in pediatric practice: a
prospective study of skill acquisition by anesthesia residents. Anesthesiology 1996;84:807-811.
95. Holzgreve W, Holzgreve B, Curry CJ. Nonimmune hydrops fetalis: diagnosis and
Management. Semi Perinatol 1985;9:52-67.
96. Carlton DP, McGillivray BC, Schreiber MD. Nonimmune hydrops fetalis: a multidisciplinary
approach. Clin Perinatol 1989;16:839-851.
97. Phibbs RH, Johnson P, Kitterman JA, et al. Cardiorespiratory status of erythroblastotic
newborn infants: III. Intravascular pressures during the first hours of life. Pediatrics
1976;58:484-493.
98. Phibbs RH, Johnson P, Tooley WH. Cardiorespiratory status of erythroblastotic newborn
infants: II. Blood volume hematocrit and serum albumin concentrations in relation to hydrops
fetalis. Pediatrics 1974;53:13-23.
99. Hey E, Jones P. Coagulation failure in babies with rhesus isoimmunization. Br J Haematol
1979;42:441-454.
100. Baum JD, Harris D. Colloid osmotic pressure in erythroblastosis fetalis. Br Med J
1972;1:601-603.
101. Phibbs RH, Johnson P, Kitterman JA, et al. Cardiorespiratory status of erythroblastotic
infants: I. Relationship of gestational age, severity of hemolytic disease and birth asphyxia to
idiopathic respiratory distress syndrome and survival. Pediatrics 1972; 49:5-14.
102. Rehan VK, Menticoglou SM. Mechanism of visceral damage in fetofetal transfusion
syndrome. Arch Dis Child Fetal Neonatal Ed 1995;73:F48-F50.
Chapter 19
Physical Assessment and Classification
Michael Narvey
Mary Ann Fletcher
The approach to the newborn examination differs in several ways from that of the adult patient. It is imperative
that one seizes opportunities as they arise, rather than force an unwilling infant to an assessment based on the
examiner's preferred order. Such forcing often culminates in the infant crying, which although informative may
render a complete examination impossible at that time. Instead, during a quiet moment, one may appreciate the
heart sounds or the clarity of breath sounds. Similarly during an awake, active period, simple observation may
yield a plethora of information regarding the neurologic status of the infant. In the acutely ill infant, delaying the
complete examination until such time that the infant may be handled safely is prudent. Inspection, palpation,
percussion, and auscultation are all important tools for examination at any age and are incorporated into the
neonatal assessment. It is important to provide a thorough systematic assessment of the newborn, but flexibility
must be inherent in the approach.
Antenatal ultrasound screening and more recently the evaluation of fetal abnormalities by MRI have provided
physicians with the capability of preparing for the delivery of newborns with anticipated problems. For example,
antenatal discovery of congenital diaphragmatic hernia allows planned delivery in a tertiary care hospital with
neonatal staff present at delivery. Despite improvements in triaging of deliveries, the physical examination at
birth remains a critical tool in the management of all newborns. In the above example, assessment of vital
signs, work of breathing, color, and the presence of other anomalies at birth determine subsequent
management, not just the diagnosis itself. In other cases, significant anomalies may be present but their impact
on the newborn not appreciated until examination after birth.
Physical assessment in neonates serves to determine anatomic normality for the first time in a new life and the
state of health in someone unable to describe their symptoms. A challenge is to determine which findings will be
transient or are merely variations of normal and which are markers of major malformations or syndromes. Most
of the clinical descriptions of specific syndromes are made of findings that become typical only after there has
been sufficient growth and maturation. An example is the subtlety or absence of obvious findings in aborted
fetuses or extremely premature infants with Down syndrome.
Because many of the physical signs of early disease also present as part of the normal physiologic changes
occurring at birth or in the newborn period, differentiating the markers of subtle illness from transitional
variations is a particular challenge in neonatal physical diagnosis. Additionally, there are unique findings that
appear quite dramatic but carry little medical significance. Once the examiner has determined that findings
represent a disease process, he or she then has to decide just how sick that infant is, or is likely to become. To
that end, there have been devised a number of acuity of illness scores that range from the very simple to
complex systems that include physiologic monitoring and laboratory values (1,2,3,4,5). The primary advantage
of such scoring systems is in forcing a systematic and quantitative assessment that can be compared among
observers and over time. Such scores have been used as indicators of mortality risk (5).
The first neonatal examination occurs immediately after birth in the assigning of Apgar scores at 1 and 5
minutes of age and every 5 minutes thereafter until the total is above 7. The scores summarize encapsulated
assessments of the cardiopulmonary and neurologic systems after inspection for color, heart rate, respiratory
efforts, tone, and muscle activity and assigning a value of 0, 1, or 2 for each of the five observations. Also part
of this first examination is inspection for designation of gender and a cursory inspection for major anomalies.
Any obvious abnormality merits more immediate evaluation, but the definitive examination in healthy infants
should take place after initial transition and the first bath.
P.328
The first complete examination ordinarily occurs within the first 24 hours after birth, but if any portion of an
assessment is deferred or abnormal at that time reexamination prior to discharge is warranted. For infants
discharged before 48 hours after delivery, an examination should take place within 48 hours of discharge by a
health professional competent in newborn assessment (6).
During the first outpatient visit, the physician should fully reevaluate the infant, especially those systems not as
easily assessed immediately after birth, e.g., the eyes, and those that undergo the greatest changes during
transition, e.g., the cardiovascular and hepatobiliary systems. The areas emphasized with subsequent well-baby
examinations include neuromuscular and sensory development, the heart, and the hips, as well as parameters of
growth including head circumference, length, and weight.
This chapter is a brief discussion of the steps for assessing the newborn infant and interpreting some of the
findings. A thorough textbook on how to perform and interpret the physical examination in neonates is available
(7).
NEWBORN HISTORY
It is tempting to start the physical examination of neonates before reviewing the history and available laboratory
information of the mother. If a newborn is critically ill, the clinician should initiate therapy for stabilization after a
cursory examination and before obtaining the complete history, but too much delay can lead to missed or partial
diagnoses. Historic information is just as important for neonates as for any other patient. Even if it is more
practical to examine an apparently healthy newborn before obtaining the history, a complete evaluation includes
all available information. Knowledge of certain historic details may increase one's vigilance for signs of drug
withdrawal in a baby born to a drug abuser, for instance. A key part is the mother's pregnancy history as well as
her prior medical and social history. Other essential elements include general family history, postnatal history,
and information about the placental examination.
Maternal history includes the mother's age, gravidity, parity, time and type of previous fetal losses, general
fertility issues, and premature births and their outcomes; maternal illness before or during pregnancy; extent
and location of prenatal care; results of any prenatal laboratory tests, especially those for hepatitis, human
immunodeficiency virus (HIV), and sexually transmitted diseases; labor and delivery history including duration,
assessments of fetal well-being, anesthesia, and route of delivery, drug, alcohol, and tobacco use; prescription
and nonprescription medication use; and her vocation. The general family history includes current or significant
past medical illnesses in other family members, including siblings; physical traits or appearance, including birth
weights of other siblings; consanguinity; social information, educational levels, and vocations; and ethnic or
racial background. Helpful information about the newborn period of siblings includes success in breast-feeding,
infections, congenital anomalies, genetic conditions, jaundice, and other concerns. To include the postnatal
course as part of the newborn history, the clinician should review events surrounding the birth and response to
resuscitation, vital signs, feeding, eliminations, and behavior. If there were any complications or requirement for
anything other than routine care, this information is a key part of the total neonatal history.
Placental Information
Often more overlooked than history in evaluating neonates is information about the placenta and the clues it
provides about the gestational history. Several features of the placenta and cord can be readily assessed on
gross examination by anyone at the time of delivery. The placenta should be examined for size, odor, color, and
the number and character of fetal membranes. In the last trimester the ratio of fresh placental weight to infant
weight is normally 1:6. There should be a uniform thickness and density throughout. Depressions and adherent
clots or changes in firmness on the maternal surface suggest abruption or infarction. The placenta is essentially
odorless except for a slight odor of fresh blood. Malodor may indicate the presence of infection although this is
controversial.
The color of the fetal surface changes with gestational age (GA), but pallor or plethora suggest aberrations in
fetal blood volume or hemoglobin level. Elevated bilirubin in the amniotic fluid stains the placenta bright yellow.
Meconium will discolor the fetal surface greenish-brown but so too can old blood. If either meconium passage or
bleeding occurred more than 1 day prior to delivery, it can be difficult to differentiate the two by gross
examination.
The fetal surface should be examined for cloudiness of fetal membranes, which suggests an inflammatory
reaction but not necessarily due to infection. Nodules on the amnion indicate prolonged, extreme
oligohydramnios. Fetal pulmonary hypoplasia is highly probable in this setting and is a key finding in renal
agenesis. Their presence suggests futility if resuscitation is underway.
In multiple gestations with a single placenta, the dividing membranes should be assessed. Membranes can be
teased apart and counted: four layers indicate dichorionic placentation and two layers indicate a monochorionic
placenta. With a dichorionic placenta or completely separate placentas and same gender twins, one cannot say if
they are identical or fraternal from the placental examination. Monochorionicity has traditionally been viewed as
patho-gnomonic for identical twins; however, a recent report refutes this assertion (8). Just as monozygotic
twins can have separate placentas, dizygotic twins can have tightly fused placentas that appear to be one mass.
If there are any remnants of vessels seen in translucent dividing membranes held to a light, there are four
membrane layers. If there is only a transparent membrane with no chorionic remnants, it is likely to contain
only amnion.
P.329
GESTATIONAL AGE ASSESSMENT

For standard reporting of reproductive health statistics and as a prerequisite to determining normality, all infants
should be classified by GA and birth weight (6). Ultrasonography has improved the accuracy of pregnancy
dating, but discrepancies in dates, physical appearance, or size require further evaluation. If there has been no
prenatal care, physi-cal assessment remains the primary clinical determinant of GA.
GA is noted in completed weeks after the onset of the last menstrual period (LMP). For example, a fetus who is
37 weeks and 2 days is in its 38th week of gestation. A term infant is any infant whose birth occurs from the
beginning of the first day of week 38 through the end of the last day of week 42 after the onset of the LMP (i.e.,
260 to 294 days of gestation). A preterm infant is one born before 37 completed weeks or at 36 weeks and 6
days (259 days) or less. A postterm infant is one whose birth occurs from the beginning of the first day of week
43 (i.e., after 294 days or at 42 weeks and 1 day or more). Correct classification of GA has assumed great
importance when discussing the outcome of preterm infants. GAs should never be rounded up. For example, a
preterm infant of 24 weeks and 6 days should be considered and classified as 24, not 25 weeks. Classifying
infants born at term, preterm, or postterm helps to establish the level of risk for neonatal morbidity and longterm developmental problems. The terms premature and postmature often are used clinically to connote
physical or physiologic findings not always correlating with GA. Postdates is an obstetric term meaning a
pregnancy that has continued to any time after the expected date of confinement but is not necessarily postterm.
Assessment Techniques
Estimation of GA by physical examination is possible because there is a predictable pattern of physical changes
that occur throughout gestation. The most popular score for GA assessment was originally developed in part by
Saint-Anne-Dargassies (9), Amiel-Tison (10), and Dubowitz and associates (11). These systems have been
applied to infants with GAs of 22 to 27 weeks despite the fact that no infants below 28 weeks were included in
their development. In these very premature infants, the GA was consistently overestimated by up to 2 weeks
when compared to reliable ultrasound dates. Conversely, in postterm pregnancies they tended to underestimate
the age of postterm infants (12). A further modification by Ballard and associates (13), which included infants
with GAs 26 weeks, purportedly improved reliability. However, even in infant cohorts of <1,500 g and <2,500
g, the Ballard score similarly yielded a 1- to 2-week error, being greatest at lower birth weights (14,15,16,17). A
final modification produced the New Ballard Score (NBS), which claimed to improve the accuracy of age
assessment to within 1 week (Fig. 19-1) (18). The strength of this study was inclusion of infants 20 weeks,
designed to improve accuracy because of larger numbers of extremely premature infants. However a recent
study in 24- to 27-week premature infants refuted these findings, showing persistent miscalculation by up to 2
weeks (19). Some have attributed the tendency to overestimate true GA to accelerated neurologic maturity
(14,18). It is likely that factors contributing to preterm birth cause stress in the developing fetus and bring
about faster neurologic maturation compared to unstressed fetuses in continuing pregnancies. Using last
menstrual period (LMP) or early ultrasound as the gold standard, some studies show a closer correlation
between physical criteria alone and GA compared to neurologic or total Ballard score. In a multicenter study,
infants who were small for their GA had consistent overestimation of their GA. (14).
Due to the inaccuracy of the NBS in extreme prematurity, one must continue to use maternal LMP and early
ultrasound as the gold standard for determining GA (see Chapter 12). This becomes imperative when deciding
aggressiveness of support in an infant born at 22 to 23 weeks by dates. Despite the aforementioned concerns,
the NBS remains the best method available to estimate GA in the presence of uncertain dates.
Other methods for GA assessment have emerged over the years. Prior to gestation of 27 weeks, the cornea is
too hazy to permit examination of intraocular structures. The vessels in the anterior vascular capsule of the lens
mature in a predictable enough pattern in the last trimester to allow determination of GA with a 2-week margin
of error in infants at 28 to 34 weeks of gestation. (20) The exam must be performed within the first 24 to 48
hours of life as the vessels atrophy rapidly after this time. It may prove useful when a neurologic abnormality
and uncertain dates render the NBS inadequate. As well, it has been validated in infants who are small for GA,
which compromises the accuracy of the NBS (21).
One final technique is the use of changes in skin reflectance during fetal development to determine GA of infants
at 24 to 42 weeks of gestation (22). A reflectance spectrophotometer at a wavelength of 837 nm obtains a
measurement of skin reflectance independent of melanin, yielding an estimate of GA unaltered by skin color.
Initially considered for use in black infants who were thought to be more neurologically mature at birth than
other ethnic groups, the requirement for special equipment and discovery that GA assessment tools were not
influenced by racial differences likely prevented widespread use of this technique (23).
Accurate estimation of GA requires experience and consideration of the infant's history and overall condition at
the time of scoring; for example, maternal medications or drugs and the infant's fetal position or sleep state
affect the neuromuscular response in a normal infant. Significant hypertonia or hypotonia is particularly powerful
in affecting the neuromotor scores but does not affect the physical maturity score. Examination as soon as
possible after initial stabilization or by 12 hours increases the accuracy in gestations shorter than 28 weeks (18).
The use of NBS is particularly
P.330
attractive for neonates who are immature and instrumented because it does not require lifting the infant.
Although described here separately from other components of the examination, the steps for assessing GA can
be done as part of the general physical examination and can provide information for the neurologic evaluation.
The neurologic examination of infants during the first year continues to use a number of these assessment
elements (24,25).
Figure 19-1 Assessment of maturity by the expanded Ballard score (18).
Neuromuscular Maturity
The resting posture is that observed with the infant in a quiet unrestrained environment. Tone increases in a
caudocephalad direction to a pattern of full flexion at term (Fig. 19-1).
The square window is assessed by flexing the wrist and measuring the minimal angle between the palm and
flexor surface of the forearm. This angle decreases with advancing
P.331
GA. Conditions of marked intrauterine compression, such as severe oligohydramnios, increase wrist flexion. As
an extremely premature newborn advances through corrected GA, he or she will not continue to develop as
much wrist flexion after birth as the infant would have had he or she stayed in utero.
The scarf sign indicative of shoulder and superior axial tone is assessed by pulling the hand across the chest to
encircle the neck as a scarf and observing the position of the elbow in relation to the midline. There is decreased
range and a higher score if there is marked obesity, chest wall edema, an abnormally shortened humerus, or
shoulder girdle hypertonicity. Brachial plexus injury or generalized hypotonia produces a spuriously low score.
With the infant supine and head midline, arm recoil is assessed by first flexing the elbow and holding the arm
against the forearm for 2 to 5 seconds. The elbow is then fully extended and released with observation of how
quickly and fully the infant resumes a flexed posture. Assessing recoil should not be done as part of testing arm
traction or with forceful extension because other responses may interfere with a normal reaction. Any pathology
affecting the motor strength or tone of the arm will decrease this score.
To determine the popliteal angle, one should first flex the hips with the thighs alongside the abdomen rather
than over the front. With the hips held in flexion and the pelvis flat, the knee is then extended as far as possible
to estimate the popliteal angle. If an infant was in frank breech presentation with legs extended, the popliteal
angles would be greater than expected for age.
In the heel-to-ear maneuver, the legs are held together and pressed as far as possible toward the ears without
lifting the pelvis from the table. The angle made by an arc from the back of the heel to the table decreases with
maturity.
Physical Maturity
Skin in the most premature infants is gelatinous and almost transparent, allowing the abdominal vessels to be
visible. It becomes opaque with maturity as it thickens and keratinizes, eventually shedding the lubricating
vernix after it dries and cracks.
Lanugo, which is the fine hair evenly distributed over the body, first emerges at 19 to 20 weeks, but for a few
weeks after initial emergence it is not readily apparent. Maximally apparent at 27 to 28 weeks, lanugo sheds
first from the areas of greatest contact. Lanugo is distinct from the more pigmented body hair that may be quite
prominent in infants of medium to dark complexion.
Assessment of the plantar surface includes measuring the foot because its length reliably corresponds to early
GA. With normal muscle activity and uterine compression, creases develop in the sole, progressing from the toes
toward the heel. Inappropriate sole creasing is seen in infants with serious neuromotor deficit in the lower
extremities (e.g., decreased creasing or only deep vertical creasing) or with oligohydramnios (e.g., increased
creasing).
The breast develops with an increase in color, stippling of the areola, and increase in the size of the breast
tissue. Although the volume of the breast somewhat depends on fetal nutrition and fat deposition, areolar
development with increasing GA is more consistent and independent of these factors.
Ear cartilage becomes firmer with gestation if there is no continuous, extrinsic pressure and the auricular
muscles have normal anatomy and activity. Concurrently, as gestation advances, the number of ear folds
increase as well. Unfusing of the eyelids can occur over several weeks, and a fused condition by itself is not a
sign of extreme, nonviable immaturity. Opening starts by 22 weeks; complete unfusing is evident by at latest 28
weeks (18).
Maturity of the external genitalia is one of the more reliable individual indicators of GA (18). Due to timed
descent in the third trimester, testicular progress through the canal into the scrotum is a GA marker. The testes
are usually palpable high in the scrotum at 36 weeks and fully descended by 40 weeks. For the scrotum to
develop fully into a pendulous, rugose, term appearance, testicular descent must occur at some time, even if the
sac is empty at the time of birth.
The appearance of term female genitalia depends on fat deposition and is abnormally immature in a poorly
nourished infant. The clitoris approaches term size well before 38 weeks so it is disproportionately large in
premature females (26,27). The appearance of a pigmented vertical line, the linea nigra, above the pubis toward
the umbilicus suggests a GA of at least 36 weeks.
Influences on Age Assessment Results

If there is a discrepancy between expected and achieved scores, factors that influence the demonstrated
findings should be sought before assigning GA.
Resting posture reflects progressive flexion into the position assumed when a term fetus occupies virtually all
available intrauterine space and is no longer in a freely floating state. To minimize the volume occupied, this
flexion logically progresses in a caudal to cephalic sequence. Compression from oligohydramnios accelerates
flexion; it gives an artificially advanced estimate of maturity. Conversely, poor fetal motor activity over a long
period spuriously decreases estimated age. Fetuses with markedly decreased motor activity and polyhydramnios
have decreased plantar and palmar horizontal creasing and inappropriately immature posturing. Any condition
affecting the position or activity of the lower extremities, such as frank breech presentation with hyperextended
knees or myelomeningocele with paresis leads to an aberrantly lowered neuromotor score. Similarly, an infant
who is hypotonic from illness or sedation has less flexion than normal for true GA.
GROWTH
Measurement Techniques
Infants weighing less than 2,500 g are low birth weight (LBW) regardless of GA; very low birth weight (VLBW)
P.332
refers to a weight less than 1,500 g, and extremely low birth weight (ELBW) indicates an infant weighing less
than 1,000 g. Classification by these weight groups helps establish the level of risk for neonatal and long-term
morbidity and mortality, particularly when the weight classification is coupled with an accurate GA (28).
Of all measurements, crown-heel length is the most subject to variability, because it depends on achieving full
extension of an infant who is more naturally in flexion. This measurement is performed with the infant supine,
neck neutral, leg fully extended, and ankle flexed to 90 degrees. If there is deviation from an expected norm,
the first step in evaluation is to remeasure the infant for confirmation. Although not part of routine practice, a
measuring board definitely improves accuracy.
When anomalies of the lower extremities make crown-heel length implausible, the crown-rump measurement
may still be feasible. Crown-rump length is measured with the infant supine and the hips flexed 90 degrees.
Congenital dwarfism may be classified as those with a short trunk, short legs, or both. These subtypes can be
readily differentiated by the crown-rump to total length ratio. From 27 to 41 weeks of gestation the value is
fairly consistent at 0.665 0.027 (29). The ratio is normal if a condition causes proportional reductions in
length of the upper and lower body; increased if the legs are shortened to a greater degree; and decreased if
the trunk is foreshortened. Standards for separate lengths of upper and lower limbs are available (30,31).
The head circumference is the largest dimension around the head obtained with a tape placed snugly above the
ears. This is the occipital frontal circumference (OFC). Head circumference undergoes a marked increase during
the last trimester, averaging 25 cm at 28 weeks and 35 cm at term (32). The average head circumference is 0.5
cm greater in male compared with female neonates (33). Due to greater reliability of repeated measurements,
paper rather than reusable cloth tape measures should be used (34). Minor changes in head circumference occur
during the first week after birth as scalp edema and molding resolve. The molding seen after prolonged breech
positioning can lead to an OFC that is as much as 2 cm higher than it will be after molding resolves.
The OFC predictably falls on the same percentile curve as the length. If the OFC differs from length by more
than one quartile, the cause should be sought because head size in part reflects brain growth. The most
frequent reason for a head percentile to exceed that of length is familial. In this situation the head
circumference follows a persistently higher but consistent growth curve. In contrast, pathologic macrocephaly
tends to cross to higher percentile curves as it progresses. A decreased rate of head growth, manifested by a
flat curve or by dropping to a lower percentile, may indicate poor brain growth, atrophy, or premature suture
fusion (craniosynostosis [CS]). As OFC may be normal in some forms of CS, the head width index (maximal
biparietal diameter divided by the OFC) and head length index (glabella to occipital prominence divided by the
OFC) may be more informative (35). For example, a patient with scaphocephaly secondary to sagittal suture
closure may have a normal OFC but have an abnormally small width index and excessive length index. Normal
ranges for these indices are available (35). The fetal head circumference exceeds the abdominal circumference
until 32 weeks. Between 32 and 36 weeks, the two circumferences are equivalent, and after 36 weeks the
abdominal circumference normally is greater.
Interpretation of Growth Parameters

Interpretation of growth parameters requires plotting the measurements on percentile charts constructed from a
similar race and environmental population. If birth weight falls between the 10th and 90th percentiles for a
given GA, the infant is appropriate for gestational age (AGA); if less than the 10th percentile, the infant is small
for gestational age (SGA); and if above the 90th percentile, the infant is large for gestational age (LGA). Some
literature cites the 3rd and 97th percentiles as outer limits, but for most clinical purposes this broader range
underselects for some at-risk infants, particularly in the lower sizes. Accuracy in GA assessment is critical in
determining if a weight is appropriate. AGA infants born at term are at lowest risk for problems associated with
neonatal mortality and morbidity (28).
Infants are considered symmetric if the three parameters of weight, length, and head circumference fall on the
same curve. The infant is asymmetric if the parameters are on different curves, usually with the weight on a
curve lower than those of the head circumference or length. If an infant has either a slowing of intrauterine
growth rate documented by serial fetal sonography or a presumed slowing by very low weight for length
measurements, he or she is classified as having intrauterine growth retardation (IUGR). All infants who fall
below the 10th percentile for weight are both SGA and IUGR. Infants above the 10th percentile are AGA but may
be IUGR if they did not achieve their expected growth potential. An infant who demonstrates a deceleration in
growth from the 50th to the 20th percentile during the last trimester due to maternal hypertension is such an
example.
Infants who are SGA, IUGR, or LGA are at risk for perinatal and long-term problems. Several of the problems
encountered by LGA infants include the following:
Iatrogenic prematurity due to overestimation of GA by late size estimates
Increased requirement for delivery by cesarean section
Pulmonary hypertension
Shoulder dystocia
Birth injuries
Ecchymoses
Local fat necrosis associated with forceps applications
Cephalohematoma, skull fracture
Brachial plexus injuries
Paralysis of diaphragm
Fracture of clavicle or humerus

P.333
Polycythemia
Jaundice
Hyperviscosity syndrome
Seizures
Renal vein thrombosis
Increased total blood volume
Poor feeding
Hypoglycemia
EXAMINATION
Examination Conditions
A routine neonatal examination, normally 5 to 10 minutes, should take place in a quiet, warm environment. The
room's light should be bright enough to detect skin markings and color but not so bright as to discourage open
eyes. When an infant is ill, attention to optimizing the environment and recognizing the potential effect of
noxious nursery surroundings on his or her state is fundamental.
Even healthy infants do not tolerate handling in extended examinations. The sicker or more immature they are,
the less they tolerate manipulation and environmental assaults. For all examinations, the prime consideration
must be that no harm should come by the process. In routine care situations, having one or both parents
present during an examination allows discussion about physical findings and offers the opportunity to point out
behaviors that can help them better understand their infant. They can address directly any questions about
history or therapy at that time as well.
General Assessment
The specifics of neonatal examination are discussed in the following sections. Some systems that are discussed
in more detail in other chapters are given less emphasis in this chapter than they would merit in an actual
examination.
Inspection
Inspection begins before making any physical contact and from enough of a distance to encompass the infant as
a whole. An immediate assessment of wellness can come from simply noting the state, color, respiratory effort,
posture, and spontaneous activity. Even simple observations of spontaneous movement patterns can suggest
future neurologic deficits or well-being (36).
State
Important indicators of infant well-being are the states or levels of arousal the infant achieves throughout the
examination and throughout the day as described by the parents or nursing staff. One categorization of states
listed here was originally defined by Prechtl and Beintema (37). Modifications have been made but are not
clinically important for general assessments (38,39,40,41).
Deep sleep
Light sleep
Awake, light peripheral movements
Awake, large movements, not crying
Awake, crying
During examination, a healthy infant should demonstrate several levels of arousal. The most useful states for
assessing an infant are those of light sleep and quiet awake so irritating maneuvers are held until the conclusion
of the assessment. What it takes to assist an infant in moving from one state to another or how well he or she
does it without assistance is noteworthy. Because the deep sleep that follows a recent feeding may give an
appearance of lethargy on arousal, knowing the feeding history and pattern is prerequisite to determining
aptness of state.
Newborns spend nearly two-thirds of each day in sleep (42). Each 24-hour period involves cycling between
periods of active sleep (AS) and quiet sleep (QS). These periods are also known as rapid eye movement (REM)
and non-REM sleep, respectively. During AS, infants demonstrate phasic limb movements, eye movements, and
irregular respirations. Breathing is typically rapid and shallow interspersed between periods of more regular
respiration (43). In comparison, QS is characterized by regular respirations and the absence of eye and limb
movements. As GA increases, the proportion of time spent in QS increases (44). Stress during the perinatal
period may alter the proportion of time spent between AS and QS as firstborns and infants born after cesarean
section spend more time in AS than babies from subsequent pregnancies or vaginal deliveries (42).
Quieting an infant may require anything from simply stopping the handling to holding and talking to him or her.
The amount of time spent in unstimulated crying is normally limited in the first 24 hours but may increase
significantly each day thereafter. Excessive crying that requires more than routine consoling, particularly if there
are no intervals of quiet alert states, indicates abnormal irritability, but other causes include a proper response
to pain or to a cold environment (45,46).
Color
Color assessment includes judging perfusion and skin color for the presence of cyanosis, (see color plate)
jaundice, pallor, plethora, or any unusual pigmentation.
Respiratory Effort
The degree of respiratory effort is a primary indicator of how distressed or comfortable a newborn infant is, even
if the cause of distress is not pulmonic. The examiner can observe the respiratory rate, depth of excursions, use
of accessory muscles with retractions or nasal flare, any emitted sounds (e.g., grunting or wheezing), and crying
pattern. Understanding the infant's pattern of respiratory
P.334
effort can suggest a specific illness and direct the examination. As the severity of a condition increases, these
distinctions may be lost (Table 19-1).
TABLE 19-1 PATTERNS OF NEONATAL RESPIRATORY EFFORTa
Conditions
Distal airway or lung parenchyma
Pattern Observed
Upper airway obstruction

Cardiac pattern
Neurodepression
Intercostal retractions, sternal retractions, flaring, tachypnea,

grunts, increased work in breathing
Suprasternal retraction, subcostal retractions
Tachypnea without effort, infant is quiet but not somnolent
Poor effort compared with physiologic need, apnea
Metabolic acidosis of sepsis
Tachypnea, apnea, lethargy, whining, minimal retractions
Early in disease process before patterns merge with multiple system involvement.
Posture
The normal resting postures at different GAs are shown in Fig. 19-1. While observing neck position, the
examiner looks for symmetry between the sides and compares the upper and lower extremities. If there is
lateral asymmetry and the head is turned to one side, there may be an asymmetric tonic neck reflex with the
extremities on the mental side in extension and those on the occipital side in flexion. In that case, the head
should be turned to the opposite side to verify that the asymmetry reverses.
If the fetal presentation is nonvertex or unknown or there is asymmetry or deformation, it is helpful to assist the
infant in assuming a position reflecting his or her intrauterine attitude. The physician can fold the extremities
into the fetal position by applying moderate pressure to a relaxed infant's feet while gently shaking the infant's
legs and by directing the arms toward the thorax through gentle pressure on the elbows.
Spontaneous Activity
The examiner should observe what the infant does in light sleep and awake states. Does the infant stretch,
move all extremities equally, open and close hands, root and start sucking when something touches his or her
face, and yawn with great facial expression, or does the infant lie quietly and move only in response to
stimulation?
Premature infants spend more time sleeping but should have spontaneous activity and resting postures
commensurate with their GA (47). Because they habituate and become disorganized and stressed quickly on
handling, inspection before contact in a benign environment is important.
Vital Signs
Temperature
It is unusual for neonates to develop fevers except in response to increased environmental temperature. If an
infant's skin temperature is above 38C and remains elevated after the environment returns to normal, a rectal
temperature should be obtained. Unless the temperature has been elevated for a prolonged time, the rectal
temperature is less likely to be affected by environment, and evaluation for infectious or neurologic causes is
indicated (48). Recurrent or profound hypothermia also requires additional evaluation.
In a warm environment, overbundling may cause temperature elevation into the febrile range (49). The infant's
postural response to hyperthermia is arm and leg extension, decreased spontaneous activity, and increased
sleep duration in order to maximally dissipate heat. Conversely, hypothermic infants assume a flexed posture to
conserve heat. During the first week of life, only 30% of infants born at less than 30 weeks are capable of limb
extension, but by 2 weeks this number increases to 87% (50). As such, premature infants in the first week of
life and hypotonic or myopathic infants are most at risk for temperature instability because they are less capable
of altering position to aid heat dissipation or conservation.
Term infants in the first day of life sweat in response to overheating, but not as efficiently as in a child or adult
(51). Infants less than 36 weeks, in comparison, are incapable of sweating on the first day but do so by 2 weeks
of age (52). Furthermore, the minimal temperatures required to induce sweating are higher in preterm than in
term infants. The first site capable of sweating is the forehead with recruitment of other sites happening in a
caudal direction. Visible sweating at rest or on feeding in an afebrile infant is abnormal and may indicate
distress, typically from cardiac disease.
Respiratory Rate and Heart Rate

The respiratory rate is obtained by looking at the upper abdomen for a full minute. As soon as an infant is
touched, the respiratory rate and depth change. The normal respiratory rate is 30 to 60 inspirations per minute
in a term infant, with lower rates occurring after the period of cardiopulmonary transition. When awake, some
normal infants breathe shallowly and rapidly but are able to slow sufficiently to feed well.
P.335
The heart rate is 110 to 160 beats per minute (bpm) in healthy term infants, but it may vary significantly during
deep sleep or active awake states. Preterm infants have resting heart rates at the higher end of the normal
range. Tachycardia, with a rate persistently greater than 160 bpm, may be a sign of many conditions, including
central nervous system (CNS) irritability, congestive heart failure, sepsis, anemia, fever, or hyperthyroidism.
Conversely, low resting heart rates may be observed following mild perinatal asphyxia.
Blood Pressure
Measuring blood pressure is not a routine part of vital signs in most newborn nurseries but is used for infants
requiring special care and for evaluating coarctation of the aorta. There are wide variations of normal at different
GAs (53,54,55,56). The Committee on Fetus and Newborn of the American Academy of Pediatrics states that
hypertension should be diagnosed only after three separate measurements (57).
The range of normal blood pressure in neonates depends on the method used for assessment and GA (see
Appendix C-1 for blood pressure values). The values obtained by the blanching and flush methods are mean
pressures and are lower than those registered by direct intravascular or Doppler monitoring. The flush method
for obtaining mean pressure is easier in an active infant and requires only a sphygmomanometer (58). The
Doppler methods, although providing diastolic and systolic pressures, require electronic equipment and a quieter
patient. Two important elements for obtaining accurate blood pressure are a quiet infant and a properly sized
cuff with a width two-thirds the length of the upper arm.
Facies
Assessment of facies includes looking for symmetry, size, shape, and the relations of all parts of the face and
how the infant holds or uses them. A seemingly unusual facial appearance dictates analyzing the individual
components to decide if the constellation represents malformation, deformation, a syndrome, or merely familial
appearance.
Head and Neck

Inspection of the head includes assessment of the shape and size relative to the rest of the body, face
distribution and character of the hair, and the underlying scalp.
Head circumference was discussed previously. Even when the OFC is normal, it is important to notice if the size
of the head seems appropriate for the size of the face. The shape of the cranial vault reflects interaction of
internal forces (i.e., brain anatomy, volume, intracranial pressure) against external forces (i.e., intrauterine and
extrauterine molding, suture mobility). Normal intrauterine molding for a vertex presentation leads to a
narrowed biparietal diameter and a maximal occipitomental dimension. After breech presentation, there may be
marked accentuation of the occipitofrontal dimension with parietal flattening, an occipital shelf, and apparent
frontal prominence. This normal breech shape requires differentiation from the abnormal occipital prominence
found in posterior fossa masses (e.g., Dandy-Walker malformation), frontal prominence due to increased cranial
volume, or the boat-shaped scaphocephaly from synostosis of the sagittal suture. Normal molding resolves
within a few weeks, but other aberrations progress. Pathologic conditions that prevent normal molding of the
fetal head should be suspected in infants who failed to engage in vertex and descend in labor. These conditions
are unusual but important to detect as early as possible.
Unusual head shapes are found in approximately 10% of newborns with the most common abnormality being
posterior or lateral plagiocephaly (i.e., flattening) (59). Associated risk factors are primiparity, assisted delivery,
prolonged labor, and twin pregnancies (59). Right-sided flattening occurs more often than left due to the more
common left occiput anterior descent during birth. It has been postulated that plagiocephaly present at birth
may contribute to the development of positional plagiocephaly in infancy. Initiatives to decrease the incidence of
sudden infant death syndrome through supine sleep positioning may accentuate plagiocephaly, as the infants
may lay on the side of the preexisting flattening (59). Plagiocephaly and torticollis often coexist; occipital
flattening with contralateral frontal prominence dictates determining range of motion for the neck. The infant's
head should turn as far as the shoulder in both directions; farther if it is premature. Frontooccipital
plagiocephaly may be manifested by a unilateral epicanthal fold or asymmetric positioning of the ears (60). To
fully assess position, the ears should be viewed en face and from the top of the head.
Hair and Scalp

The hair is inspected for color, texture, distribution, and directional patterns. Although hair color may change,
there should be racial concordance. For example, reddish or blond hair in a dark-skinned infant may indicate
albinism. The hair color should also be fairly uniform. Random patches of white hair are familial or sporadic and
inconsequential, but white forelocks with other pigment defects and anomalies are associated with deafness and
retardation (61). The texture of hair at birth is relatively fine. Most of the conditions associated with abnormal
textures or fragility appear some time later. With immaturity, the hair is even finer and more sparse.
The hairline may vary at the frontal margin, and normal but hirsute infants have hair well down the forehead but
without synophrys (i.e., eyebrows which have grown together). The posterior hairline has a more consistent
limitation, so that hair roots well below the neck creases, particularly at the lateral margins, suggest syndromes
associated with short or webbed necks. Although neonatal hair initially appears quite disheveled, its growth
direction normally is consistent.
P.336
In 97.5% of newborns, there is a single parietal hair whorl, located in or just to the right of the midline
approximately 80% of the time (62). Nearly 90% of these whorls have a clockwise rotation. More than two, an
isolated frontal or a significantly abnormally positioned whorl may be a sign of abnormal development of
underlying structures. Due to hair kinkiness, whorls are identified in only 10% of black infants. Consequently,
the absence of a whorl should not trigger unnecessary investigations of the CNS in this population. It is unusual
for hair to be strongly upswept at the neck or forehead at birth. If there is extreme hair unruliness with multiple
directions of growth, particularly with unusual facies, microcephaly, or SGA, there may be poor brain growth of
early fetal onset (63). Unruliness is typical but not diagnostic of a number of genetic syndromes, including
Cornelia de Lange and Down syndromes.
Superficial ecchymoses and abrasions of the scalp are common after vaginal deliveries, especially after
extraction by forceps or vacuum. Incision sites for fetal scalp electrodes or blood sampling should be small and
inconsequential, although some are deep enough to require closure. A small defect of the scalp, cutis aplasia,
that appears coincidentally at a potential monitoring spot may be confused with an electrode lesion because it
sometimes appears ecchymotic or blistered. Telangiectatic or staining lesions appear over the scalp, neck, and
face, ranging from the superficial and transient flammeus nevus or stork bite to the more intense, permanent
port wine stain. Scalp nodules less than 3 cm in size and covered by a port wine stain are not benign. This
lesion, encircled by a long thick whorl of hair is known as the hair collar sign. All of these nodules contain a
small encephalocele and thus neuroimaging is indicated (64).
An unusual finding is scalp rugae along with a normal scalp hair pattern. The skull beneath the rugae has
collapsed during fetal development and portends a very poor neurologic outcome (65).
Figure 19-2 Lateral and sagittal views of the cranial sutures. (From full reference citation, with permission).
Transillumination of the skull may detect large fluid collections, but the method has been supplanted by more
precise diagnostic techniques. Transillumination remains an important adjunct to examining the chest, abdomen,
and genitalia for fluid or air accumulations.
Palpation of the Head

Palpation detects motility and firmness of adjacent bones, size of the sutures, and bony or dermal defects. There
are six bony plates to the cranial vault; one frontal, two parietal, two temporal, and one occipital. Normally at
birth, these bones are separated by suture lines of which there are also six: metopic, sagittal, and paired coronal
and lambdoid (Fig. 19-2). Depending on the extent and direction of molding, there is variable overlap in the
sagittal, coronal, or lambdoid sutures. With CS, the fused sutures do not move freely when alternate sides are
pressed. Primary CS implies that the suture fusion is present at birth whereas secondary CS occurs when poor
brain growth culminates in premature closure of the sutures. This is an important distinction because while the
neurologic outcome for primary CS is often normal, that of secondary CS is almost invariably poor. If secondary
CS is present, one should evaluate the infant for other signs of thyrotoxicosis and exclude it as a cause (66).
Sutures that are overlapping but not fused feel like a mountain cliff, with a drop-off from the overriding bone to
the lower bone. Sutures that are fused feel more like a mountain range, with equal build-up on both sides.
Growth of the skull plates is perpendicular to the suture lines. If the sutures fuse prematurely, growth is
impaired, giving rise to the various forms of CS (Fig. 19-3). Isolated CS occurs in 0.6 of 1,000 live births, in a 2
to 3:1 male to female ratio, with affected sutures found to be metopic in 50%, sagittal in 28%, coronal in
16.5%, and lambdoid in 5.5% (67).
Any suture, particularly the metopic, may be normally widened in the absence of increased intracranial pressure;
P.337
the exception is a wide lambdoid suture, which ordinarily indicates increased pressure. Palpation of the bones
adjacent to the sagittal suture may reveal a give and return similar to that felt when pressing on an aluminum
can. These softer areas, indicating craniotabes, occur most often in premature infants or term gestations if the
fetal head was compressed against the maternal bony pelvis for several weeks. Physiologic craniotabes resolves
within a few weeks after birth (68). Pathologic craniotabes occurs in syphilis and rickets but may not be evident
from these associations in the early neonatal period.
Figure 19-3 The various forms of craniosynostosis displaying lack of growth perpendicular to the prematurely
fused suture line. (From full reference citation, with permission.
The fontanelles vary in size between and within race and by GA (69,70,71,72). There is little clinical application
for measurements in otherwise normal infants because head growth can occur despite apparently closed
fontanelles. Regardless of size, a pulsatile bulging fontanelle is a strong indicator of raised intracranial pressure.
The rate of fontanelle closure is independent of gender, growth parameters, and bone age (71,73,74). Infants
with a closed posterior fontanelle at birth have smaller anterior fontanelles, and those with smaller anterior
fontanelles also have smaller head circumferences (72). Although aberrantly large fontanelles are seen in
genetic syndromes and metabolic or endocrine diseases, they are not pathognomonic.
There are several other palpable findings on the head unique to the neonatal period. The most frequent, caput
succedaneum, presents at birth with pitting edema and initially is most prominent over the presenting area. It
represents fluid accumulation within and under the scalp. Although a caput initially may be limited to overlying a
single bone, it will shift to dependent regions and be more apparent in crossing sutures.
Cephalohematoma is less common and rarely is present immediately upon delivery. Any nonfluctuant swelling
that is palpable in the delivery suite is more likely a caput. Typically a cephalohematoma develops after delivery
and expands during the first few hours as blood accumulates between the surface of a calvarial bone and its
pericranial membrane (75). The cephalohematoma is rounded and discrete with boundaries limited by suture
lines. There may be pitting edema if a caput overlies a cephalohematoma, but an isolated cephalohematoma
feels fluctuant. Because of periosteal reflection at the margins, there is often a false sensation of underlying
bony depression. The blood contained in a cephalohematoma may take several weeks to resorb and prolong
neonatal jaundice. Any resolving cephalohematoma that begins to increase in size and becomes erythematous
should be aspirated to rule out infection. Unless there are neurologic indicators, radio-graphs to look for the
occasional underlying skull fracture are not indicated.
The least frequent scalp injury is a subgaleal hematoma, which may feel crepitant with less pitting and more
discoloration than found in caput succedaneum. Because there is little anatomic restriction to accumulation of
fluid under the aponeurosis, large amounts may redistribute
P.338
and deplete total body volumes in massive subgaleal hemorrhage (76). Cephalohematomas and subgaleal
hematomas occur most often after vacuum or difficult forceps extractions but may develop spontaneously even
in cesarean or unassisted vaginal deliveries (77). If a subgaleal hematoma is suspected, observation for tracking
of the swelling towards the nape of the neck and bluish discoloration will confirm the diagnosis. Serial
measurements of OFC should be made in the first 24 hours in all cases of suspected subgaleal hemorrhage; an
increasing OFC will allow for early identification and treatment of hypovolemia.
A thorough examination of the head includes auscultating for bruits over the temporal arteries and anterior
fontanelle, particularly if there are conditions involving high-output cardiac failure or neuropathology.
The neck should be extended for maximal exposure to look for branchial clefts or cysts anywhere from the ear
along the anterior border of the sternocleidomastoid. The isthmus of a normal thyroid is just palpable in the
sternal notch on neck extension; midline enlargements rarely represent a goiter. Other congenital neck masses
include cystic hygroma, lymphangioma, and cervical teratoma (78). A teratoma will usually be found in the
anterior midline with extension to the right, which differs from the more exclusively midline goiter (79). Any
mass may produce feeding difficulties, torticollis, and respiratory distress if airway compression is present.
Finally, a firm mass in the belly of the sternocleidomastoid is a fibroma, which may cause torticollis. If present,
the head tilt is to the involved side with the chin pointing away from the involved muscle.
Eyes
Examining the eyes of a neonate requires patience and a cooperative infant. Stimulating sucking or taking the
infant from a supine to an upright position and rocking gently back and forth may encourage spontaneous eye
opening in dim light; however, if the infant is crying inconsolably, delaying the examination is prudent. It is
inadvisable to pry the infant's eyes open as doing so will usually elicit crying and result in tighter eye closure.
The emphasis of the neonatal eye examination is on the structure and appearance of the eye and its
surroundings rather than assessment of visual acuity or extraocular muscles. The examination may proceed
from anterior to posterior structures: eyebrows, eyelids, eyelashes, eye socket, conjunctiva, sclera, cornea, iris,
and pupils.
The eyebrows are examined for symmetry and synophrys. Synophrys may raise suspicion that the infant has a
syndrome such as Cornelia de Lange or others. Do the eyelids rise and fall symmetrically? Failure of one side to
elevate may indicate a congenital ptosis, which may be due to several causes. A miotic pupil on the ipsilateral
side suggests the presence of congenital Horner's syndrome. Absent eyelashes may be a clue to a disorder of
the ectoderm and should prompt further evaluation.
Additional observations include determining the relative size, shape, and position of the eye in its socket and
demonstrating if the infant appears to have vision by reacting to light. One looks for symmetry and observes
whether the eyes seem to fit their sockets; they should sit neither too deeply nor too far forward. The size and
shape of the socket needs to be related to the size and shape of the surrounding skull. Marked molding with
depression of the forehead may make normal eyes appear to sit too far forward unless their position relative to
the cheeks is considered. Standard measurements for the eyes are avail-able (80).
Subconjunctival hemorrhages may be present following vaginal delivery. These are transient, disappearing after
a few days, and related to increased intravascular pressure during delivery.
Tearing or persistent eye crusting after the first 2 days calls for evaluation for glaucoma, infection, corneal
abrasion, mass lesions with obstruction of the nasolacrimal duct, or absence of the puncta. The signs of
congenital glaucoma that may be noticed during the neonatal examination include photophobia, excessive
tearing, cloudy cornea, or eyes that appear large (81).
The pupillary response to light requires a relatively dark room with only a moderately bright beam to avoid
stimulating reflex eye closure. The pupil diameter decreases toward term as its response to light increases. Pupil
reaction occurs consistently only after 32 weeks but may develop as early as 28 weeks of gestation. Pupillary
reaction and size are not affected by the presence of intraventricular hemorrhage as had been previously
thought (82). Pupils of term infants are anomalously dilated if their diameter is larger than 5.4 mm or
anomalously constricted if smaller than 1.8 mm (83). The pupils may appear discrepant in size (anisocoria) but
if more than 1 mm in diameter different, a cause should be sought (84).
Iris color is poorly defined at birth. The iris should form a continuous circle without interruptions or unusual
stretching or banding. In infants less than 28 weeks of gestation, corneal cloudiness permits only a cursory
inspection of the iris and pupils.
A thorough fundoscopic examination with mydriasis is not routine, but there should be an attempt to detect
cloudiness, masses, or large hemorrhages. Assessment of vision is best with an alert, quiet infant, but a startle
in response to a bright light flash through closed lids even with the infant asleep indicates intact optic pathways.
The corneal edema normally present in the first 2 days after birth may prevent an accurate detection of a red
reflex or visualization of the fundus without mydriasis. This is especially problematic in darkly pigmented eyes.
The red reflex is best visualized using a direct ophthalmoscope with the widest beam of light possible, held 30
cm from the infant such that both eyes are within the field of light (85). An abnormal red reflex indicates an
ocular abnormality anywhere from the cornea to the retina and warrants an urgent ophthalmologic assessment
to determine its exact location.
P.339
Ears
Each ear is examined for shape, size, position, presence of a canal, and any extraneous tags or pits. The ear
canal should be examined, especially in the presence of an auricular abnormality. However, the presence of
vernix and debris from birth may obscure visualization of the tympanic membrane until after the first few days
of life. The shape of the external ear is determined in part by intrauterine forces and by the activity of extrinsic
and intrinsic auricular muscles. Abnormal formation may be a sign of neuromuscular weakness or abnormalities
in auricular muscles (86,87). This is quite uncommon, with a recent estimate of 0.011% having a malformed
auricle at birth (86). The length of the ear roughly approximates the vertical distance from the arch of the brow
to the bottom of the nose. Ear length measurements at different GAs are available (88). Posteriorly rotated or
low-set ears occur when cephalad migration and anterior rotation fail to complete. Molding or deformation from
the birth process may also yield an abnormal position, but this will resolve after a few days. The position at term
should be similar on both sides, with at least 30% of the pinna above a line extended between the medial canthi
(89). Because a line extended between the medial and lateral canthi on one eye reflects variable slanting of the
eye, it is better to use a more consistently positioned marker such as the two medial canthi.
A behavioral reaction to a standardized sound excludes only gross bilateral deficits, but it should be elicited in all
neonates. Assessment of hearing by brainstem-evoked potential or otoacoustic emission is specifically indicated
in infants at risk for hearing deficits, particularly those with anomalies of the head and neck, a family history of
childhood deafness, very low birth weight, severe asphyxia, fetal infection, meningitis, severe jaundice, and
intracranial hemorrhage. Even though congenitally acquired hearing deficits may result from anomalies,
infections, or other perinatal conditions, loss may not develop for months or be detectable until behavioral
audiometry is feasible. Early detection of hearing deficits may improve with the institution of universal hearing
screening, which is becoming standard in many institutions worldwide.
Nose
The nose is assessed for shape, size, patency, presence of swelling over the nasolacrimal duct, size of the
philtrum, and definition of the nasolabial folds. It should appear appropriately sized for the face when viewed
laterally and en face.
Nasal deformation with asymmetry of the nares and apparent deviation occurs as part of facial compression and
molding. The triangular cartilage rarely may be dislocated during delivery causing septal deviation, which is best
treated by surgical relocation during the first week. With depression of the tip of the nose, a dislocated septum
appears even more angled within the nares but a normal septum merely compresses. After release, a dislocated
septum does not return to upright nor can it be readily molded into a normal shape (90).
Nasal patency is assessed by free passage of a small catheter through both nares and into the stomach. Air flow
is detected by holding a strand of thread in front of each nostril and observing fluttering with breathing.
Congenital obstruction of the nasolacrimal duct occurs in approximately 20% of newborns, 95% of whom are
symptomatic within the first month of life (91). Common signs are a large tear meniscus at the lower lid, tearing
without stimulation, dried mucoid residue after a nap, or a discharge during waking. A distally obstructed
nasolacrimal duct is diagnosed by pressing the finger over the lacrimal sac and sliding it along the course of the
duct toward the eye to express material from the puncta. Dacryocystocele is a dilation of the lacrimal drainage
system due to obstruction at both ends and filling of the enclosed space. Dacryocystoceles are observed at birth
as immobile, tense, sometimes blue-gray cystic swellings no more than 1 cm in length located just below the
medial canthal tendon (92).
Mouth and Throat

The shape of the mouth is a marker of fetal position and neuromotor activity. For the mouth to develop properly,
there must be muscle activity of the tongue against an intact hard palate. If there is a large palatal defect so the
tongue does not meet superior resistance, the mandible is more receded and small. If the hard palate is intact
but the tongue inactive, the hard palate may have a high arch or have prominent lateral palatine ridges.
The more common oral findings have counterparts in neonatal dermatology and are benign (Table 19-2). The
mouth should be observed with the infant at rest and crying. The shape and size of the mouth is best
determined by looking at the mandible and how well it fits the maxilla. It should open at equal angles bilaterally.
If the head was tilted in utero for an extended period just before delivery, there may be mandibular deviation
causing the jaw to open at an angle. This deformation resolves spontaneously, but significant oral asymmetry
causes difficulty in breast-feeding on one side compared to the other. Asymmetry on crying occurs with facial
nerve paresis, in which the nasolabial folds are asymmetric, or with absence of the depressor anguli oris muscle,
in which the folds are symmetric (93, 94). The side with the absent muscle feels thinner.
The tongue, buccal surface, palate, uvula, and back of the mouth should be visualized. The gums and hard
palate are best assessed by palpating with a gloved finger while the strengths of the suck and gag reflex also
are assessed. A bifid uvula should alert one to the presence of a submucous cleft palate.
Oral cysts in various locations may be present in up to 80% of newborns (95). One to six pairs of small benign
midline cysts known as Epstein pearls may be present at the junction of the hard and soft palates. Other cysts
may be found along the maxillary or mandibular alveolar ridges or on the buccal surfaces. These cysts are
benign and often
P.340
resolve by 3 months of age. Epstein pearls are always found in the midline except when there is a submucous
cleft, in which case they appear as paired cysts along either side of the median raphe (95).
TABLE 19-2 NEONATAL ORAL FINDINGS
Finding
Palatal cysts (e.g., Bohn nodules,

Epstein pearls)
Alveolar or gingival cysts

Alveolar lymphangioma
Alveolar eruption of cysts with or

without teeth
White (%)
Nonwhite (%)
Comments
73-85
65-79
Yellow-white elevated cysts 1 mm

in diameter; nests of epithelial
cells in the midpalatal raphe at the
54
0
<0.1
40
4
<0.1
fusion points of the soft and hard

palates
Appear similar to palatal cysts
Blue-domed, fluid-filled cysts in
posterior regions; no more than
one per quadrant; may cause
discomfort during feeding if cysts
are large
Clear, fluid-filled cysts; mandibular
central incisor; rates range from
1:2,500 in Hong Kong to 1:3,392
in Canada
Leukoedema
11
43
Median alveolar notch
16
26
Ankyloglossia
~2
~2
Filmy, white hue of mucosa,

nonblanching; of no significance
compared with thrush
Reduces when teeth erupt or
persists as notch between central
incisors
Male to female ratio of 3:1; lingual
frenum prevents protrusion of
tongue, extends to papillated
surface of tongue, or causes
fissure in tip
Commissural lip pits
Blind-ended pits at corners of

mouth; autosomally dominant;
associated with preauricular pits;
medial pits more syndromic
Thrush
Bifid uvula
Ranula
Epulis
<1
<1
<<1
<<1
<<1
<<1
Adherent white plaques on tongue

and buccal and palatal surfaces;
will scrape off; caused by Candida
sp.
Associated with submucous cleft
palate
Cyst of sublingual salivary gland
Large, pedunculated cyst of incisor
region
Data from refs. 120,121,122,123,124,125.

Natal teeth, often in pairs, may be found most commonly along the mandibular alveolar ridge. These fairly
mobile teeth are present in 1:2,000 to 3,000 live births and are often removed due to concerns of aspiration
(96). Despite the probable low risk of aspiration, removal should be considered due to ulceration of the ventral
surface of the tongue, which may occur if they are retained (97).
A prominent tight frenulum (ankyloglossia or tongue tie) has been cited as causing poor latch and nipple pain
during breast-feeding. While prophylactic frenuloplasty cannot be advocated, it may be considered for the infant
with severe persistent feeding problems or nipple pain unresponsive to usual therapeutic interventions (98).
Other abnormalities involving the tongue include one that is too large, preventing mouth closure, and tongue
protrusion because of low oral-facial tone or small oral cavity.
Skin and Lymph Nodes

The skin is assessed for general color, presence of any extra markings or rashes, texture, turgor, edema or
areas of induration, thickness of underlying fat, and maturity. Icterus progresses in a cephalocaudal pattern and
is best appreciated in natural light. Detection may be improved by gently blanching the skin to remove the blood
from the dermal capillaries. Levels requiring phototherapy are seldom attained if icterus is not present past the
nipple line (99).
The distribution and types of pigmentation should also be noted. Areas of hypo- or hyperpigmentation such as
caf au lait spots are important to document as they may be associated with neurologic disorders.
A common finding is the presence of a line of demarcation with pallor above and deep erythema in the
dependent area below (harlequin color change). Another frequent skin pattern is marbling of the extremities
from vasoconstriction (cutis marmorata) occurring when the infant undergoes hypothermic stress.
If any mark on the skin is to be considered to be a birth mark, there is rarely an infant born without several.
In view of the many variations of normal and the important signs of other diseases or syndromes that are
manifested on the skin, understanding the common physical findings of the neonatal skin is fundamental.
Lymph nodes are palpable in more than one-third of all neonates, most commonly in the inguinal region and
independent of perinatal history. These nodes, from 3 to 12 mm in diameter, tend to persist (100). A benign
node commonly palpable in the femoral triangle serves as a reliable landmark for finding the femoral pulse
adjacent to it.
The most frequent, abnormal congenital lymphatic masses are cystic hygroma or cystic lymphangiomas, which
P.341
are soft, compressible, and often poorly defined masses in almost any part of the body, but commonly in the
head, neck, abdomen, and axilla. If markedly distended, the hygroma may transilluminate. Ultrasonography
reveals their cystic anatomy. A discussion on the congenital abnormalities of the lymphatic system is available
(101).
Chest and Abdomen

Size and Symmetry
In term neonates, the chest circumference is 1 to 2 cm less than the head circumference; it is relatively smaller
with lower GA. The thorax is normally symmetric and wider than its anteroposterior dimension. The ribs are
compliant, and their shape is easily impacted by external and internal form. Compression from the infant's own
arm or a twin's body part may lead to marked asymmetry in thoracic shape and pattern on inspiration. By
encouraging the infant to assume the fetal position, the cause of a chest deformation may become apparent.
The abdomen is mildly protuberant compared with the chest. It should be softly rounded, with a diameter
slightly greater above than below the umbilicus. The abdominothoracic relation is reversed in diaphragmatic
defects, with herniation of abdominal contents into the thorax leaving a scaphoid abdomen. Diastasis recti, a
separation between the rectus abdominus muscles, is a normal finding during the newborn exam. Supraumbilical
fullness is increased in the presence of duodenal atresia with gastric distension or hepatomegaly, and
infraumbilical fullness is increased with distension of the urinary bladder or in severe cases of IUGR with an
abnormally small liver. Any significant abdominal visceral enlargement causes distension, as does forced
depression of the diaphragm.
Retractions
Mild subcostal and intercostal retractions are common even in healthy neonates because of their compliant chest
walls. Suprasternal retractions, indicating proximal airway resistance, are normally less pronounced;
supraclavicular retractions are never normal. In conditions notable for loss of lung volume and poor compliance,
respiratory movements may become paradoxical (i.e., seesaw) with a collapse of the chest wall on inspiration as
the abdominal wall expands. With air trapping and increased thoracic volume, there is an increase in the
anteroposterior dimension and abdominal distension as the diaphragm is pushed down.
Because the diaphragm is the primary muscle of breathing with little contribution by accessory muscles, quiet
breathing is abdominal, with only mild but equal subcostal retractions. The umbilical stump moves caudally in
the midline with each contraction of the diaphragm. In the absence of abdominal abnormalities, any lateral
deviation of the umbilicus with inspiration suggests a diaphragmatic paresis with the deviation toward the
nonfunctioning side (102). This belly dancer sign is lost during mechanical ventilation. Albeit rare, neonatal
diaphragmatic paresis occurs most often with brachial plexus injuries, and it should be considered if an arm is
weak.
Auscultation of Breath Sounds

Neonatal lung sounds are relatively more bronchial than vesicular because of better transmission of large airway
sounds across a small chest. Changes in the pitch of bronchial sounds from one side to the other or between
regions most likely represent main stem or conducting bronchial narrowing. The coarse crackle of a distal airway
opening is particularly common in neonates in whom there is a natural tendency in many conditions for
microatelectasis. If heard at the end of inspiration, adventitial sounds represent more distal disease compared
with those in beginning inspiration, which usually represent conducting airway secretions. A characteristic sound
of crushing styrofoam or walking on dry snow signals pulmonary interstitial emphysema.
Locating abnormal sounds requires listening carefully bilaterally. Breath sounds should be symmetric when
auscultating right and left corresponding regions of the chest. If they are not, air leaks or fluid collections or
consolidations may be present and the cause should be identified. If bowel sounds are heard, the examiner
should differentiate direct transmission from herniated contents or referred abdominal sounds.
If there is stridor or wheezing, auscultation of the nose or throat may reveal a site of extrathoracic obstruction.
Because there must be sufficient air flow to cause stridor or wheezing in the first place, stimulating the infant
may accentuate an obstruction. If an endotracheal tube is in place and an air leak is present, a whistling sound
corresponding to ventilator breaths may be evident in the periphery of the lung fields but is loudest over the
upper trachea or outside the mouth.
Clavicles
Clavicular fracture is the most common form of birth trauma. If carefully sought by radiographs or repeated
examinations, clavicular fractures are found in at least 1.7% to 2.9% of term deliveries and more frequently on
the right side (103). These fractures most commonly occur in pregnancies complicated by macrosomia, shoulder
dystocia, or operative vaginal delivery. Most clavicular fractures are asymptomatic and are incidental findings on
a chest radiographs. When symptomatic, the most frequent findings are swelling from hematoma, crepitations,
asymmetrical bone contour, and crying with passive movement. There may be an associated brachial plexus
injury or pseudoparesis, with poor movement in the affected arm and an asymmetric Moro reflex. As well, an
infant may not be willing to breast-feed on one side because of discomfort with positioning. Standing at the foot
of the infant, the examiner feels each clavicle, compares ease of outlining the distinct borders of the bones, and
assesses tenderness,
P.342
swelling, or crepitation. Finally, the clavicles may be hypoplastic or absent as in cleidocraniodysostosis. If they
are absent, the shoulders may be made almost to touch in the anterior midline.
Nipples
The breasts of term infants vary in diameter from 0.5 cm to several centimeters, with clinically insignificant
differences between genders. The internipple distance varies with GA and body weight, but its relation to chest
circumference is more constant. If the internipple distance divided by the chest circumference is greater than
0.28 cm, the space is more than 2 standard deviations above the mean regardless of body size (104). Larger
breasts, influenced by maternal hormones, may secrete a thin, milky substance (i.e., witches' milk) for a few
days or weeks. Although the degree of enlargement may not be the same in both breasts, they should never be
hot, red, or notably tender. Unless there are specific signs of inflammation, enlarged breasts should be left alone.
Supernumerary nipples occur in 1.2% to 1.6% of darkly pigmented infants but are more unusual in lightly
pigmented infants. These supernumerary nipples, seen in the milk line below and lateral to the true breast, are
rudimentary, occasionally only distinguishable because of the presence of a small pigmented mark or dimple.
Speculation that infants with supernumerary nipples have a higher incidence of renal and urinary tract
malformations has been refuted (105).
Umbilicus
The umbilicus normally is positioned approximately halfway between the xiphoid and pubis. A caudally placed
insertion occurs in conditions of caudal regression or underdeveloped lower body segment. The neonatal
appearance of the umbilicus does not indicate what the adult appearance will be because most are relatively
protuberant with redundant skin.
The umbilical cord is assessed for appearance, length and diameter, number of vessels, and insertion site. The
cord is a uniform ivory color ranging in length from 30 to 100 cm; a shorter cord suggests decreased fetal
movement and a reason for fetal distress, failed descent, or avulsion. Deep green staining of the cord is a sign
of prior fetal distress reflecting the passage of meconium at least several hours prior to delivery. Superficial
staining reflects very recent passage of meconium. Longer cords are more likely to result in fetal entanglement
or prolapse. At term, the cord diameter is an average of 1.5 cm and is relatively uniform throughout its length,
without strictures. If the base of the umbilical cord itself is especially broad or remains fluctuant after vascular
pulsations have stopped, there may be a herniation of abdominal contents into the cord.
At birth, the presence of two arteries and a single vein should be identified. Single umbilical arteries occur in
approximately 1% of pregnancies with nearly 10% of identified cases having another congenital malformation. A
thin cord with a paucity of Wharton's jelly is present in neonates with IUGR and may be compressed more easily
by fetal parts. As the cord dries, it should remain odorless. The base should not appear red or indurated. After
the cord falls off, the umbilicus should be examined for granuloma or continued leakage through a patent
urachus.
Palpation of Abdomen
The infant tolerates palpation of the abdomen best when the organs are brought to the examining hand rather
than the fingertips pushing into the abdomen and probing for the organs. Standing at the right side of the
infant, with the left hand lifting the legs and raising the pelvis slightly off the mattress to relax the abdominal
muscles, the examiner can keep the right hand flat and use the fingerpads rather than fingertips to palpate the
abdominal organs. Palpation should start below the umbilicus on both sides and proceed toward the diaphragm.
In some instances, it is helpful to palpate the abdomen with the infant in the decubitus or prone position,
allowing the contents to fall toward the hand rather than being pushed away (106). Palpation of the abdomen in
ill neonates increases centrally measured blood pressure by as much as 25% above baseline (107). The liver is
normally palpable 3.0 0.7 cm below the costal margin in the midclavicular line and across the midline as a left
lobe that is distinguishable from the spleen, which is felt more laterally (108). Finding a left lobe larger than the
right may reflect situs inversus. A liver edge that is palpable greater than 4.4 cm below the costal margin is
indicative of hepatomegaly although hyperinflated lungs may falsely give this impression (108). Assessing the
liver span in this circumstance will provide a better assessment of liver size (109). By 34 weeks, the normal liver
span, determined by percussing the upper and lower margins, is at least 6 cm in the midclavicular line (110). An
exception to this is infants who are SGA in whom the span may be up to 1 cm less (110). An effective technique
to outline the margins of the liver or any solid abdominal mass is to scratch lightly across the skin surface while
auscultating with the diaphragm of the stethoscope held over the mass. The pitch elevates when the stroking
overlies the solid mass or liver.
The normal edge of the liver is thin and soft, and the hepatic surface is smooth. A full or firm edge commonly
represents a marked increase in total blood volume, increased extramedullary hematopoiesis, chronic infection,
early cirrhosis, or an infiltrative process. Hepatomegaly is a late and inconsistent finding in cardiac failure.
Cardiac pulsations in the liver occur in right-sided obstructive cardiac lesions, but these hepatic pulsations
should be differentiated from a normally transmitted cardiac impulse or res-piratory excursions. In the first 24 to
48 hours after birth, the liver often decreases markedly in size, probably reflecting redistribution of circulating
blood volume.
Although the spleen is often not palpable in the newborn period, an attempt should be made to palpate it.
Beginning in the right lower quadrant, one gently palpates with the
P.343
right hand in a direction cephalad and to the left side of the abdomen until the fingers come to rest under the
left subcostal margin. Successful palpation may be achieved by placing the left hand under the left posterior
subcostal margin and applying gentle anterior pressure to bring the spleen forward. If this maneuver is not
fruitful, then positioning in the right lateral decubitus may displace the spleen in an anteriomedial direction,
allowing for easier palpation.
The kidneys are palpable if the abdomen is soft, and they are moderately firm and rounded. It is often easier to
palpate the right kidney as it is displaced more caudally than the left by the liver. An enlarged ureter simulates a
filled segment of large bowel, although it is less mobile. A fullness in the lower abdomen may be a distended
bladder in an infant with infrequent voiding.
An infant reveals abdominal tenderness by a grimace, cry, or drawing up of the legs on light palpation. True
guarding is unusual. Rebound tenderness is difficult to detect, and infants with significant peritoneal disease are
often too obtunded to show a reliable response. The presence of localized edema or discoloration of the
abdominal wall is an important indicator of intraperitoneal disease. An unusual exception is ecchymosis caused
by leaking of an umbilical vessel or urine edema from a patent urachus leaking into the subcutaneous space
above the peritoneum. In either case, the dramatic findings are limited to the abdominal wall below the
umbilicus. A thin abdominal wall allows transillumination of fluid- or gas-filled masses to outline their position
and size. Meconium-filled bowel loops do not transilluminate, but stomach or bowel distended with air,
hydronephrotic kidneys, or a distended bladder will. A transillumination pattern that shifts with patient rotation
suggests free air.
Auscultation of the abdomen includes listening for pitch and activity of bowel sounds and for bruits. Infants
normally have relatively inactive bowel sounds on their first day of life or if they are extremely premature and
not fed for several days or weeks. Even in infants with clinical ileus, bowel sounds tend to persist to some
extent; however, a true absence of bowel sounds is always significant. Detecting changes in the pattern of bowel
sounds is more helpful than the findings of a single examination. Auscultation may reveal the presence of a bruit
over the liver, indicating an arteriovenous fistula, or over the kidneys in the presence of renal artery stenosis.
Cardiovascular System
The changes that occur in the cardiovascular system during the neonatal period complicate the cardiac
examination until the pulmonary and systemic pressures reverse their fetal associations, all communications
have closed, and the left ventricle becomes predominant. The role for most clinicians in the newborn
examination is not to determine precisely what the cardiac anatomy is but to rule out cardiac disease as part of
a routine newborn examination, and in a symptomatic infant, to determine if the cause of the symptoms is
cardiac.
The physician must determine the urgency of the condition by asking some basic questions. Is this a cardiac
disease that could be fatal if not immediately diagnosed and treated (e.g., ductal-dependent lesions, cyanotic
heart disease)? Is its presence aggravating or relieving other conditions (e.g., patent ductus arteriosus [PDA] in
the presence of lung disease or pulmonary hypertension)? Is this something that requires following the patient
and potential future intervention but is not emergent and should not interfere with newborn and parental
adjustment (e.g., mild pulmonic stenosis or a small septal defect)?
Palpation of the chest may be informative in several ways. The position of the point of maximal impulse (PMI)
may be displaced downward and laterally from its common location in the fourth or fifth intercostal space in the
midclavicular line. This finding suggests cardiac enlargement, which should be followed by palpation at the lower
edge of the xiphoid. A strong impulse in this location is indicative of significant right ventricular enlargement,
which could reflect right-sided or biventricular enlargement as the cause of PMI displacement. Additional
information obtained from palpation may be parasternal heaves, another sign of ventricular enlargement, or
thrills in the presence of significant murmurs.
Careful auscultation of the chest will reveal two heart sounds with an occasional splitting of the second heart
sound due to changes in pulmonary blood flow with normal respiration. This split may be difficult to appreciate,
however, in infants with heart rates in the upper range of normal or in those with frank tachycardia. In addition
to examining the heart sounds, it is important to palpate the strength of the peripheral pulses, taking special
note of the intensity of the femoral pulse relative to the brachial pulse. The femoral pulse is located just lateral
to the femoral triangle beneath the inguinal ligament. The typical radial-femoral delay observed in older children
with coarctation of the aorta is difficult, if not impossible, to appreciate in the setting of such rapid heart rates. It
is critical, therefore, to establish the strength of the femoral pulses and if weak or not palpable, urgent
echocardiographic examination is required. Bounding peripheral pulses are indicative of a run-off situation, such
as a PDA with marked left to right shunt or less commonly an arterio-venous malformation.
Murmurs persisting after the first 12 hours are likely to reflect structural abnormalities even though they may
not be hemodynamically significant. In one study of infants referred to the cardiology department for murmurs
between 12 hours and 14 days, 84% had identifiable lesions, with ventricular septal defects (39%), pulmonic
stenosis (15%), and PDA (15%) most common (111). The remaining 16% had normal hearts with innocent
murmurs due to tricuspid regurgitation or peripheral pulmonic stenosis. Very commonly a systolic murmur from
a closing PDA will be present in the first 24 to 48 hours of life. The key to diagnosing this murmur is serial
examinations to confirm that in fact the murmur has disappeared. Whether or not echocardiac evaluation is
warranted requires clinical judgment if there is no option for reexamination in a timely manner. It is important
P.344
to remember that infants with the most serious forms of congenital heart disease may have no murmur but
often will have other clues in the cardiac examination.
Evaluation of the cardiovascular system begins in the delivery room with assessment of the Apgar scores and
includes evaluation of heart rate, color, and respiratory effort. Frequently, a line of demarcation is observed,
with the head, right arm, and right side of the chest pink and the rest of the infant pale or cyanotic until there is
functional closure of the ductus. With vigorous crying, its disappearance indicates an appropriate drop in
pulmonary vascular resistance and transductal shunting. Another reassuring milestone in cardiac transition often
noticed at the first bath by nursing staff is a brief but bright red flush over the entire body and extremities. This
blush, reminiscent of cooked lobster, is distinguishable from the darker, ruddy, plethoric color of polycythemia,
which is accentuated in the mucous membranes and less so on the palms and soles and is more persistent. The
blush is not seen in infants with cyanotic cardiac disease. Specific points to be considered in the cardiac
examination are outlined in Table 19-3.
Genitourinary System
In the delivery room, one of the first documented observations of the neonate is assignment of gender. Genital
abnormalities are relatively uncommon but cause significant stress to new parents, and so it is important to
distinguish the variations of normal from pathologic malformations. It is always urgent to start an appropriate
evaluation of gender if it is in question.
The male infant should be examined by stretching the penis for an expected penile length at term of at least 2.5
cm. Although previously unrecognized, a recent study has indicated that there may be racial differences in the
size of infant genitalia. As such, infants having measurements at the extremes of the standard length curves
should be compared to others from their racial background if this information is available (112). The presence of
chordee prevents complete stretching, but a twisted median raphe is of no significance. In obese infants, the
shaft may be retracted and covered by suprapubic fat, appearing to be too small unless it is stretched. The
observation of the presence of erectile tissue essentially eliminates true micropenis as a consideration. The
meatal opening should be located although completely retracting the foreskin is unnecessary. Any significant
glandular hypospadias generally is accompanied by incomplete foreskin and therefore is readily apparent on
simple inspection. Fortunately, the newborn frequently provides opportunities to observe the origin, direction,
and force of his stream on urination.
The presence of both testes deep in the scrotal sac indicates term gestation. If a testicle is not felt within the sac
or canal, use a lubricated finger to sweep from the anterior iliac crest along the canal while palpating the
scrotum. The volume of the testes should be estimated. Table 19-4 summarizes the normal values. If the
scrotum or a testis is distended but soft and nontender, transillumination may reveal a hydrocele. Deep
discoloration suggests hematoma or torsion and a need for immediate surgical evaluation, but superficial scrotal
cyanosis (see color plate) may represent benign ecchymosis after breech presentation. Only in rare instances
will a salvageable testicle be found as most often it is a remote vascular insult and was resorbed before birth
(vanishing testicle). Hydrocele of the cord, a harbinger of inguinal hernia, is not so likely to transilluminate but
is easily felt.
The female genitalia should be inspected for size and location of the labia, clitoris, meatus, vaginal opening, and
the relations of the posterior fourchette to the anus (Table 19-4). Virtually all female newborns have redundant
hymenal tissue. Hymens tend be annular with a smooth or fimbriated edge and a central or ventrally displaced
opening. Tags of tissue may extend from 1 to 15 mm beyond the rim of the hymen and occur in at least 13% of
female neonates. These tags disappear within a few weeks. A complete review of hymenal variations in
newborns is available (113). An imperforate hymen can present with a hydrometrocolpos, a build-up of mucoid
or bloody secretions causing a mass protruding from the vagina, which usually resolves with spontaneous
rupture or regression but can enlarge significantly and cause urinary obstruction or apparent discomfort.
Assessment for virilization in the female is difficult because there are varying degrees of clitoral hypertrophy and
labioscrotal fusion. With clitoral size realized by 27 weeks of gestation but with little deposition of fat in the
labia, there is particular confusion about clitoral hypertrophy in premature infants. As gestation progresses, the
labia majora enlarge and by term should completely cover the labia minora. Masculinization causes posterior
fusion of the labioscrotal folds independent of clitoral hypertrophy. The distance of the anus from the posterior
fourchette varies by GA and body size, but its relation relative to other genital landmarks is more constant
(Table 19-4). Measurements are made with the hips flexed and the infant relaxed so that the perineum does not
bulge. It is important in both genders to identify a normally positioned anus. Anterior displacement of the anus,
while not problematic in the early months of life, frequently causes significant constipation after the stools
become more formed.
Musculoskeletal System
Examination of the spine includes observation for abnormal curving and cutaneous manifestations of underlying
deformities such as sacral agenesis or spina bifida. A pilonidal sinus is suspected if the bottom of a sacral pit is
not visible or there is moisture in an otherwise dry area. Long tufts of hair, an overlying hemangioma, or
pigmented nevus potentially indicate a tethered cord unless well below the origin of the cauda equina. A
palpable mass usually indicates a lipoma if covered by normal skin that moves with it. A sacrococcygeal
teratoma tends to be a fixed mass just lateral to midline, and spinal dysraphism presents as a midline mass,
most frequently without full skin coverage. One assesses the extremities for symmetry, size and length, range
P.345
P.346
of active and passive motion, and obvious deformity. The length of the upper extremities should allow the
fingers to reach to the upper thighs on extension. The muscles are not well defined but should not feel atrophic
or fibrotic.
TABLE 19-3 NEONATAL CARDIAC EXAMINATION

Finding
Color
Key Location
Over entire surface

except presenting
part; inside oral
mucous membranes
Points to Consider
Peripheral cyanosis may

include area around
mouth but not inside
mucous membranes
Prominent venouscapillary plexus around
mouth and eyes simulates
cyanosis
Acrocyanosis of
extremities reverses with
warming
Mild cyanosis may appear
as pallor or mottling
Infant with PDA run-off
looks washed out,
particularly in the feet
Respiratory patternLateral view of chest Most often have
and abdomen
respiratory rate within
Alae nasi
normal range
May be cyanotic but
tachypneic without
distress (e.g., retractions,
labored breathing) unless
there is pulmonary edema
or severe acidosis
Heart rate rhythm PMI
Resting rate 120-130
bpm (range 100-150);
higher second to fourth
weeks and in premature
infants
Most premature beats are
transient and benign
Precordial bulge
PMI
Thorax compared
side to side and to
the abdomen
Thoracic asymmetry
indicates bulge with AVM,
tricuspid regurgitation (i.
e., Ebstein anomaly),
tetralogy with absent PV,
intrauterine arrhythmia,
or myocardopathy
Most commonly,
asymmetry indicates
pneumothorax,
diaphragmatic hernia,
atelectasis, or lobar
emphysema
Left parasternal area Visible until 4-6 h of life
during transition; beyond
12 h, associated with
volume overload lesions
(e.g., AP shunt,
transposition, or outflow
obstruction)
Normally more visible in
premature infants but
increases with PDA
Abnormal to have PMI
beyond 1-2 cm left of LSB
at less than 1 wk of age
Right sided indicates
dextrocardia versus shift
due to intrathoracic
pressures
Absence of increased
impulse with cyanosis
indicates pulmonary
atresia, tetralogy, and or
tricuspid atresia
Increase with cyanosis
indicates transposition
BP
Right arm and leg
Pulses
Right and left

brachial and
simultaneous
femoral and right
brachial
Thrill: gross insufficiency

of AV valve, severe
pulmonary stenosis,
absent PV
Pressure in lower
extremities is equal to or
minimally higher than
pressure in upper
extremities in the first
week
Pressures are preserved
by ductal flow in the
presence of severe leftsided obstructive disease
Norms vary by
gestational and
chronologic age and
method
Look for equality of
intensity and timing,
synchronicity, slope of
impulse curve, no delay
in peak between
preductal and postductal
pulses
Easily seen axillary pulses
suggest run-off or wide
pulse pressure
Pulse pressure
S1
Systolic minus
diastolic BP
Upper LSB
Lower LSB
S2
Upper LSB
25-30 cm of H2O in term;

15-25 cm of H2O in
preterm
Narrow indicates
myocardial failure,
vasoconstriction, and or
vascular collapse
Widened indicates AV
malformation, truncus
arteriosus, AP window,
and or PDA; may not be
widened until pulmonary
vascular resistance has
dropped
Usually single and
relatively accentuated;
audible split indicates
Ebstein anomaly or slow
heart rate; decreased
with CHF, prolonged AV
conduction
Increased accentuation
with increased flow across
AV valve indicates PDA,
MI, VSD, TAPVR, AVM,
and or tetralogy
Two components should
be heard by 6-12 h of age
Single sound indicates
aortic atresia, pulmonary
atresia, truncus
arteriosus, and or
transposition of great
arteries
Wide split indicates
pulmonary stenosis,
S3 and S4
Base or apex
Ebstein anomaly, TAPVR,

tetralogy, and or
occasionally left-to-right
atrial shunts
Loud sounds indicate
systemic or pulmonary
hypertension
S3 indicates increased
atrioventricular valve
flow, PDA, and or CHF
S4 indicates severe
Murmur
Precordium, back,
under both axilla
myocardial disease with

diminished LV compliance
Many serious cardiac
malformations do not
have their classic
murmurs in early
neonatal period but will
have some combination
of signs suggesting
pathology; absence of
murmur does not
preclude presence of
serious malformation (e.
g., transposition, TAPVR)
At least 60% of infants
have murmurs during the
first 48 h of life; PDA,
peripheral pulmonary
stenosis, and or tricuspid
regurgitation may be
indicated
Quiet is necessary to
auscultate murmurs; may
need to disconnect the
infant from the ventilator
for a few beats

Persistent murmurs first
heard right after birth
indicate ventricular
outflow obstruction; most
often pulmonic stenosis
Venous pulse
Jugular vein, liver
Jugular a and v waves in
sleeping infant
In presence of cyanosis,
pulsating liver suggests
RA or RV obstruction
Abdomen
Liver (left and right) Span greater than 5.5 cm
at term; late sign of CHF,
presence of left-sided or
central liver suggests
likely cardiac anomaly
Edema
Presacrum, eyelids, Causes are more often
legs and feet;
noncardiac except when
Chest: hydrops
associated with
abnormalities of renal
blood flow (e.g., leftsided obstruction of
severe hydrops
associated with
myocardiopathy, such as
severe anemia)
AP, aortopulmonary; AV, atrioventricular; AVM, arteriovenous
malformation; BP, blood pressure; bpm, beats per minute; CHF,
congestive heart failure; LA, left atrium; LSB, left sternal border;
LV, left ventricle; MI, mitral insufficiency; PDA, patent ductus
arteriosus; PMI, point of maximal impulse; PV, pulmonary valve;
RA, right atrium; RV, right ventricle; S1, first heart sound; S2,
second heart sound; S3, third heart sound; S4, fourth heart sound;
TAPVR, total anomalous pulmonary venous return; VSD, ventricular
septal defect.
From refs. 85 and 86.
Hand examination consists of observing its activity and appearance, including the nails, joints,
and palmar creases. The creases of the fifth digit should be parallel. If there is shortening of
the mid phalanx, the nonparallel creases mark a radial deviation, clinodactyly. Any curve less
than 10 to 15 degrees is normal. The thumb should reach just beyond the base of the index
finger. Extra digits that are postaxial or on the ulnar side are most often equivalent to skin
tags and of no significance; they may be familial, most often in families of color. Extra digits
on the preaxial or radial side are often enough associated with hematologic and cardiac
abnormalities to warrant further evaluation, regardless of racial background.
The neonate's hips require assessment with each visit because dislocations may not be
detectable on every examination. If the femur freely dislocates, it may appear to jerk
spontaneously when the infant extends or flexes his or her hip. The legs should be symmetric
in length on extension and with the knees flexed as the feet rest on the bed. If they are
unequal, suggesting dislocation of the shorter leg (i.e., Galeazzi sign), the next maneuver is
to attempt reduction on the shorter side while stabilizing the pelvis (i.e., Ortolani maneuver).
With the hip and knee flexed, the thigh is grasped with the third finger over the greater
trochanter and the thumb near the lesser trochanter. The other hand stabilizes the pelvis. As
the thigh is abducted, gentle pressure applied to the greater trochanter reduces the
dislocated femoral head into the acetabulum with a clunking sensation. The commonly felt,
benign clicks are distinct from the pathologic clunks, which often are seen as much as they
are felt when the femoral head jerks into place. If the legs are of equal length or if they rest
in full abduction, the first maneuver is to attempt to dislocate the head (i.e., Barlow
maneuver). With the hip and knee flexed, the thigh is grasped and adducted to 15 degrees
beyond midline while applying downward pressure. If the hip dislocates on the maneuver, the
Ortolani maneuver should then reduce it. If the hip rides to the edge but not out of the
acetabulum during the Barlow maneuver, it is subluxable. Even if dislocation is undetectable,
there may be telescoping with free movement of the femur up and down, indicating some
degree of instability. The Ortolani maneuver may be negative if a teratologic hip dysplasia
cannot be reduced. Unless both hips are involved, discrepancy
P.347
P.348
in leg length and inability to abduct fully to the affected side should be present.
TABLE 19-4 NEWBORN GENITALIA
Parameter
Penis
Testis
Anus
Location, male
Location, female
Size
Masculinization (i.e.,
labioscrotal fusion)
a
From refs. 87 and 88.
Normal Range
Abnormal Range
Length
Width
Volume
3.5 cm 1 cm
0.9-1.2 cm
1-2 cm
<2.5 cm
Anus to scrotum/
coccyx to scrotum
Anus to fourchette/
coccyx to fourchette
Diameter
Anus to fourchette/
anus to clitoris
0.58 0.06 cm
0.44 0.05 cm
7 mm +(1.3
weight in kg)
<0.46 cm
<0.34 cm
<0.5 cm
>0.5 cm
TABLE 19-5 NEONATAL NEUROLOGIC EVALUATION

Test
Resting posture
Technique
Observe unswaddled
Normal for Term
Moderate flexion of Constant tight flexion
infant without contact 4 limbs, held off

in quiet awake, quiet bed
active, or light sleep
states
Deep sleep
Light sleep
lower if head is in
midline
Elbows flexed with

dorsum of hands on
Extension of neck
in face
bed
Tight, persistent fisting
presentation or
ATNR persistent 30
seconds
Phasic (i.e.,
passive) tone:
resistance to
movement Tendon
reflexes
Strong lateral
preference
Moves from one to Is difficult to move

the other with
from one to the other
Awake, light
appropriate stimuli
peripheral movements Self calms
Awake, large
Modulated cry with
movements, not
expression
crying
Awake, crying
Motor activity
Full extension, flaccid

or forced
Equal side-to-side Knees abducted to

and upper-tobed (i.e., frog leg)
legs in breech
presentation
State
Deviant for Term
Observe throughout Appropriate for

physical examination state of alertness
Symmetric, fairly
smooth
Expressive face
with yawn or cry
Measure resistance to Response
extension (limb recoil) appropriate for
Scarf, heel to ear
gestational age
Test patellar reflex
Patellar reflex only
with head midline
one reliably
present at birth
Stays too alert or

cries without physical
reason
Does not come to fully
awake state
Weak or monotonous
cry
Bicycling, swatting
without stimulus
Asymmetric, weak
Jittery while sucking
Flat facial expression
Resists too much or
too little
Asymmetry
Sustained clonus
Postural (i.e.,
active) tone:
Pull to sitting while
resistance to gravity grasping infant's
Traction response hands
Infant pulls back

with flexion at
Asymmetry in pulling
back
elbows, knees, and No resistance

ankles
Full head lag
Head comes with Pull to stand instead
body with minimal Head does not fall
lag and falls
forward when
forward as infant goes

past upright
sitting is obtained
Upright
suspension
Suspend infant facing Infant supports

Infant falls through
examiner with both
himself then yields immediately
hands in axillae
slowly
Holds head erect,
Legs extend
Eyes fail to open
flexes hips, knees, Infant fails to relax

and ankles
and fall through after
Ventral
suspension
Positive support
Eyes open
Hold infant under
Flexes arms,
chest and suspend in extends neck,
prone position
holds back straight
Galant: stroke
Curves toward side
adjacent to spine
of stimulus
Landau: stroke
Extends back, lifts
caudocephalad along head and pelvis,
spine
micturates
Hold infant to support Infant extends
trunk with feet
hips to bear his or
touching firm, flat
her own weight
and relaxes after 1

min
Integrated reflexes Hold infant in supine Spreading: arms
Moro reflex
position; support
abduct, extend;
head and neck with
hands open
hand; allow head to
Hugging: arms
drop while still
adduct and flex;
supporting it
hands close
Tonic neck reflex Infant in supine,
Mental extension,
neutral position; turn occipital flexion
head to one side;
primarily of arms;
repeat opposite side; does not remain in
position for >30
seconds
1 min
Hangs limply or
excessively rigidly
Asymmetric incurving
Weak or absent
response
Infant fails to bear

weight or extends too
much or too long
surface
Unequal laterality
Absence of spread
Asymmetry
Exaggeration with
disorganization in state
Exaggerated response
and stays in position
>30 seconds
Withdrawal reflex Painful stimulus to

one foot
Withdrawal of
Absence of flexion in
stimulated foot;
stimulated leg
variable extension
of opposite leg
ATNR, asymmetric tonic neck reflex.

TABLE 19-6 ASSESSMENT OF CRANIAL NERVES
Cranial Nerve
II
III, IV, VI
Assessment
Pitfalls
Withdrawal or grimace to
Rarely tested clinically; rapid
strong odor (e.g., peppermint,

oil of cloves)
habituation
Behavioral response to light (i.

e., blinking fixing, follwing,
Room too bright; infant too

deeply asleep; overstimulation
turning to light source);

searching nystagmus
of other senses
Ocular movement, doll's eyes,

oculovestibular response, pupil
Should not force eyes to open;

ocular alignment usually poor in
size, gemini red reflex

VII
V, VII, XII
VIII auditory a
portion
IX, X
VII, IX
XII
neonates; light for pupil

response causes eye closing
Facial muscle tone at rest and Poor mouth opening due to
during crying
absence of depressor anguli oris
muscle
Sucking strength, rooting reflex Gestational age dependent;
infant should be hungry
Behavioral response to horn (i. Room too loud; difficult to
e., blinking widened eyes);
distinguish unilateral loss; rapid
quieting to voice
habituation
Swallowing with normal gag
Irritated throat after suctioning
Facial expression to strong
Rapid habituation
flavor
Tongue fasciculation, thrust,
Macroglossiaa
ability to shape around nipple
Nervous System
Neurologic evaluation begins with the initial observations made on approaching the infant and
continues as the infant is positioned and stimulated for the remainder of the routine physical
examination. As discussed earlier, assessing GA includes many of the steps also used to
evaluate motor tone and symmetry. Much can be learned about the neurologic state just by
observing what the infant does on his or her own; little more is needed unless the
observations indicate abnormality or there are particular risk factors. Subtle differences in
tone or use require more specific evaluation.
A neonate with facial asymmetry while crying and who has a flattened or absent nasolabial
fold has a facial palsy. These are most often acquired during forceps delivery and have an
incidence of 1.8 per 1,000 deliveries. More than 90% are expected to recover in the first few
years of life (114).
Infants display patterns of spontaneous muscular activity referred to as general movements.

In the first three months they are typically writhing of slow to moderate speed and small
amplitude (115). There appears to be a correlation between abnormal general movements
and later development of cerebral palsy (116,117).
These abnormalities are characterized as those confined to one or few body parts (poor
repertoire), involving simultaneous contraction of all limb and trunk muscles (crampedsynchronized), or are chaotic with large amplitude. The most worrisome are the crampedsynchronized as they have been most closely linked to the later development of cerebral
palsy (117).
The most common peripheral nerve injury associated with vaginal delivery is a brachial plexus
injury. Injuries may be incurred secondary to traction on an arm during birth. The upper
roots, C-5 and C-6, are most commonly damaged, leaving the infant with a prone adducted
arm at his or her side. It is generally advised that infants undergo operative repair if function
is not recovered by 3 months of age (114).
Fortunately quite rare, spinal cord injury initially may be confused with bilateral brachial
plexus injury if it occurs below C-5. Careful exam, however, will disclose paresis or paralysis
in the lower extremities as well. Lesions above this level are likely to incapacitate the
diaphragm and present with apnea or severe respiratory distress and may also be
complicated by cardiovascular instability (114).
Jitteriness, characterized by rhythmic tremors of equal amplitude around a fixed axis in an
extremity or the jaw, may occur in 41% to 44% of healthy newborns in the first hours of life
and thus deserves special comment (118). It occurs more often if the infant is awake, after a
startle, or after crying. Accompanying symptoms of CNS irritation, hypertonicity, and lowthreshold startle may also be present. Jitteriness is distinguishable from clonic seizure activity
because of its more rapid rate and its cessation either by pressure or by stimulating sucking;
it is most often a physiologic activity. If it fails to cease during sucking, it may be a sign of
hypoglycemia, hypocalcemia, or drug withdrawal (118). Jitteriness may persist in some
infants for many months with the infants having more severe accompanying CNS symptoms
having longer persistence (119).
The basics of the neonatal neurologic examination include assessment of state; spontaneous
muscle activity for assessing amount, quality, and strength; passive and active muscle tone;
and functioning of the cranial nerves. The steps are described in Tables 19-5 and 19-6.
P.349
REFERENCES
1. Silverman WA, Andersen DH. A controlled clinical trial of effects of water mist on
obstructive respiratory signs, death rate and necropsy findings among premature infants.
2. Morley CJ, Thornton AJ, Cole TJ, et al. Baby Check: a scoring system to grade the
severity of acute systemic illness in babies under 6 months old. Arch Dis Child 1991;66:100105.
3. Gray JE, Richardson DK, McCormick MC, et al. Neonatal therapeutic intervention scoring
system: a therapy-based severity-of-illness index. Pediatrics 1992;90:561-567.
4. Richardson DK, Gray JE, McCormick MC, et al. Score for Neonatal Acute Physiology: a
physiologic severity index for neonatal intensive care. Pediatrics 1993;91:617-623.
5. Richardson DK, Corcoran JD, Escobar GJ, et al. SNAP-II and SNAPPE-II: simplified
newborn illness severity and mortality risk scores. J Pediatr 2001;138:92-100.
6. American Academy of Pediatrics and the American College of Obstetricians and

Gynecologists. Guidelines for perinatal care, 5th ed. Elk Grove Village, Ill and Washington,
DC: AAP and ACOG, 2002.
7. Fletcher MA. Physical diagnosis in neonatology. Philadelphia: Lippincott-Raven

Publishers, 1998.
8. Souter VL, Kapur RP, Nyholt DR, et al. A report of dizygous monochorionic twins. New
Engl J Med 2003;349:154-158.
9. Saint-Anne-Dargassies S. Neurological development in the full-term and premature

neonate, 1st ed. Amsterdam: Elsevier, 1977.
10. Amiel-Tison C. Neurological evaluation of the maturity of newborn infants. Arch Dis
Child 1968;43:89-93.
11. Dubowitz LM, Dubowitz V, Goldberg C. Clinical assessment of gestational age in the
newborn infant. J Pediatr 1970;77:1-10.
12. Alexander GR, de Caunes F, Hulsey TC, et al. Validity of postnatal assessments of
gestational age: a comparison of the method of Ballard et al. and early ultrasonography.
13. Ballard JL, Novak KK, Driver M. A simplified score for assessment of fetal maturation of
newly born infants. J Pediatr 1979;95: 769-774.
14. Constantine NA, Kraemer HC, Kendall-Tackett KA, et al. Use of physical and neurologic
observations in assessment of GA in low birth weight infants. J Pediatr 1987;110:921-928.
15. Sanders M, Allen M, Alexander GR, et al. Gestational age assessment in preterm
neonates weighing less than 1500 grams. Pediatrics 1991;88:542-546.
16. Alexander GR, de Caunes F, Hulsey TC, et al. Validity of postnatal assessment of
gestational age: a comparison of the method of Ballard et al. and early ultrasonography.
17. Wariyar U, Tin W, Hey E. Gestational assessment assessed. Arch Dis Child 1997;77:
F216-F220.
18. Ballard JL, Khoury JC, Wedig K, et al. New Ballard score, expanded to include extremely
premature infants. J Pediatr 1991;119: 417-423.
19. Donovan EF, Tyson JE, Ehrenkranz RA, et al. Inaccuracy of Ballard scores before 28
weeks' gestation. J Pediatr 1998;135:147-152.
20. Hittner HM, Hirsch NJ, Rudolph AJ. Assessment of gestational age by examination of the
anterior vascular capsule of the lens. J Pediatr 1977;91:455-458.
21. Hittner HM, Gorman WA, Rudolph AJ. Examination of the anterior vascular capsule of
the lens: II. assessment of gestational age in infants small for gestational age. J Pediatr
Ophthalmol Strabismus 1981;18:52-54.
22. Lynn CJ, Saidi IS, Oelberg DG, et al. Gestational age correlates with skin reflectance in
newborn infants of 24-42 weeks gestation. Biol Neonate 1993;64:69-75.
23. Stevens-Simon C, Cullinan J, Stinson S, et al. Effects of race on the validity of clinical
estimates of gestational age. J Pediatr 1989;115:1000-1002.
24. Ellison P. The infant neurological examination. Adv Dev Behav Pediatr 1990;9:75.
25. Amiel-Tison C, Grenier A. Neurological assessment during the first year of life, 1st ed.
New York: Oxford University Press, 1986.
26. Litwin A, Aitkin I, Merlob P. Clitoral length assessment in newborn infants of 30 to 41

weeks gestational age. Eur J Obstet Gynecol Reprod Biol 1991;38:209-212.
27. Oberfield SE, Mondok A, Shahrivar F, et al. Clitoral size in full-term infants. Am J
Perinatol 1989;6:453-454.
28. Wilcox AJ, Russell IT. Birthweight and perinatal mortality: II. on weight-specific
mortality. Int J Epidemiol 1983;12:319-325.
29. Merlob P, Sivan Y, Reisner SH. Ratio of crown-rump distance to total length in preterm
and term infants. J Med Genet 1986;23: 338-340.
30. Sivan Y, Merlob P, Reisner SH. Upper limb standards in newborns. Am J Dis Child
1983;137:829-832.
31. Merlob P, Sivan Y, Reisner SH. Lower limb standard in newborns. Am J Dis Child
1984;138:140-142.
32. Amiel-Tison C, Gosselin J, Infante-Rivard C. Head growth and cranial assessment at

neurological examination in infancy. Dev Med Child Neurol 2002;44:643-648.
33. Raymond GV, Holmes LB. Head circumference standards in neonates. J Child Neurol
1994;9:63-66.
34. Sutter K, Engstrom JL, Johnson TS. Reliability of head circumference measurements in
preterm infants. Pediatr Nurs 1997;23: 485-490.
35. Sivan Y, Merlob P, Reisner SH. Head measurements in newborn infants. J Craniofac
Genet Dev Biol 1984;4:259-263.
36. Prechtl HF, Einspieler C, Cioni G, et al. An early marker for neurological deficits after
perinatal brain lesions. Lancet 1997;349: 1361-1363.
37. Prechtl H, Beintema D. The neurologic examination of the full-term newborn infant.
Clinics in developmental medicine, vol. 12. London: SIMP Heinemann, 1964.
38. Brazelton TB. Neonatal behavioral assessment scale, 2nd ed. Clinics in developmental
medicine, vol. 88. Philadelphia: JB Lippincott, 1984.
39. Lester BM, Boukydis CF, McGrath M, et al. Behavioral and psychophysiologic
assessment of the preterm infant. Clin Perinatol 1990;17:155-171.
40. Thoman EB. Sleeping and waking states in infants: a functional perspective. Neurosci
Biobehav Rev 1990;14:93-107.
41. Haddad GG, Jeng HJ, Lai TL, et al. Determination of sleep state in infants using
respiratory variability. Pediatr Res 1987; 21:556-562.
42. Sadeh A, Dark I, Vohr BR. Newborns' sleep-wake patterns: the role of maternal
delivery and infant factors. Early Hum Dev 1996; 44:113-126.
43. Hathorn MK. The rate and depth of breathing in new-born infants in different sleep
states. J Physiol 1974;243:101-113.
44. Stern E, Parmelee AH, Akiyama Y, et al. Sleep cycle characteristics in infants. Pediatrics
1969;43:65-70.
45. Poole SR. The infant with acute, unexplained, excessive crying. Pediatrics 1991;88:450455.
46. Heine RG, Jaquiery A, Lubitz L, et al. Role of gastro-oesophageal reflux in infant
irritability. Arch Dis Child 1995;73:121-125.
47. Als H, Lester BM, Tronick EC, et al. Manual for the assessment of preterm infants'
behavior (APIB). In: Fitzgerald HE, Lester BM, Yogman MW, eds. Theory and research in
behavioral pediatrics, vol. 1. New York: Plenum Press, 1982:65-132.
48. Grover G, Berkowitz CD, Lewis RJ, et al. The effects of bundling on infant temperature.
Pediatrics 1994;94:669-673.
49. Cheng TL, Partridge JC. Effect of bundling and high environmental temperature on
neonatal body temperature. Pediatrics 1993; 92:238-240.
50. Harpin VA, Chellappah G, Rutter N. Responses of the newborn infant to overheating.
Biol Neonate 1983;44:65-75.
51. Rutter N, Hull D. Response of term babies to a warm environment. Arch Dis Child
1979;54:178-183.
52. Harpin VA, Rutter N. Sweating in preterm babies. J Pediatr 1982;100:614-619.
53. Hegyi T, Carbone MT, Anwar M, et al. Blood pressure ranges in premature infants. I.
The first hours of life. J Pediatr 1994;124: 627-633.
54. Park MK, Lee DH. Normative arm and calf blood pressure values in the newborn.
Pediatrics 1989;83:240-243.
55. Perry EH, Bada HS, Ray JD, et al. Blood pressure increases, birth weight-dependent
stability boundary, and intraventricular hemorrhage. Pediatrics 1990;85:727-732.
56. Engle WD. Blood pressure in the very low birth weight neonate. Early Hum Dev
2001;62:97-130.
57. American Academy of Pediatrics Committee on Fetus and Newborn. Routine evaluation
of blood pressure, hematocrit, and glucose in newborns. Pediatrics 1993;92:474-476.
58. Goldring D, Wohltmann HJ. Flush method for blood pressure determinations in newborn
infants. J Pediatr 1952;40:285- 289.
P.350
59. Peitsch WK, Keefer CH, LaBrie RA, et al. Incidence of cranial asymmetry in healthy
newborns. Pediatrics 2002;110:e72.
60. Jones MD. Unilateral epicanthal fold: diagnostic significance. J Pediatr 1986;108:702704.
61. Jones KL, ed. Smith's recognizable patterns of human malformation, 4th ed.
Philadelphia: WB Saunders, 1988.
62. Samlaska CP, James WD, Sperling LC. Scalp whorls. J Am Acad Dermatol 1989;21:553556.
63. Smith DW, Greely MJ. Unruly scalp hair in infancy: its nature and relevance to problems
of brain morphogenesis. Pediatrics 1978; 61:783-785.
64. Drolet BA, Clowrey L Jr, McTigue MK, et al. The hair collar sign; marker for cranial
dysraphism. Pediatrics 1995;96: 309-313.
65. Corona-Rivera JR, Corona-Rivera E, Romero-Velarde E, et al. Report and review of the
fetal brain disruption sequence. Eur J Pediatr 2001;160:664-667.
66. Johnsonbaugh RE, Bryan RN, Hierlwimmer R, et al. Premature craniosynostosis: a

common complication of juvenile thyrotoxicosis. J Pediatr 1978:93:188-191.
67. Shuper A, Merlob P, Grunebaum M, et al. The incidence of isolated craniosynostosis in

the newborn infant. Am J Dis Child 1985;139:85-86.
68. Graham JM Jr, Smith DW. Parietal craniotabes in the neonate: its origin and
significance. J Pediatr 1979;95:114-116.
69. Popich GA, Smith DW. Fontanels: range of normal size. J Pediatr 1972;80:749-752.
70. Faix RG. Fontanelle size in black and white term newborn infants. J Pediatr
1982;100:304-306.
71. Duc G, Largo RH. Anterior fontanel: size and closure in term and preterm infants.
Pediatrics 1986;78:904-908.
72. Adeyemo AA, Omotade OO. Variations in fontanelle size with gestational age. Early
Hum Dev 1999;54:207-214.
73. Lloyd FA, Finkelstein SI. Normal head growth in infant with nonidentifiable anterior
fontanel. J Pediatr 1975;87:490-494.
74. Kataria S, Frutiger AD, Lanford B, et al. Anterior fontanelle closure in healthy term
infants. Infant Behav Dev 1988;11:229.
75. Potter EL, Craig JM. Pathology of the fetus and the infant, 3rd ed. Chicago: Year Book
Medical Publishers, 1975.
76. Benaron D. Subgaleal hematoma causing hypovolemic shock during delivery after failed
vacuum extraction: a case report. J Perinatol 1993;13:228-231.
77. Govaert P, Vanhaesebrouck P, De Praeter C, et al. Vacuum extraction, bone injury and
neonatal subgaleal bleeding. Eur J Pediatr 1992;151:532-535.
78. Gundry SR, Wesley JR, Klein MD, et al. Cervical teratomas in the newborn. J Pediatr
Surg 1983;18:382-386.
79. Carr MM, Thorner P, Phillips JH. Congenital teratomas of the head and neck. J
Otolaryngol 1997;26:246-252.
80. Sivan Y, Merlob P, Reisner H. Eye measurements in preterm and term newborn infants.
J Craniofac Genet Dev Biol 1982;2:239- 242.
81. Crouch ER Jr, Crouch ER. Pediatric vision screening: why? when? how? Contemp Pediatr
1991;8:9-30.
82. Isenberg SJ, Vazquez M. Are the pupils of premature infants affected by intraventricular
hemorrhage? J Child Neurol 1994;9: 440-442.
83. Isenberg SJ. Clinical application of the pupil examination in neonates. J Pediatr
1991;118:650-652.
84. Roarty JD, Keltner JL. Normal pupil size and anisocoria in newborn infants. Arch
Ophthalmol 1990;108:94-95.
85. Goldbloom RB, ed. Pediatric clinical skills, 2nd ed. New York: Churchill Livingstone,
1997.
86. Smith DW, Takashima H. Ear muscles and ear form. Birth Defects Orig Artic Ser
1980;16:299-302.
87. Zerin M, Van Allen MI, Smith DW. Intrinsic auricular muscles and auricular form.
88. Ruder RO, Graham JM Jr. Evaluation and treatment of the deformed and malformed
auricle. Clin Pediatr (Phila) 1996;35: 461-465.
89. Sivan Y, Merlob P, Reisner SH. Assessment of ear length and low set ears in newborn
infants. J Med Genet 1983;20:213-215.
90. Silverman SH, Leibow SG. Dislocation of the triangular cartilage of the nasal septum. J
Pediatr 1975;87:456-458.
91. MacEwen CJ, Young JD. Epiphora during the first year of life. Eye 1991;5:596-600.
92. Ogawa GS, Gonnering RS. Congenital nasolacrimal duct obstruction. J Pediatr
1991;119:12-17.
93. Levin SE, Silverman NH, Milner S. Hypoplasia or absence of the depressor anguli oris
muscle and congenital abnormalities, with special reference to the cardiofacial syndrome. S
Afr Med J 1982;61:227-231.
94. Miller M, Hall JG. Familial asymmetric crying facies. Its occurrence secondary to
hypoplasia of the anguli oris depressor muscles. Am J Dis Child 1979;133:743-746.
95. Richard BM, Qiu CX, Ferguson MW. Neonatal palatal cysts and their morphology in cleft
lip and palate. Br J Plast Surg 2000;53:555-558.
96. Hayes PA. Hamartomas, eruption cyst, natal tooth and Epstein pearls in a newborn.
ASDC J Dent Child 2000;67:365-368.
97. Zhu J, King D. Natal and neonatal teeth. ASDC J Dent Child 1995;62:123-128.
98. Ballard JL, Auer CE, Khoury JC. Ankyloglossia: assessment, incidence, and effect of
frenuloplasty on the breastfeeding dyad. Pediatrics 2002;110:e63.
99. Moyer VA, Ahn C, Sneed S. Accuracy of clinical judgment in neonatal jaundice. Arch
Pediatr Adolesc Med 2000;154:391-394.
100. Bamji M, Stone RK, Kaul A, et al. Palpable lymph nodes in healthy newborns and
101. Hilliard RI, McKendry JB, Phillips MJ. Congenital abnormalities of the lymphatic
system: a new clinical classification. Pediatrics 1990;86:988-994.
102. Nichols MM. Shifting umbilicus in neonatal phrenic palsy (the belly dancer's sign). Clin
Pediatr (Phila) 1976;15:342-343.
103. Joseph PR, Rosenfeld W. Clavicular fractures in neonates. Am J Dis Child

1990;144:165-167.
104. Hassan A, Karna P, Dolanski EA. Intermamillary indices in premature infants. Am J

Perinatol 1988;5:54-56.
105. Grotto I, Browner-Elhanan K, Mimouni D, et al. Occurrence of supernumerary nipples

in children with kidney and urinary tract malformations. Pediatr Dermatol 2001;18:291-294.
106. Senquiz AL. Use of decubitus position for finding the olive of pyloric stenosis.
107. Sinkin RA, Phillips BL, Adelman RD. Elevation in systemic blood pressure in the
neonate during abdominal examination. Pediatrics 1985;76:970-972.
108. Ashkenazi S, Mimouni F, Merlob P, et al. Size of liver edge in full-term, healthy
infants. Am J Dis Child 1984;138:377-378.
109. Reiff MI, Osborn LM. Clinical estimation of liver size in newborn infants. Pediatrics
1983;71:46-48.
110. Brion L, Avni FA. Clinical estimation of liver size in newborn infants. Pediatrics
1985;75:127-128.
111. Du ZD, Roquin N, Barak M. Clinical and echocardiographic evaluation of neonates with
heart murmurs. Acta Paediatr 1997;86: 752-756.
112. Phillip M, De Boer C, Pilpel D, et al. Clitoral and penile sizes of full term newborns in
two different ethnic groups. J Pediatr Endocrinol Metab 1996;9:175-179.
113. Berenson A, Heger A, Andrews S. Appearance of the hymen in newborns. Pediatrics

1991;87:458-465.
114. Medlock MD, Hanigan WC. Neurologic birth trauma. Intracranial, spinal cord, and
brachial plexus injury. Clin Perinatol 1997;24: 845-857.
115. Einspieler C, Prechtl HF, Ferrari F. The qualitative assessment of general movements
in the preterm, term and young infantsreview of the methodology. Early Hum Dev
1997;50:47-60.
116. Prechtl HF, Einspieler C, Cioni G, et al. An early marker for neurological deficits after
perinatal brain lesions. Lancet 1997;349: 1361-1363.
117. Ferrari F, Cioni G, Einspieler C, et al. Cramped synchronized general movements in

preterm infants as an early marker for cerebral palsy. Arch Pediatr Adolesc Med
2002;156:460-470.
118. Linder N, Moser AM, Asli I, et al. Suckling stimulation test for neonatal tremor. Arch
Dis Child 1989;64:44-46.
119. Kramer U, Nevo Y, Harel S. Jittery babies: a short-term follow-up. Brain and Dev
1994;16:112-114.
120. Jorgenson RJ, Shapiro SD, Salinas CF, Levin LS. Intraoral findings and anomalies in
neonates. Pediatrics 1982;69:577.
121. Levin LS, Jorgenson RJ, Jarvey BA. Lymphangiomas of the alveolar ridge in neonates.
122. Fromm A. Epstein's Pearls, Bohn's nodules and inclusion-cysts of the oral cavity. J
Dent Child 1967;34:275.
123. King NM, Lee AMP. Prematurely erupted teeth in newborn infants. J Pediatr
1989;114:807.
124. Leung AKC. Natal teeth. Am J Dis Child 1986;140:249.
Chapter 20
General Care
Ian Laing
This chapter describes the general care of the well infant and discusses how a
pediatrician may meet the needs of the majority of families whose infants do not
require admission to a neonatal unit.
No pediatrician should act independently of midwives, neonatal nurses,
obstetricians, fetal medicine specialists, radiologists, geneticists, and practitioners
of many other disciplines. The pediatrician should be a member of a
multidisciplinary group that meets weekly to discuss the anticipated antenatal
problems and to provide an audit of outcomes from the previous months.
There are four phases of care:
Antenatal
Intrapartum
Neonatal
Postdischarge
ANTENATAL
The past 20 years have seen a revolution in the general public's expectations
regarding the standard of care received and the quality and extent of information
provided. Midwives and obstetricians are not the only professionals responsible for
providing this care and information. The pediatrician must work with these other
professionals to ensure the clarity and accuracy of information provided
antenatally to mothers and their partners. The team must also grapple with the
legal and ethical aspects of informed consent. Wherever possible, all
professionals should strive toward evidence-based medicine. The concept of risk is
not yet well understood by the general population.
Information
The developed world has now emerged from the days when underinformed
patients conveyed their decisions to professionals. The patient-professional
partnership is now crucial. Information should be given to mother and partner as
fully as possible, in verbal and written form. Such information should maintain
perspective, while being spoken or written in language that the patient can
understand.
Risk
Today our communities strive toward natural childbirth, emphasizing that birth
should ideally be a normal physiologic phenomenon for a healthy woman and
fetus. Nevertheless, even the healthiest pregnancy carries a small risk of maternal
mortality and a significant risk of morbidity. For the fetus and neonate the risks
are much higher: The perinatal mortality is usually between 6 and 10 deaths per
1,000 total deliveries in the developed world. The anticipated risk may change.
New possibilities may emerge unexpectedly. Unforeseen emergencies occur
without time to discuss them completely. Further-more, some pregnancies may
appear to cause initial concern, and such worries may resolve as the pregnancy
progresses. The clinician must address the question of whether the mother should
be told of all possibilities, even those that may occur once in 500 events. In
striving to do so, there is a chance that all perspective may be lost for the patient.
A value judgment must be taken in each individual case. This risk, however great
or small, is communicated to mother and a plan is chosen accordingly. Written
information should capture the idea that the plan may have to change in light of
forthcoming events.
Because labor is usually a tiring process for the mother, it is not an ideal time to
discuss options that could be better addressed beforehand. Mode of feeding,
vitamin K prophylaxis, and immunization schedules are examples of subjects that
should be explored in depth before the mother enters labor.
Informed Consent
In theory, there should be fully informed consent at all times. This is an
important goal, but it is not entirely achievable. Pregnancy, labor, delivery, and
the neonatal period are all times when unexpected events may arise. Anticipating
them all is impossible and would be detrimental to the perspective that a woman
and her partner require. Fully informed consent can be aimed at exploring all
likely events and their outcomes, test findings, including the implications of falsepositive and false-negative results, and
P.352
the concept of changing risk. These goals are commendable, very timeconsuming, and potentially bewildering, even for educated parents. Compromises
therefore must be embraced. Professionals should obtain written informed consent

where possible, and those who administer, manage, and finance the health
program must recognize that the informed consent procedure is time-consuming
and requires increased staff resources.
Infant Feeding Information

There is great variation throughout the world in rates of breast-feeding. This
variation is largely cultural in origin. Some parts of the developed world, including
Europe, the United States, and Canada, have low rates of breast-feeding and
physicians are required to provide mothers with clear information about the
proven advantages of breast-feeding. Advantages include lower prevalence of
maternal breast cancer, lower incidence of infant infections (1) (including
necrotizing enterocolitis), lower prevalence of obesity in childhood (2), and lower
incidence of sudden infant death syndrome (SIDS) (3,4). Breast-feeding may also
provide some protection against allergic disease in childhood.
Vitamin K Information
Hemorrhagic disease of the newborn (HDN) can cause significant neonatal
morbidity and mortality. It may present in the early days of neonatal life as
bleeding from the umbilicus, skin, or as intestinal hemorrhage. Late-onset HDN
may cause intracranial bleeding, with consequent death or brain damage. Infants
who are breast-feeding or born of mothers taking anticonvulsants are at particular
risk (5). Because of these hazards, all parents should be advised to choose
vitamin K prophylaxis for their newborn infants. One intramuscular dose of
vitamin K can prevent almost all episodes of HDN. This is a simple and reliable
method of prophylaxis. Some reports suggest that there is an increased incidence
of childhood malignancy after intramuscular vitamin K prophylaxis, whereas other
studies fail to confirm this.
Oral prophylaxis may be given on days 1, 8, and 28, and is effective at
dramatically reducing the incidence of HDN (6). Oral prophylaxis requires a
motivated parent to remember to administer the second and third dose, or else an
efficient community support program must be in place to ensure that the infant is
similarly protected.
TABLE 20-1 IMMUNIZATION SCHEDULE
Age (Months After Delivery)
At birth
2
3
Immunization
BCG (high risk) Hep B (mother carrier)

Dip Tet Pertuss Polio HIB MenC
4
BCG, Bacille Calmette-Gurin (tuberculosis); Dip, diphtheria; HIB,
Haemophilus influenzae B; MenC, meningococcal C; Pertuss, pertussis;
Tet, tetanus.
Written information should be available for all parents to examine the evidence,
and professionals should be available to discuss the issues and provide parents
with a recommendation. Anxieties expressed in the literature, whether justified or
not, should not lead parents to reject vitamin K prophylaxis altogether. HDN is a
potentially lethal condition.
Immunization Information
Throughout the world there are differences in immunization programs. These
depend on varying risks of disease in different parts of the planet, availability of
vaccines and ability to refrigerate them, and community education programs.
Preterm infants should be immunized at the same postnatal age as their term
counterparts. Provided the infant is clinically well, the timing of immunization
should be unaffected by gestation. Table 20-1 shows an example of recommended
immunization schedules in the developed world. Because each country has its own
schedule, the pediatrician should check both the local and current
recommendations.
INTRAPARTUM
Ideally the pediatrician attending the delivery of a newborn infant has full
information available. The reality is often different. An urgent call to a labor suite
is followed gradually by an unfolding history. The pediatrician relies on close
observation of the neonate to guide emergency care while the history is gathered.
Information from case records should include the following:
Maternal health: Diseases prior to and during the pregnancy are

documented; these include diabetes mellitus and positive human
immunodeficiency virus (HIV) status. In addition, there should be an

account of diseases associated with the pregnancy, such as pregnancyinduced hypertension and gestational diabetes.
Prescribed drugs during pregnancy: Some therapeutic drugs may have

significant effects on the fetus and neonate (see Chapter 15).
P.353
Drug abuse: A mother may be a chronic drug abuser. Frequently this

involves a cocktail of drugs, only some of which are known by professionals.
The chronic drug abuser may be a caring mother with feelings of guilt and
sincere intentions of creating a healthy environment for the new baby. The
mother however, may have a chaotic lifestyle, and there might also be
coincident malnutrition, heavy smoking, alcoholism, and positive HIV status.
Fetal growth and apparent well-being, including all tests of fetal

assessment: These may involve fetal ultrasonography scans for structural
anomalies and nuchal translucency. Data from amniocentesis or chorionic
villus sampling may be available. Doppler studies of uterine arteries may
have been used to predict preeclampsia and fetal growth retardation.
Umbilical arterial Doppler studies may have predicted placental damage and
vascular occlusion. Doppler studies of fetal arteries and veins can also be
used to assess fetal hypoxemia. Biophysical profile scoring involves
cardiotocography, amniotic fluid volume, and three dynamic
ultrasonography variables (fetal movement, fetal breathing, and fetal tone).
Chapter 12 includes a detailed discussion of these techniques.
Gestation: It is important to know the best estimate of gestation and how

this was calculated. An ultrasonography examination in the first trimester is
accurate to within 2 to 3.5 days. The mother's history of the first day of
the last menstrual period may be helpful but is not always available.
Remember too that a mother may experience a menstrual period even after
conception has taken place, and in these circumstances an early
ultrasonography scan may correctly declare that the pregnancy is 4 weeks
further advanced than the mother's history would suggest.
Date and time of rupture of membranes: The definition of prolonged rupture

of the amniotic membranes is not consistent and is variously described as
12, 18, 24, 48, and 72 hours. It is associated with chorioamnionitis and an
increased risk of fetal and neonatal infection.
Events during labor: Assessment of fetal well-being including

cardiotocography, passage of meconium, maternal pyrexia, and the results
of any fetal scalp pH measurements may guide the resuscitation. The
combination of profound fetal acidosis and neonatal apnea requires the
presence of an experienced pediatric team that can begin immediate
assisted ventilation.
Mode of delivery: Resuscitation of the infant in the room where mother has
labored allows the pediatrician to learn firsthand about the ease or difficulty
of delivery. Forceps or ventouse (vacuum) deliveries may account for local
trauma observed.
NEONATAL
Chapter 18 describes resuscitation of the neonate in detail.
The transformation from fetus to neonate is a remarkable one. The placenta,
which until this point has been the provider of food and oxygen and the remover
of fetal waste products, is clamped off and the neonate must immediately adapt to
take responsibility for these functions. Collapsed, liquid-filled lungs become
inflated within seconds, and the capillaries and lymphatics drain most of the
pulmonary fluid in a few hours. The fetal partial pressure of arterial oxygen (PaO2)
changes from 32 mm Hg to 80 mm Hg in minutes and achieves 100 mm Hg in a
few days. Approximately 95% of infants establish spontaneous respiration by 1
minute after delivery and thereafter become pink and vigorous spontaneously.
The purpose of resuscitation is to intervene when these natural processes are
disturbed pathologically. Dr. Virginia Apgar's scoring system for assessment of the
neonate at 1 and 5 minutes is still used widely today, but the pediatrician should
concentrate on assisting the infant to achieve a heart rate greater than 100 per
minute and active, crying respirations. In the absence of congenital abnormalities,
central pinkness is soon achieved thereafter.
General Principles of Resuscitation

The following principles are guides to the general care of the newborn in the early
minutes of life:
Resuscitation in the delivery room can be readily described to parents while

it is happening. Parents fear silences.
Identification anklets or wristbands are important, but not at the expense of
causing a tiny infant to become cold.
The loudly crying baby never needs to be intubated or to be given bag-andmask insufflation. Enrichment of the inspired gas may be required but never
under positive pressure. The infant is more efficient than the clinician.
Do not ever sound frustrated or critical of colleagues during resuscitations.

Such sentiments result in a breakdown of essential teamwork and sap the
confidence of parents.
Meconium
The contents of the fetal bowel may have been passed prematurely and may have
been inhaled during asphyxial gasping prior to delivery. Meconium is a chemical
irritant to the lungs as well as a marker of fetal distress. In 9% to 11% of
deliveries, meconium is present. The caregiver should thoroughly suction any
meconium from the mouth and nares with the head on the perineum. Whether or
not meconium should be aspirated from the newborn's oropharynx is controversial
(7); the current author is in favor of this practice, but recognizes that more data
are required to support or refute this maneuver. It is clear that those infants who
are vigorous do not require invasive care. Those who are in poor condition should
be intubated and any inhaled meconium aspirated from the trachea.
Cord Clamping
Data are currently being gathered about the ideal time of cord clamping and
whether the infant should be held for a
P.354
few seconds below the placenta in order to receive a gravitational transfusion (8).
Consequences of receiving too much blood volume are fluid overload,
polycythemia, and (later) hyperbilirubinemia. Until the data from research studies
are complete, it may be reasonable to allow a 15-second delay to achieve a
modest blood transfusion. Because delayed clamping of the cord increases the
infant's blood volume, it may be advantageous in extreme preterm infants to
ensure an adequate intravascular volume, and perhaps to decrease the necessity
for subsequent blood transfusion. There is no place for stripping of the cord by
milking the blood from placenta to baby. The cord should initially be clamped
several centimeters from the umbilicus and then be cut (9). Then the cord should
be clamped 1 or 2 cm from the umbilicus using a disposable clamp, and cut with
sterile scissors distal to the clamp. The cord clamp can be safely removed during
the second day of life, at which time the cord should be inspected to insure that
there is no residual hemorrhage.
Drying
The neonate is delivered covered in amniotic fluid that immediately extracts latent
heat by evaporation from the infant's body. Consequently, most infants should be
thoroughly dried with warm towels.
Temperature Control
Newborn infants lose heat by four physical mechanisms: evaporation, conduction,
convection, and radiation. Evaporation is at its most important in the early
seconds of life when the newborn is covered in amniotic fluid. Conduction in
general contributes little to heat loss because the infant is in contact with warm
garments of low conductivity. Convection losses become important when a child is
exposed to drafts, and most especially when a child is being nursed naked on an
open radiant warmer. Radiant heat loss occurs to cold surrounding objects, that
is, the incubator wall in a cool room. Recognition of these physical principles
allows heat loss to be minimized by early drying, warm clothing, freedom from
drafts, and ensuring that incubators are double-walled with warm gas between the
walls. Physical examination, weighing, and bathing of the infant should always be
carried out in a warm environment. The neonatal unit and postnatal wards should
audit the incidence of neonatal hypothermia and identify the sources of baby
cooling. The areas or procedures at fault should be corrected and the quality of
care surveyed prospectively.
Identification
Before leaving the delivery room, every infant should be identified by the fixing of
a wristband and anklet; the parent(s) should witness the bands being attached. In
addition, many institutions now routinely do footprinting, handprinting, and
fingerprinting, although it is not yet clear that this is reliable and worthwhile (10).
Security
Because of sporadic baby abductions in maternity units, each unit should have a
written policy to protect the infants. Both antenatally and postnatally, parents
should be given instructions to maximize the safety of the infant. Parents and staff
must verify the identity of anyone requesting to remove the baby from the room.
Parents should be encouraged, when in any doubt, to summon another member of
the staff to check the identification of the person asking to take the child.
Bonding
The early minutes and hours of a child's life are important times for establishing a
close bond between mother and infant. After the majority of deliveries it should be
possible to put the infant immediately to the maternal breast for close physical
contact even if the mother does not intend to breast-feed.
Control of Infection
Even today infection is one of the major causes of mortality and morbidity in
newborn infants. Gowning does not decrease bacterial colonization of the baby nor
the incidence of neonatal sepsis (11). The maternity service, in all its
departments, must have high-quality handwashing facilities. The troughs should
be large with elbow-operated, knee-operated, or automatic taps that produce
warm water. Soap must be plentiful. Alcohol rubs should be readily available
wherever infants are cared for. Staff should be free of all clothes and jewelry from
elbows to fingertips. Each institution should have regular (at least annual)
education sessions for staff, and a culture of nonthreatening criticism by peers
should be ever-present. All staff on entering a nursery must wash their hands and
forearms with an antiseptic (e.g., chlorhexidine or hexachlorophene). Before and
after touching each baby, staff should douse hands and forearms in at least 5 mL
of alcohol rub or thoroughly wash their hands (12).
Eye Prophylaxis
Throughout the developed world there is great variation in the use of eye
prophylaxis. Each individual community should make a risk assessment depending
on the incidence of identified Chlamydia and gonococcal infections. Erythromycin
0.5% and tetracycline 1% are effective antibiotics against sensitive gonococci, but
penicillinase-producing gonococci are increasing in numbers and are best
prevented with silver nitrate 1% (13).
If prophylaxis is to be given, the ointments should be instilled within the first hour
of life into the lower conjunctival sac. The lids are then gently massaged. It is
difficult to produce effective prophylaxis against Chlamydia. If conjunctivitis is
identified, scrapings of the tarsal conjunctiva should be sent to the laboratory to
identify typical cytoplasmic inclusion bodies. Treatment with combined
P.355
oral and topical erythromycin has the advantage of eradication of nasopharyngeal
carriage of Chlamydia. Equally, if the mother is known to have gonococcal
disease, prophylaxis with ointments is insufficient for the infant, who should be
fully treated with parenteral antibiotics (13).
General Concepts of Clinical Examination

Chapter 19 describes the details of clinical examination. The purpose of newborn
clinical examination is to identify the child who is unwell and the child who has an
evident congenital abnormality. Reassurance is given, along with an opportunity
for parents to ask questions of the examining doctor. There is no universally
accepted ideal time for carrying out the examination. The pediatrician cannot
exclude all congenital abnormalities but rather is highlighting those that can be
identified at that time. On day 1 the systolic murmur of ventricular septal defect is
not yet manifest. On day 4 the infant may seem well yet may perish on day 5
from aortic atresia as the ductus arteriosus closes. Congenital metabolic diseases
may occasionally take weeks, months, or even years to be symptomatic. It should
be made clear in writing to parents that the examination is a description of the
child's condition on a particular day and time, and that a normal examination does
not guarantee that the neonate is perfect in every respect.
Birth weight and head circumference should be accurately measured. This is
documented as a baseline for any further consultations, which may arise, and also
to contribute to epidemiological data for the population.
Screening Tests
The developed world has the opportunity to use bloodspot screening for conditions
not readily identifiable by other means. These diseases screened for vary from
country to country and even from region to region. In the United Kingdom, the
national universal newborn screening program for phenylketonuria was introduced
in 1969, and the program for congenital hypothyroidism was introduced in 1981.
Galactosemia, tyrosinemia, and maple syrup urine disease have been screened for
in the past but have been discontinued. Cystic fibrosis and anonymous testing for
HIV-positive blood are screened for in some regions, and there are plans to screen
for sickle cell disorders in the future.
The screening may be done in the hospital, or the midwife or public health nurse
may do it during a home visit. The baby's heel is pricked, and four bloodspots
are collected on a filter paper (the Guthrie card). Each country must decide
whether they have the time to obtain fully informed written consent from a parent
in every case. Some authorities recognize that making written consent compulsory
may have the undesirable side effect of reducing uptake of the screening
program, with consequent detriment to the individual children with undetected
disease. Written information should be provided to parents both antenatally and
postnatally. There should also be high-quality verbal communication with all
mothers on the subject of newborn bloodspot screening. Pediatricians caring for
newborns must be acquainted with their local newborn screening program, not
only for the diseases being screened but also for the timing and implementation of
the screening and the procedure to be followed for questionable and abnormal
results.
Fluids and Growth

Of all infants, 95% are observed to pass urine in the first 24 hours of life (14).
The 5% who have dry diapers have probably voided during labor or in the labor
suite. Failure to pass urine in 48 hours should trigger careful examination and
investigation.
Meconium is usually passed in the first day of life. Failure to do so should prompt
the clinician to ask whether meconium was passed prior to or during labor. If
failure to pass meconium is confirmed, careful examination should exclude
anorectal atresia. Abdominal examination should ensure that there is no evidence
of distention. Hirschsprung's disease and meconium ileus may both present with
delayed passage of meconium, and the child should be evaluated for these
conditions. The infant should be observed until there is reliable documentation of
the passage of meconium (14).
Meconium is a dark green viscous substance consisting of bile-stained intestinal
secretions, intestinal debris, and cells. It is usually observed more than once per
day for the first 3 days of life. After approximately 3 days of milk feedings, the
meconium appears mixed with milk stool (changing stool) and in 2 more days
there are well-established yellow milk stools. Delay in appearance of milk stools
should trigger the clinician to question the adequacy of oral feedings. Passage of
stool in the neonatal period varies in frequency and consistency. Breast-fed
infants typically pass stool several times daily, and the laxity is occasionally
confused with diarrhea. The stool frequency of formula-fed infants is usually less,
and the stool is of a firmer consistency, occasionally confused with constipation
(Table 20-2).
Breast-feeding
Breast-feeding should be actively promoted in all societies. Whenever possible the
infant is put directly to the breast at delivery. This maintains baby's temperature,
promotes bonding, and allows early suckling, which in turn stimulates prolactin
production by the anterior pituitary, resulting in milk production in the maternal
breast. Nipple stimulation also promotes oxytocin release by the posterior
pituitary gland. Oxytocin leads to contraction of the myoepithelial cells in the
breast, and this causes the lactiferous sinuses to release milk (the let-down
reflex). For the first 3 days mother produces colostrum, which is high in protein,
immunoglobulins, hormones, and white cells. These probably have a role in
protecting the infant against infection. Initially the volumes of fluid produced are
low,
P.356
and the infant must break down stored glycogen to maintain blood glucose
concentrations.
TABLE 20-2 FREQUENCY OF VOIDING AND STOOLING IN THE EARLY

DAYS OF LIFE
Days 1-2
Passage of urine 2 or more
Days 3-4
Day 5 Onward
3 or more
5 or more heavy
Passage of stool 1 or more meconium 2 or more
diapers
2 or more yellow
changing stool,
brownish green
After the first 3 days, mother reports that her milk has come in. The clear
colostrum is now replaced by creamy-colored breast milk, rich in fat. The exact
composition of breast milk varies from mother to mother, and the hind-milk, that
which is produced later in the feed, is richer in fat and calories. Once the mother
and infant have established satisfactory breast-feeding, almost all of the milk
taken is in the first 10 minutes; any suckling thereafter is for social purposes.
Suckling for too long, such as greater than 20 minutes, might exhaust both
mother and baby and can result in trauma to the nipple with consequent
interference with pleasurable breast-feeding.
Pediatricians tend to delegate the instruction and encouragement of breastfeeding to midwives and nurses. Although it is highly probable that these
professionals give the best advice to mothers, pediatricians should possess a
working knowledge of the mechanics of breast-feeding and the most common
associated problems. Infants who fail to breast-feed may develop hypernatremic
dehydration, a condition that can be fatal or cause major morbidity. For a detailed
review of this lethal and underrecognized condition, see reference 15.
All staff should be familiar with the following principles of breast-feeding:
Mother and baby should be relaxed and comfortable.
No artificial teats (soothers, dummies) should be used.
No artificial milks should be given unless medically indicated.
Babies are commonly sleepy in the first 24 hours but may demand feeds
every 2 to 3 hours thereafter.
After the full fatty milk has been established on day 3 or 4, infants settle
with demand feeds every 3 to 4 hours.
Babies should suckle often, especially in the early days of life.
Successful breast-feeding involves a wide-mouthed approach of the baby,

the tongue underneath the nipple, the chin pressed into the breast tissue,
and the nose high.
Sucking should be powerful and rhythmic with pause periods for breath
while the child remains latched on.
If mothers are separated from their infants out of necessity, they should be
shown how to manually express breast milk manually.
Posseting and small vomits are common, but all bile-stained vomiting
should be considered pathologic until proven otherwise.
Term babies lose up to 10% of their birth weight and then should show a
growth velocity thereafter of approximately 150 to 200 g per week. If there
is any doubt about the adequacy of breast-feeding, accurate weighing can
be a simple and helpful intervention.
There is no place for test weighing, a procedure whereby a baby is

weighed before and after a feed.
Despite these important written recommendations, there is no substitute for an

experienced breast-feeding counselor sitting with mother during the early feeds,
optimizing mother's position, baby's position, observing latching onto the breast,
ensuring that the tongue is able to milk the ducts, and noting that milk supply is
adequate and that the infant successfully receives the nutrition required. Every
neonatal unit and postnatal ward should possess a copy of the WHO/UNICEF
recommendations on The 10 Steps to Successful Breast-feeding (16).
Bottle-feeding
If a mother has made an informed decision to bottle-feed her infant, there is no

reason why a small formula feed should not be offered in the early minutes or
hours of life. Some professionals prefer to offer a water feed first (17), but in
healthy newborn infants with no history of polyhydramnios, it is not clear that this
is required. Glucose feeds should not be offered, on the grounds that inhalation of
glucose can cause a severe chemical pneumonitis. Newborns may take 15 to 30
mL of formula every 3 to 4 hours in the first day, rising to 90 mL feeds 4 to 5
times a day by day 5 (18). A healthy infant can establish demand feeds in the
early days of life.
Poor Feeding
Poor breast-feeding is common. Poor bottle-feeding is not. The infant who is
thought to be feeding poorly at the breast should be seen by an experienced
professional to ensure that there is no underlying serious structural or systemic
problem. Never forget infection. If the child is thought to be well, then the most
important therapeutic measure is the prolonged patient attentions of a member of
staff who is an expert on teaching breast-feeding.
P.357
Primigravid mothers often experience difficulties in achieving good breast-feeding,
but even mothers who have breast-fed before may require further encouragement
and assistance. They might say, I have breast-fed other babies before, but I
have never breast-fed that baby.
Poor bottle-feeding is much simpler to approach. The milk is available and the
intake is readily measured. Again, however, structural and systemic problems
should be considered, because the baby with meningitis may feed neither at the
breast nor at the bottle. If there is thought to be no underlying pathology, the
technique of feeding should be observed. Is the teat (nipple) adequate to allow
milk flow? How is the mother holding the infant and bottle? Is the child lethargic,
indicating some disease process? Is the child irritable and frustrated, indicating
some mechanical difficulty in obtaining the milk desired?
Posseting
Posseting refers to the recurrent production of mouthfuls of milk from a recently
fed, satisfied child. This is normal and perhaps universal. The experienced clinician
must provide reassurance to the parent on the basis of a careful history and
observation of the phenomenon. Adequate weight gain strongly implies that the
infant is thriving. In the first few days, when weight loss is the norm, a healthy
eager child who produces mouthfuls of milk is unlikely to be concealing a serious
disease.
Vomiting
Vomiting is the production of larger amounts of emesis than mere mouthfuls. The
occasional episode is common, even in the healthy infant, but this symptom
should always be taken seriously. The vomiting infant who is otherwise clinically
well and gaining weight at a normal rate is being overfed: there is no other logical
explanation. This simple principle can be very reassuring to parents. The infant
who is vomiting and failing to gain weight adequately has a problem that must be
taken very seriously. The differential diagnosis is large (see Chapter 40). The
following general principles may be helpful to the pediatrician.
Establish the timing and quality of the vomiting. Huge projectile vomits may
indicate pyloric stenosis (note that this is more common in 4 to 8 weeks of
life and in boys, but it has been described in newborns and in females).
Meningitis may also present as projectile vomiting. Warning: Both pyloric
stenosis and meningitis may present in more subtle ways.
Always observe a feed and try to establish whether it is accomplished

satisfactorily. Mechanical problems, such as cleft palate, may present as
vomiting during the feed itself.
Any child who looks unwell and is vomiting may be septic. Other possibilities
include metabolic disease.
A full clinical examination is required for any newborn presenting with

vomiting. Abdominal palpation may reveal the tenderness and guarding
associated with necrotizing enterocolitis. Tachypnea may point to infection
or metabolic disease, triggering a full infection screen, often including
lumbar puncture, and preliminary investigations for underlying metabolic
conditions.
Common Neonatal Problems

Potential neonatal problems include cold and hypoglycemia.
Cold
Delivery suites that are designed for maternal comfort may be at inadequate
temperatures for the infant. The neonate is limited in compensatory mechanisms,
including shivering and vasoconstriction. An acute cold insult may cause the infant
to become hypoglycemic and apneic. Chronic cold injury may result in failure to
thrive (19).
Hypoglycemia
Over the years neonatal hypoglycemia has been defined in a multitude of ways.
An infant may be said to be hypoglycemic when the blood glucose concentration is
less than 2.6 mmol/L (20). Indications for testing blood glucose concentration
include growth retardation, jitteriness, lethargy, cold injury, suspected infection,
asphyxia, and if the infant is born of a mother with diabetes mellitus. When
present, the predisposing condition should be treated. Meanwhile, a strategy must
be adopted that will return the plasma glucose to normal levels promptly. This
may require the establishment of an intravenous infusion of glucose. There is no
evidence-based level for such treatment, but if the blood glucose is less than 1.7
mmol/L consideration should be given to the use of intravenous therapy. If the
baby is systemically well, it is usually possible to avoid admission to a neonatal
unit. Increasing the frequency of oral feeds, hand expression of breast milk, and
supplementation with proprietary milk may keep the infant by mother's side.
Asymptomatic hypoglycemia warrants careful observation, and the need for
intervention is carefully assessed.
Skin Appearance
Chapter 55 addresses this topic in detail. The most common skin appearances that
cause parental anxiety are capillary hemangioma; dry, cracked skin; an inflamed
perineum; and a maculopapular rash.
Capillary hemangioma (stork's beak mark), is seen most frequently at the
bregma and the nape of the neck. This red or pink capillary nevus blanches on
pressure and fades over the early months of life, being most prominent during
crying.
Dry, cracked skin is common particularly in postmature infants. Weeping and
erythema should lead the clinician to further pursuit of a diagnosis.
Inflamed perineum, is usually caused by one of two events: ammoniacal
dermatitis and candida. Ammoni-acal
P.358
dermatitis tends to show itself as acute erythema on areas of skin exposed to the
diaper. Candida is similar except that the abnormality also invades the perineal
crease including the groins and often produces satellite lesions over the
suprapubic area. This finding should trigger the search for thrush in the mouth;
especially on the buccal mucosa.
In the early days of life a maculopapular rash is commonly seen. The papules are
typically pale in the center and surrounded by a halo of erythema. In the well
child, this is almost certainly erythema toxicum, which is an entirely benign selfresolving appearance requiring only reassurance. In its most flamboyant form,
even experienced neonatologists may wish to culture any weeping area to ensure
that the infant does not have staphylococcal dermatitis.
The Clicky Hip

In the absence of a screening program, the incidence of developmental dysplasia
of the hip (DDH) would be approximately 1 in 1,000 infants delivered (21).
Chapter 19 describes how to examine an infant for DDH. Only after examining
many hundreds of neonatal hips can the clinician be confident of differentiating
between a lax dislocatable hip and one that is stable. The dislocatable hip provides
a clunk in the examiner's hands. A click is frequently felt and can be
differentiated because it is not associated with a movement of the femoral head at
the acetabular lip. Indeed the source of the click is often tendon movement at the
knees. Where there is doubt, the clinician should not hesitate to refer the child for
an orthopedic opinion. It is much better to have a click confirmed than to miss
DDH. The American Academy of Pediatrics has produced an extensive review of
this subject as well as clinical practice guidelines (22,23). Primary screening of all
newborn infants with ultrasonography examination of the hips is unlikely to be
helpful (24). Ultrasonography may, however, be a useful adjunct in studying the
hips of infants suspected to have DDH on clinical examination.
The Systolic Murmur

The population of babies with systolic murmurs changes according to the
discharge policy of the hospital. If infants are discharged home in the early hours
of life, there has frequently not been time for the common systolic murmurs to be
heard, because the pulmonary arterial pressures are still similar to aortic
pressures, and therefore shunts are small in volume. Thus the ductus arteriosus
and ventricular septal defect (VSD) are probably inaudible on day 1. The VSD may
first present at the community pediatrician's office when the second examination
takes place.
Murmurs have been found in 1% to 2% of babies undergoing routine newborn
examination (25), but less than half of these are caused by congenital
malformations (26). Furthermore, of babies found to have congenital cardiac
malformations, fewer than half have a murmur heard in the newborn period (27).
What should the clinician do in the hospital when a systolic murmur is heard? It is
a council of perfection that all parents should be told the diagnosis before
discharge home. Currently this ideal may not be achievable because of lack of
local expertise in echocardiography. In practice, the pediatrician must be satisfied
that the child is well, in no distress, with a normal oxygen saturation in room air
and with no cardiac failure. Careful auscultation may lead to a probable diagnosis
based on the character, position, and transmissibility of the murmur. Even the
most experienced pediatrician should avoid being dogmatic about the complete
diagnosis: the typical murmur of VSD may be the only evidence of tetralogy of
Fallot, and parents will rightly feel aggrieved by being told that this is a small hole
in the inner heart wall that should close off by itself. An electrocardiogram is
carried out, and, provided this is normal, the family is provisionally reassured and
an appointment is made for early review when the ductus arteriosus might be
expected to have closed. On review, persistence of a murmur requires early
referral to a pediatric cardiologist for a definitive opinion including
echocardiography.
Jaundice
Jaundice is a yellow pigmentation of the skin and sclera, and is manifest to some
degree in two thirds of babies in the first week of life (28). Most jaundiced babies
are normal (29). The differential diagnosis and treatment of neonatal jaundice are
described in Chapter 35. Each jaundiced infant should be carefully examined to
exclude infection. The pediatrician should also inquire about mother and infant
blood groups and the result of the Coombs (direct antiglobulin) test. In a child
established on milk feeds, test a sample of urine for reducing substances: the
finding of a nonglucose sugar may indicate a diagnosis of galactosemia. If the
child is well, however, the majority of infants manifest jaundice as a transient
phenomenon that is more common in breast-feeding babies (30). This is often a
result of lower volumes of milk swallowed in early breast-feeding, and the
difference is largely eliminated by encouraging the mother to feed her infant every
2 to 3 hours (31). Jaundice should never be a reason for changing a baby from
breast-feeding to bottle-feeding. Transcutan-eous bilirubin monitoring is a useful,
noninvasive technique that can help identify a population of babies who require to
be tested by plasma bilirubin levels. Each perinatal unit should have clear
protocols for initiation of phototherapy, a therapy whose main purpose is to avoid
exchange transfusion, which, in turn, is intended to avoid bilirubin encephalopathy.
Cryptorchidism
At birth approximately 2% of males exhibit failure of descent of one or both
testes. Most of these will descend in the next four months of life, perhaps because
of a postnatal surge of testosterone. It is true that cryptorchidism is associated
with decreased fertility and increased malignancy, but it is not yet
P.359
clear that orchidopexy is preventive. Nevertheless, pediatric surgeons may wish to
operate when the child is around 1 year of age.
SPECIFIC CONGENITAL INFECTIONS

Syphilis
Syphilis is again beginning to show itself as a significant pathogen worldwide, and
consequently congenital syphilis is a resurgent problem. Adequate treatment of
the mother prior to 18 weeks gestation almost always protects the fetus. If the
mother is untreated, there is up to a 70% chance that the newborn will be
infected. There may be fetal loss or hydrops. A surviving newborn infant may
present with anemia, hepatosplenomegaly, and pneumonia.
The subject of screening is a vexed one, and each community may reasonably
develop a program based on the incidence of congenital syphilis and its
anticipated occurrence in years to come.
The Centers for Disease Control and Prevention (CDC) (32) recommends that
Pregnant women should be screened early in pregnancy.
In areas of high syphilis prevalence, or in patients at
high risk, screening should be repeated in the third
trimester and again at delivery. An infant should not be
released from the hospital until the serologic status of its
mother is known.
Diagnosis in the mother is based on the visualization of Treponema pallidum by
dark ground microscopy or complement fixation (Wasserman) or flocculation tests
(Kahn, Venereal Disease Research Laboratory [VDRL]). The T. pallidum
immobilization test (TPI) and the direct fluorescent treponema antibody
absorption (FTA-Abs) immunoglobulin (Ig) M test are more specific. The
obstetrician and pediatrician should work closely with the microbiology laboratory
to ensure that the results are correctly interpreted. Such interpretation may vary
based on the mother's treatment or partial treatment during pregnancy.
Evaluation of the infant includes the following measures (32):
Complete physical examination
Serologic test for syphilis (the same test performed on the mother so that
titers can be compared)
Lumbar puncture for analysis of cells, protein, and VDRL
Long-bone radiographs
The evaluation and treatment of maternal and congenital syphilis are complicated
and should always be managed by a multidisciplinary team.
Hepatitis B
The mother who tests positive for hepatitis B surface antigen (HBsAg-positive) can
transmit hepatitis B virus (HBV) perinatally to her newborn. In up to 90% of
newborns, the immune response to infection is incomplete and a chronic infective
carrier state occurs. The disease in the newborn may present as a very wide
spectrum from apparent normality to severe, fatal hepatitis. Perinatally infected
infants are at particular risk of becoming chronic carriers, and, even if free of
chronic hepatitis, they are at increased risk of developing later cirrhosis or
hepatocellular carcinoma. Because immunoprophylaxis against HBV initiated at
birth is 98% to 99% effective in preventing virus acquisition by the infant,
identification of HBsAg-positive pregnant women is essential (33). In the
developed world each country must make decisions about a screening program
based on prevalence of hepatitis B, incidence of congenital disease, and the
resources available to tackle this major problem. In the United States and
Canada, confining testing to high-risk women (e.g., Asian or African race,
intravenous drug abuse, multiple sexual partners) may miss 50% of those who
are HBsAg-positive; therefore, prenatal HBsAg testing of all pregnant women is
recommended (33).
Treatment of the infant whose mother is HBsAg-positive or hepatitis B e antigen
(HBeAg)-positive or who has had acute hepatitis during pregnancy consists of
immediate, thorough bathing; administration of 0.5 mL of hepatitis B
immunoglobulin (HBIg) intramuscularly within 12 hours of birth; and the first dose
of hepatitis B vaccine (0.5 mL) intramuscularly, concurrently with HBIg but at a
different site.
Additional doses of vaccine are given at 1 and 6 months of age (33).
Babies born to mothers who are HBsAg-positive and anti-HBeAg-positive should
have hepatitis B vaccine but not HBIg. When in doubt, the pediatrician must
always consult with a virologist expert in perinatal disease.
Note that no special isolation is necessary. Although HBV is found in breast milk,
breast-fed infants, even if they are not receiving immunoprophylaxis, are not at
increased risk of acquiring HBV infection, and breast-feeding therefore is allowed.
In the United States, infants born to known HBsAg-positive women should receive
HBIg and the recommended 3 doses of vaccine, but there is no need to delay the
initiation of breast-feeding until after the infant is immunized (33).
In some countries a hepatitis B immunization series is recommended for all
infants, including those of HBsAg-negative mothers. National guidelines must be

consulted and local programs adapted to the community need. This requires the
input of an expert multidisciplinary team.
Recently, because of concerns about thiomersal, which is an organic compound
containing mercury, thimerosal-free hepatitis B vaccines have been developed.
Human Immunodeficiency Virus (HIV)

Human immunodeficiency virus (HIV) is a retrovirus that can cause meningitis and
encephalopathy. Vertical transmission may occur transplacentally, and also
acutely during labor. Such transmission occurs in 13% to 45% of cases, but this
risk
P.360
can be reduced by zidovudine and perhaps by cesarean section delivery.
Approximately one-third of infected children will become symptomatic in the first
year of life.
Maternal antibodies are transferred passively to the infant and are detectable for
15 to 18 months; therefore, HIV culture and identification of HIV proviral
deoxyribonucleic acid (DNA) by means of the polymerase chain reaction (PCR) are
the preferred diagnostic tests for the newborn. Because false-negative results are
common within the first month of life, retesting at 4 to 6 months is recommended.
All exposed infants need comprehensive follow up, preferably by a pediatrician
with a particular expertise in HIV infection and its profound effects on families.
There are documented cases of infants acquiring HIV infection through breastfeeding. The American Academy of Pediatricians (AAP) recommendation states,
In the United States, where safe alternative sources of feeding are readily
available and affordable, HIV-infected women should be counseled not to
breastfeed their infants (34). This recommendation is now widely adopted
throughout the developed world. No special isolation for mother or infant is
necessary as long as universal precautions are followed.
POST-DISCHARGE
Car Safety
Car safety ideally should be planned before delivery of the infant. Staff should be
aware of the regulations of the national car safety authorities to ensure that the
guidelines available in the perinatal service are compatible with the law and are
up to date.
The following principles are important:
The child car seat should be suitable for the weight and size of the child.
The seat should have a United Nations E mark, a British standards

kitemark, or a European e mark.
The parent fitting the car seat should ensure that the adult seat belt passes
through all the correct guides.
If the adult seat has an airbag in front of it, do not fit a rear-facing child
restraint.
It should be possible to fit an adult hand between the child's chest and the
restraining harness.
Timing and Planning

Throughout the developed world there is great variation in the timing of discharge
home. It should be dictated primarily by the well-being of the infant and of the
mother. Mothers with other children at home may be anxious to achieve early
discharge, perhaps within hours of delivery. In the United Kingdom this is
exemplified by the DoMInO (Domiciliary Midwife In and Out) scheme, where a
mother is cared for antenatally, intrapartum and postpartum by the same midwife
or team of midwives. No matter what the program available, the professionals
should ensure that mother and baby are healthy and that they have been
adequately prepared for life in the community.
A checklist should be completed and signed by parents and professionals prior to
discharge. The checklist should include at least the following items:
Parents are able to care for their infant independently.
Feeding is satisfactory and will be further supported in the community.
If the child is bottle-feeding, parents demonstrate ability to sterilize bottles

and teats.
Parents have been given information on choking, crying, frequency and
consistency of passage of stool.
Parents given advice on baby's day and night clothing, outdoor clothing, and
how to assess baby's temperature.
Parents given verbal and written information about prevention of sudden

infant death syndrome.
Audiologic screening has been performed (where this is available).
Arrangements for follow up are clear, whether this is domiciliary visiting, at

the practitioner's office in the community, or by return to a hospital clinic.
Parents have discussed car seat safety and have a car seat that meets the
national safety requirements.
Parents have demonstrated satisfactory knowledge of elementary

cardiopulmonary resuscitation.
Parents possess any drug prescriptions and have demonstrated ability to

administer drugs to the baby.
Parents possess a written record of any immunizations given and have clear
written plans for future immunizations.
It is also important that the discharge plan for mother and infant be
individualized. Factors that may alter the plan include baby's birth weight, any
illnesses or congenital abnormalities present, parents' wishes and confidence,
educational attainment of the parents, staff knowledge of parents' lifestyle and
support available, whether the home is urban or rural, and whether transportation
is readily available. Special planning is required to support an adolescent mother
and her infant: the responsibility of care must be clear, and robust systems should
be put into place to allow the teenage mother to return to school whenever this is
practical.
The hospital should communicate clearly with the team responsible for care of the
family in the community to ensure that continuity of care is seamless.
REFERENCES
1. Howie PW, Forsyth JS, Ogston SA, et al. Protective effect of breastfeeding
against infection. BMJ 1990;300:11-16.
2. Gillman MW, Rifas-Shiman SL, Camargo CA Jr, et al. Risk of overweight

among adolescents who were breast-fed as infants. JAMA 2001;285:2461-2467.
3. Alm B, Wennergren G, Norvenius SG, et al. Breastfeeding and the sudden

infant death syndrome in Scandinavia, 1992-1995. Arch Dis Child 2002;86
(6):400-402.
4. McVea KL, Turner PD, Peppler DK. The role of breast feeding in sudden infant
death syndrome. J Hum Lact 2000;16(1):13-20.
P.361
5. Lane PA, Hathaway WE. Vitamin K in infancy. J Pediatr 1985; 106:351-359.
6. AAP Committee on Fetus and Newborn and ACOG Committee on Obstetrics:

Maternal and fetal medicine. Postpartum and follow-up care. In: Freeman RK,
Poland RL, eds. Guidelines for perinatal care, 3rd ed. Elk Grove Village, IL:
American Academy of Pediatrics and American College of Obstetricians and
Gyneco-logists, 1992:91.
7. Wiswell TE, Gannon CM, Jacob J, et al. Delivery room management of the
apparently vigorous meconium-stained neonate: results of the multicenter,
international collaborative trial. Pediatrics 2000;105:1-7.
8. Mercer JS. Current best evidence: a review of the literature on umbilical cord
clamping. J Midwifery Womens Health 2001;46: 402-414.
9. Cunningham FG, MacDonald PC, Gant NF. Williams Obstetrics, 18th ed.
Norwalk, CT: Appleton & Lange, 1989:307.
10. Thompson JE, Clark DA, Salisbury B, et al. Footprinting the newborn infant:
not cost effective. J Pediatr 1981;99:797-798.
11. Birenbaum HJ, Glorioso L, Rosenberger C, et al. Gowning on a postpartum
ward fails to decrease colonization in the newborn infant. Am J Dis Child

1990;144:1031-1033.
12. Hudome SM, Fisher MC. Nosocomial infections in the neonatal intensive care
unit. Curr Opin Infect Dis 2001;14(3):303-307.
13. Committee on Infectious Diseases, American Academy of Pediatrics.

Prevention of neonatal ophthalmia. In: Peter G, Lepow ML, McCracken GH Jr, et
al., eds. Report of the committee on infectious diseases, 22nd ed. Elk Grove
Village, IL: American Academy of Pediatrics, 1991:546.
14. Clark DA. Times of first void and first stool in 500 newborns. Pediatrics
1977;60:457-459.
15. Laing IA, Wong CM. Hypernatraemia in the first few days: is the incidence
rising? Arch Dis Child Fetal Neonatal Ed 2002;87:F158- F162.
16. World Health Organization. Protecting, promoting and supporting breast

feeding: the special role of maternity services (a joint WHO/UNICEF statement).
Geneva: World Health Organization, 1989.
17. Olson M. The benign effects on rabbits' lungs of the aspiration of water
compared with 5% glucose or milk. Pediatrics 1970;46: 538-547.
18. Driscoll JM Jr. Routine and special care. In: Fanaroff AA, Martin RJ, eds.
Neonatal-perinatal medicine, 4th ed. St. Louis: CV Mosby, 1987:441.
19. Oliver TK Jr. Temperature regulation and heat production in the newborn.
Pediatr Clin North Am 1965;12:765-799.
20. Lucas A, Morley R, Cole TJ. Adverse neurodevelopmental outcome of

moderate neonatal hypoglycaemia. BMJ 1988;297:1304-1308.
21. Leck I. Congenital dislocation of the hip. In: Wald N, Leck I, eds. Antenatal
and neonatal screening. Oxford: Oxford University Press; 2000.
22. Committee on Quality Improvement and Subcommittee on Developmental
Dysplasia of the Hip. Clinical practice guideline: early detection of

developmental dysplasia of the hip. Pediatrics 2000;105:896-905.
23. Lehmann HP, Hinton R, Morello P, et al. Developmental dysplasia of the hip
practice guideline: technical report. Committee on Quality Improvement, and
Subcommittee on Developmental Dysplasia of the Hip. Pediatrics 2000;105:e57.
24. Gardiner HM, Dunn PM. Controlled trial of immediate splinting versus
ultrasonographic surveillance in congenitally dislocatable hips. Lancet
1990;336:1553-1556.
25. Arlettaz R, Archer N, Wilkinson AR. Natural history of innocent heart

murmurs in newborn babies: controlled echocardiographic study. Arch Dis Child
1998;78:F166-F170.
26. Ainsworth SB, Wyllie JP, Wren C. Prevalence and significance of cardiac
murmurs in neonates. Arch Dis Child 1999;80:F43-F45.
27. Wren C, Richmond S, Donaldson L. Presentation of congenital heart disease

in infancy: implications for routine examination. Arch Dis Child 1999;80:F49F53.
28. Maisels MJ, Newman TB. Jaundice in the healthy full-term infant: time for
reevaluation. In: Klaus MH, Fanaroff AA, eds. 1990 Yearbook of neonatal and
perinatal medicine. St. Louis: Mosby Year Book, 1990:iv.
29. Newman TB, Maisels MJ. Does hyperbilirubinemia damage the brain of
healthy full-term infants? Clin Perinatol 1990; 17:331-358.
30. Schneider AP II. Breast milk jaundice in the newborn. JAMA 1986;255:3270.
31. Yamauchi Y, Yamanouchi I. Breast-feeding frequency during the first 24

hours after birth in full-term neonates. Pediatrics 1990;86:171-175.
32. Centers for Disease Control and Prevention. 1989 Sexually transmitted
diseases treatment guidelines. MMWR Morb Mortal Wkly Rep 1989;38:9.
33. Committee on Infectious Diseases, American Academy of Pediatrics.

Hepatitis B. In: Peter G, Lepow ML, McCracken GH Jr, et al, eds. Report of the
committee on infectious diseases, 22nd ed. Elk Grove Village, IL: American
Academy of Pediatrics, 1991; 238.
34. Committee on Infectious Diseases, American Academy of Pediatrics. AIDS

and HIV infections. In: Peter G, Lepow ML, McCracken GH Jr, et al, eds. Report
of the committee on infectious diseases, 22nd ed. Elk Grove Village, IL:
American Academy of Pediatrics, 1991:115.
Chapter 21
Fluid and Electrolyte Management
Edward F. Bell
William Oh
Infants who are born prematurely or who are critically ill cannot regulate their
own intake of fluids and nutrients. Moreover, enteral feeding is often limited by
feeding intolerance or medical problems that preclude or limit use of the
gastrointestinal tract for feeding. In other cases, the infant presents with
disordered fluid and electrolyte balance as a primary result of an underlying
illness. In all of these situations, water and electrolytes must be provided by
prescription of the health provider. Prescribing the correct amounts of water and
electrolytes helps to assure the infant's healthy recovery.
The goal of fluid and electrolyte management is to replace losses of water and
electrolytes so as to maintain normal balance of these essential substances during
growth and recovery from disease. A subsidiary aim in the first days of life is to
allow successful transition from the aquatic environment of the uterus into the
arid extrauterine milieu. The principles of fluid and electrolyte management in the
neonatal period are similar to those established for older children, except for some
variations and specific features of body composition, insensible water loss (IWL),
renal function, and neuroendocrine control of fluid and electrolyte balance.
To manage fluid therapy of newborns appropriately, the clinician should
understand the normal physiologic mechanisms that govern water and electrolyte
balance and the variations in these mechanisms that can occur in sick or
premature infants. The clinician should develop a systematic approach to the
estimation of fluid and electrolyte requirements for correction of deficits and
replacement of ongoing losses, both normal and abnormal. Finally, the results of
fluid and electrolyte management must be carefully monitored so that the intakes
of water and electrolytes can be adjusted as needed.
BODY COMPOSITION OF THE FETUS AND

NEWBORN INFANT
Changes in Body Water During Growth
The total body water (TBW) is divided into two major compartments, intracellular
(ICW) and extracellular (ECW). The ECW is further divided into the interstitial
water and the plasma volume, which is the intravascular component of the ECW
(Fig. 21-1).
In the early stages of fetal development, a large part of the body consists of water
(1). It has been estimated that TBW is 94% of the body weight during the third
month of fetal life. As gestation progresses, the TBW per kilogram declines. By 24
weeks the TBW is approximately 86%, and by term it is about 78% of body
weight (Fig. 21-2). There also are characteristic changes in the partition of body
water between ECW and ICW during development. ECW decreases from 59% of
body weight at 24 weeks of gestation to about 44% at term, and ICW increases
from 27% to 34% of body weight during the same period (Table 21-1)
(1,2,3,4,5,6). Infants born prematurely thus have higher TBW and ECW per
kilogram than their term counterparts (7,8,9), and small-for-gestational-age
infants have higher TBW per kilogram than do appropriate-for-gestational-age
infants (9).
After birth, TBW per kilogram of body weight continues to fall, due primarily to
contraction of the ECW (2,7,8,10,11,12,13). This mobilization of extracellular fluid
occurs in conjunction with the improvement in renal function that takes place
following birth (14,15), which is thought to occur as a result of increasing
glomerular filtration rate and perhaps, too, as a result of increasing levels of the
epithelial transport proteins involved in renal tubular function (16). It has also
been suggested that atrial natriuretic peptide
P.363
plays a role in the postnatal contraction of the ECW (13). Various studies have
shown an increase, decrease, or no change in the ICW after birth. ICW probably
increases roughly in proportion to body weight in the first weeks of postnatal life
(2,7,8,17). Thereafter, ICW increases faster than body weight and exceeds ECW
by 3 months of age (Fig. 21-2) (1,2). These postnatal changes in body water and
its partition between ECW and ICW are influenced by the intake of water and
electrolytes (11,18). Failure to allow the normal postnatal contraction of ECW in
premature infants may increase the risk of significant patent ductus arteriosus
(PDA) (19), necrotizing enterocolitis (NEC) (20,21,22), and bronchopulmonary
dysplasia (BPD) (23,24).
Figure 21-1 Distribution of body water in a term newborn infant.
Solute Distribution in Body Fluids

The major cation in the blood plasma is sodium (Fig. 21-3). Potassium, calcium,
and magnesium constitute the balance of the cation fraction. The primary anion is
chloride, with protein, bicarbonate, and some undetermined anions constituting
the balance of the anions. The interstitial fluid (i.e., nonplasma ECW) has a solute
composition that is similar to plasma except that its protein content is lower. The
ICW contains potassium and magnesium as its primary cations, and phosphate,
both organic and inorganic, is the major anion, with bicarbonate contributing a
smaller fraction.
TABLE 21-1 CHANGES IN BODY WATER AND ELECTROLYTE COMPOSITION
DURING INTRAUTERINE AND EARLY POSTNATAL LIFE
Component
Total body water (%)

Extracellular water (%)
Intracellular water (%)
Sodium (mEq/kg)
Potassium (mEq/kg)
Chloride (mEq/kg)
24
Gestational Age (Weeks)

1 to 4 Weeks
28
32
36
40 After Term Birth
86
59
27
99
40
70
84
56
28
91
41
67
82
52
30
85
40
62
80
48
32
80
41
56
78
44
34
77
41
51
74
41
33
73
42
48
Data from Friis-Hansen B. Changes in body water compartments during

growth. Acta Paediatr 1957;6
(Suppl 110):168; Friis-Hansen B. Body water compartments in children:
changes during growth and related changes in body composition. Pediatrics
1961;28:169181; Ziegler EE, O'Donnell AM, Nelson SE, et al. Body
composition of the reference fetus. Growth 1976;40:329341; Forbes JB,
Perley A. Estimation of total body sodium by isotopic dilution. II. Studies
on infants and children: an example of a constant differential growth ratio.
J Clin Invest 1951;30:566574; Cheek DB. Observations on total body
chloride in children.
Pediatrics 1954;14:510; Romahn A, Burmeister W. [Body composition
during the first two years of life: analysis with the potassium 40 method].
Klin Pdiatr 1977;189:321327.
Figure 21-2 Changes in body water during gestation and infancy. (Adapted from
Friis-Hansen B. Changes in body water compartments during growth. Acta
Paediatr 1957;46(Suppl 110):1-68., with permission.)
The electrolyte composition of the body fluids of the newborn infant is largely
determined by gestational age. Premature infants contain more sodium and
chloride per kilogram of body weight than term infants (3,4,5) because of their
larger ECW (Table 21-1). Total body potassium content largely reflects ICW and is
similar or slightly lower
P.364
per kilogram of body weight in premature infants than at term (3,6). These
concepts are important in the management of fluid and electrolyte therapy for
newborn infants.
Figure 21-3 Ion distribution in the blood plasma, which represents extracellular
fluid, and in the intracellular fluid compartment.
In the fetus, fluid and electrolyte balance depends on maternal homeostasis and
placental exchange. Thus, fluid and electrolyte status at birth is influenced by the
maternal fluid and electrolyte management in labor (25,26,27,28).
INSENSIBLE WATER LOSS

The loss of water by evaporation from the skin and respiratory tract is known as
insensible water loss (IWL). About 30% of IWL normally occurs through the
respiratory tract as moisture in expired gas (29,30,31), with the remaining 70%
lost through the skin. IWL can be expressed in reference to body surface area
(m2) or weight (kg). IWL depends more on surface area than weight, but it is
commonly expressed per kilogram because weight is more easily determined than
area.
TABLE 21-2 FACTORS AFFECTING INSENSIBLE WATER LOSS IN

NEWBORN INFANTS
Effect on Insensible Water
Factor
Level of maturity (32,33,35, 36 and

37)
Loss (IWL)
Inversely proportional to
birth weight and gestational
age (Fig. 21-4)
Respiratory distress (hyperpnea) (38) Respiratory IWL increases
with rising minute ventilation
when dry air is being
breathed
Environmental temperature above
Increased in proportion to
neutral thermal zone (29,39,40)
increment in temperature
Elevated body temperature (29,39) Increased by up to 300%
Skin breakdown or injury
Increased by uncertain
magnitude
Congenital skin defect (e.g.,
Increased by uncertain
gastroschisis, omphalocele, neural
magnitude until surgically
tube defect)
corrected
Radiant warmer (33,41,42,43,44 and Increased by about 50%
45)
Phototherapy (43,46,47)
Increased by about 50%
Motor activity and crying (29,49,50) Increased by up to 70%
High ambient or inspired humidity
Reduced by 30% when
(29,31)
ambient vapor pressure
isincreased by 200%
Plastic heat shield (32,44,52)
Reduced by 30% to 70%
Plastic blanket (52, 53,54) or
Reduced by 30% to 70%
chamber (54,55)
Semipermeable membrane (56, 57
Reduced by 50%
and 58)
Topical agents (59,60)
Reduced by 50%
Figure 21-4 Relation between insensible water loss (IWL) and birth weight of 5-day-old (mean) infants in incubators.
(Data from Wu PY, Hodgman JE. Insensible water loss in preterm infants: changes with postnatal development and nonionizing radiant energy. Pediatrics 1974;54:704-712, as redrawn in Shaffer SG, Weismann DN. Fluid requirements in
the preterm infant. Clin Perinatol 1992;19: 233-250, with permission.)
A number of factors are known to influence IWL in a predictable manner (Table 21-2)
(29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,
55,56,57,58,59,60). When expressed per kilogram of body weight, IWL is inversely
proportional to birth weight and gestational age (Figs. 21-4 and 21-5) (32,33,35).
P.364
In other words, smaller, more immature infants have larger IWL per kilogram (Table 21-3). The
same is true if IWL is expressed per square meter of body surface (36,37). Therefore,
although the greater IWL of smaller premature infants is partly due to the increased ratio
of surface area (skin and respiratory tract) to body weight, it also is thought to be related
to their thinner skin, greater skin blood flow, larger body water per kilogram of body weight,
and higher respiratory rate. Because skin permeability to water varies inversely with
gestational age, the degree of immaturity is an important determinant of cutaneous IWL
independent of birth weight.
Figure 21-5 Insensible water loss (IWL) as a function of birth weight in

premature infants nursed under radiant warmers. (Adapted from
Costarino AT, Baumgart. Controversies in fluid and electrolyte therapy
for the premature infant. Clin Perinatol 1988; 15:863-878, with
permission.)
Factors That Increase Insensible Water Loss

An increase in minute ventilation increases the respiratory IWL (38) as
long as the water vapor pressure is less in the inspired than in the
expired gas. Increased minute ventilation may occur in infants with
cardiac disease, pulmonary dysfunction, or metabolic acidosis.
TABLE 21-3 AVERAGE INSENSIBLE WATER LOSSa OF PREMATURE

INFANTS IN INCUBATORS
Birth Weight Range (kg)
Age (d) 0.50-0.750.75-1.001.00-1.251.25-1.501.50-1.751.75-2.00
07
714
100
80
65
55
40
20
15
60
50
40
30
20
Insensible water loss (mL/kg/day).

Data from Wu PY, Hodgman JE. Insensible water loss in preterm
infants: changes with postnatal development and non-ionizing
radiant energy. Pediatrics 1974;54:704712; Okken A, Jonxis JH,
Rispens P, et al.
Insensible water loss and metabolic rate in low birthweight
newborn infants. Pediatr Res 1979;13:10721075; Hammarlund K,
Sedin G. Transepidermal water loss in newborn infants. III.
Relation to gestational age. Acta Paediatr Scand 1979;68:795801.
Environmental temperature above the neutral thermal zone increases IWL in

proportion to the increment in temperature (29,39,40). This effect can
occur even without a rise in body temperature. In contrast, a subneutral
environmental temperature is not associated with reduced IWL, although
metabolic heat production is increased (40). Increased body temperature,
whether caused by fever or environmental overheating, elevates IWL
(29,39).
Skin breakdown or injury disrupts the barrier against cutaneous evaporation
and raises IWL. Skin trauma from thermal, chemical, or mechanical injury is
common among critically ill, small, premature infants. Such injury may
result from removal of tape and adherent monitoring devices or from
prolonged skin exposure to disinfectant solutions. IWL also is increased in
conjunction with the skin manifestations of essential fatty acid deficiency, a
potential problem in infants receiving fat-free parenteral nutrition.
Congenital skin defects, such as those seen in gastroschisis, omphalocele,
and neural tube defects, are associated with increased IWL until surgically
corrected.
Use of nonionizing radiant energy, in the form of either a radiant warmer or
phototherapy, has been shown to increase IWL by about 50%
(33,41,42,43,44,45,46,47,48). For infants in incubators with controlled air
temperature, the increase in IWL with overhead phototherapy is most likely
a result of increased body temperature because of the warmer incubator
walls (46). For infants in incubators operated to control skin temperature,
the rise in IWL with phototherapy can be explained by the lower absolute
humidity resulting from the reduced air temperature that accompanies the
warming of the incubator walls by the phototherapy. The impact on IWL of
phototherapy delivered by fiberoptic blankets or pads is not known but is
probably negligible unless the blanket produces a warmer or moister
microenvironment around the infant. Investigators using direct
measurements of transepidermal and respiratory water loss have obtained
P.366
conflicting results regarding the effect of overhead phototherapy on IWL.
One group (47) found an increase in transepidermal water loss with
phototherapy, but another did not (61,62).
TABLE 21-4 RELATIVE AND ABSOLUTE HUMIDITY AS RELATED TO

INSENSIBLE WATER LOSS IN INCUBATORS AND UNDER RADIANT
WARMERS
Measurement
Incubator
Radiant Warmer
Air temperature (C)
35.0
27.6
Saturation pressure (mm Hg)
42.1
27.7
Relative humidity (%)
31.4
39.0
Absolute humidity (mm Hg)
13.2
10.8
Insensible water loss
2.37
3.40
(mL/kg/hour)
Data from Bell EF, Weinstein MR, Oh W. Heat balance in premature

infants: comparative effects of convectively heated incubator and
radiant warmer, with and without plastic heat shield. J Pediatr
1980;96: 460465.
If an infant's IWL is measured at the same skin temperature under a radiant

warmer and in an incubator, the IWL is higher (by about 50%) under the
radiant warmer. IWL is higher because absolute humidity (water vapor
pressure) is lower under the radiant warmer than in the incubator (44). This
may be true even if relative humidity is higher under the radiant warmer
(43,44) because the lower air temperature with the radiant warmer means
that the saturation pressure of water vapor is considerably lower than in
the incubator (Table 21-4). This finding has been confirmed using direct
measurements of transepidermal water loss (45). It is now understood that
the higher IWL with radiant warmers arises from the lower ambient water
vapor pressure and not from higher air velocity or a direct effect of
nonionizing radiation on the skin. The same phenomenon explains the effect
of phototherapy on IWL of infants in incubators operated by skin
temperature servocontrol. The effects on IWL of radiant warmers and
phototherapy are additive; the IWL with the combination is approximately
twice as large as in an incubator without phototherapy (43,48).
Increased motor activity and crying increase IWL by up to 70% (29,49,50).
This effect may be partly due to elevated minute ventilation.
Factors That Reduce Insensible Water Loss

Increasing the humidity or water vapor pressure of inspired gas reduces
respiratory IWL. The inspired humidity is raised by humidifying the airoxygen mixture delivered to a head hood or directly to the infant's upper
airway (e.g., via nasal cannula, face mask, or endotracheal tube) if
respiratory support is required. If the temperature and water content of the
inspired and expired gas are the same, the respiratory IWL will be entirely
eliminated (51). Increasing ambient humidity, for example in an incubator,
reduces total IWL, but respiratory IWL is decreased more than cutaneous
IWL (29); a threefold increase in ambient water vapor pressure, from an
average of 7 to 25 mm Hg, resulted in a 30% reduction in total IWL.
Increasing ambient humidity is facilitated by the design of certain recent
models of incubators. The use of incubator humidification systems should
not be overlooked as a way of reducing IWL.
Plexiglas heat shields are effective in reducing the IWL of small premature
infants in incubators (32,44), especially if the ends are at least partially
enclosed to decrease air movement near the skin. Plexiglas heat shields are
not effective for infants under radiant warmers (44,52) because Plexiglas is
opaque to the infrared energy produced by the radiant heaters. Thin
barriers of saran and other materials reduce IWL of infants under radiant
warmers while allowing the infrared heat to reach the skin (52). These heat
shields presumably reduce IWL by limiting air movement and raising water
vapor pressure near the infant's body surface.
Thin plastic blankets have been found to reduce IWL by 30% to 70% for
infants under radiant warmers and in incubators (52,53,54). Chambers
made of thin plastic material also reduce IWL by a similar amount (54,55).
Semipermeable membranes (56,57,58) and waterproof topical agents

(59,60) reduce IWL from the covered areas by an average of approximately
50%.
Knowledge of these factors that affect IWL is essential for estimating the
water intake required by newborn infants and for making appropriate
adjustments in water intake with changes in care. Of all infants, premature
and critically ill infants are the ones whose IWL is most significantly
influenced by these factors. This is especially true for the extremely
premature infant, of whom more are surviving each year. However, these
are exactly the infants for whom precise maintenance of fluid and
electrolyte balance is most important and for whom the margin of error is
smallest.
NEUROENDOCRINE CONTROL OF FLUID AND

ELECTROLYTE BALANCE
The pituitary gland, adrenal cortex, parathyroid glands, and heart are the
major organs producing hormones involved in the regulation of water and
electrolyte balance in the body. The basic mechanisms by which the
antidiuretic hormone, arginine vasopressin (AVP), is produced and secreted
by the posterior pituitary gland appear to be intact in newborn infants
(63,64,65), even in those who are born prematurely (66). It is not clear at
what age precise quantitative hypothalamic control of AVP production is
established. However, it is known that even in the first week of life, breastfed term infants release AVP in response to a 10% loss of body weight (65).
Aldosterone is the most potent mineralocorticoid produced and secreted by
the adrenal cortex. Its synthesis is regulated by the renin-angiotensin
system, adrenocorticotropic hormone, and plasma concentrations of sodium
P.367
and potassium. These mechanisms appear to be intact in newborn infants,
even those born prematurely (67,68,69,70,71). Increased sodium loss in
the urine of premature infants in the presence of high plasma
concentrations of aldosterone and elevated urinary aldosterone excretion
suggests that the renal tubule is less responsive in premature than in term
infants (69,70). Under conditions of low sodium intake, however, the high
plasma aldosterone concentrations found in sick newborn infants seems to

promote increased sodium reabsorption in the distal nephron (71).
Calcium concentration in the blood of newborn infants is regulated by the
balance between parathyroid hormone (PTH), which is produced by the
parathyroid glands, and calcitonin, which is produced in the thyroid. Serum
PTH concentration is low at birth and rises slowly during the first few days,
both in term and premature infants (72,73,74). The same pattern has been
observed with serum calcitonin concentrations, low at birth and rising
thereafter (73,74). Intravenous calcium infusion to large premature and
term infants caused elevation of serum calcitonin and a corresponding fall
in serum PTH (75). These data together indicate that the hormonal
regulation of calcium metabolism is basically intact in newborn infants, even
those born prematurely.
Atrial natriuretic factor (ANF) is present in the fetal heart early during
development (76). In the human fetus, cardiac atrial levels of ANF increase
during gestation and, by the beginning of the third trimester, exceed adult
human levels; during the same period, fetal ventricular ANF levels decrease
(77). Plasma ANF levels rise after birth, peaking at the time of maximal
postnatal diuresis, usually 48 to 72 hours, and then returning to levels
below those at birth (78,79,80). ANF secretion is stimulated by volume
loading (81), and ANF levels correlate with atrial size (82). ANF in turn
stimulates diuresis and natriuresis and seems to play an important role in
the regulation of extracellular fluid volume in newborn infants (82,83,84).
However, studies of the effects of sodium supplementation on ANF levels
and sodium excretion (83) indicate that premature infants are less
responsive to ANF than are adults.
RENAL FUNCTION IN RELATION TO FLUID AND

ELECTROLYTE THERAPY
Most aspects of renal function are incompletely developed at birth,
especially in premature infants
(15,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,8
0,81,82,83,84,85,86,87,88,89,90). Both glomerular and tubular functions
increase with gestational age at birth (15,85,86,87,90) and with postnatal
age (15,87,88,89). This development seems to depend most directly on
postmenstrual age (gestational age at birth plus postnatal age) and occurs
at approximately the same rate, regardless of whether the infant has been
born or is in utero (86,87,89).
In spite of the immaturity of some aspects of renal tubular function at birth,
the tubules seem to respond to AVP from the first day of life, even in small
premature infants (66). However, the maximal urine concentration of
premature infants, typically 600 mOsm/L, is less than that of term newborn
infants (800 mOsm/L) or adults (1200 mOsm/L) (91,92). Both term and
premature infants can excrete urine with osmolarity as low as 50 mOsm/L
when challenged with an acute water load (92,93,94,95). Although they can
produce dilute urine, newborn infants cannot excrete a water load as rapidly
as adults can (93).
The limitations in renal function in premature infants contribute to the
problems of fluid and electrolyte regulation in various disease states. The
glomerular and tubular functions of premature infants allow them to handle
some physiologic variations in water and electrolyte load, but imbalance
readily occurs when estimations of the water and electrolyte needs are
misjudged, particularly in the case of extremely premature infants.
PRINCIPLES OF FLUID AND ELECTROLYTE

THERAPY
As in older children, three steps should be followed in the management of
infants with fluid and electrolyte disorders:
Estimate the deficits of fluid and electrolytes.
Calculate the amounts of fluid and electrolytes required for

replacement of deficits, maintenance, and replacement of ongoing
abnormal losses.
Institute a system of monitoring the response to therapy.
Estimation of Fluid and Electrolyte Deficits

Fluid Deficit
The body water deficit can be estimated by the degree of dehydration. If
serial body weight measurements are available, the acute weight loss is
considered to represent the water deficit. During the first week of life,
however, weight loss of up to 15% occurs normally as a result of
contraction of extracellular water and tissue catabolism. Even those small

premature infants who can be enterally fed lose an average of 10% of their
body weight during the first 5 days of life (96). Among premature infants
who must be nourished intravenously, it is not uncommon to observe weight
loss of 15% or more without evidence of circulatory or renal insufficiency
resulting from dehydration (97). The precise amount of weight loss desired
during the first week of life has not been established because of a lack of
reliable physiologic data. In general, smaller infants tend to lose larger
fractions of their weight following birth (97). For small premature infants, a
weight loss of 2% to 3% per day is a reasonable target in the first week of
life. Efforts to completely prevent postnatal weight loss risk overhydration
and problems with symptomatic PDA (19). Beyond the first week of life,
acute weight loss should be considered to indicate nonphysiologic
dehydration, and the calculated deficit of water should be replaced.
P.368
If serial body weight data are not available, urine volume and concentration
and physical signs can be used to estimate the degree of dehydration.
Infants with 5% isotonic (i.e., serum sodium concentration of 130 to 150
mEq/L) dehydration have dry mucous membranes, subnormal tear
production with crying, flat or slightly sunken anterior fontanel (when quiet
in the upright position), and oliguria. Infants with 10% isotonic dehydration
have dry mucous membranes, absent tears, sunken eyes and fontanel, cool
extremities, poor skin turgor, and oliguria. Infants with 15% isotonic
dehydration have the aforementioned signs as well as signs of shock, such
as hypotension, tachycardia, weak pulses, mottled skin, and altered
sensorium. Infants with hypertonic dehydration (i.e., serum sodium
concentration above 150 mEq/L) have less severe symptoms than infants
with isotonic dehydration who have lost the same fraction of body water;
the intravascular volume is preserved better with hypernatremia than with
isotonic dehydration. However, infants with hypotonic dehydration (i.e.,
serum sodium below 130 mEq/) may have more severe symptoms with the
same degree of dehydration.
The usual clinical signs of dehydration are more difficult to evaluate in
small premature infants. Their skin and mucous membranes may appear dry
because of thermal or mechanical injury, particularly in infants kept under
radiant warmers. In addition, skin turgor is harder to judge because of the

lack of subcutaneous fat.
Electrolyte Deficits
The nature and extent of electrolyte disturbances often can be determined
by history and physical examination and by measurement of electrolyte
concentrations in serum. Based on serum sodium concentration, electrolyte
disturbances are divided into isotonic, hypertonic, and hypotonic
abnormalities. The type of electrolyte disorder seen in a clinical situation
depends on the cause of fluid and electrolyte abnormality. For example,
severe acute diarrhea usually leads to isotonic dehydration. High IWL, such
as may occur in small premature infants under radiant warmers, may result
in hypernatremic dehydration. Inadequate replacement of salt losses from
diarrhea may produce hypotonic dehydration. Although it may be possible to
anticipate the type of electrolyte disorder accompanying dehydration in
some situations, confirmation must be made by measurement of serum
electrolyte concentrations.
TABLE 21-5 CALCULATION OF SODIUM DEFICIT
Calculation
Serum
of Total
Sodium
Solute
Solute
Sodium
Type of
Concentration
Deficit
Deficit
Deficit
Dehydration
(mEq/L)
(mOsm/kg) a
(mOsm/kg)
(mEq/kg) b
140
(0.7
28
14
12
Isotonic
(10%)
280)(0.6
280)
Hypertonic
(10%)
153
(0.7
280)(0.6
306)
Hypotonic
(10%)
127
(0.7
44
22
280)(0.6
254)
Total solute deficit = (TBW e solute e - (TBW o solute o ), where
subscripts e and o indicate expected and observed, respectively. TBW e =

0.7 L/kg; TBW o = 0.7 - 0.1 = 0.6 L/kg; solute e 140 2 = 280 mOsm/L,
assuming total solute concentration in body water is twice the sodium
concentration in serum;solute o = observed serum sodium 2.
b
Total solute deficit is assumed to be half sodium. Although the serum (and
ECW) has lost this amount ofsodium, only half this amount has been lost to
the environment; the other half has been lost into the cells inexchange for
potassium, which in turn has been lost from the body. In practice,
therefore, only half the amountlisted as sodium deficit should be replaced
as sodium, and the other half should be given as potassium. TBW, total
body water. ECW, extracellular water.
Calculation of Fluid and Electrolyte

Requirements
After replacement of any fluid and electrolyte deficits, the requirements of
newborn infants for water and electrolytes are determined by the rates of
loss of these substances from the body by various routes and by the net
amounts retained by body tissues during changes in body weight and
composition. Knowledge of the usual rates of loss and of expected changes
in body weight and composition in the first postnatal days and during
subsequent growth helps to estimate water and electrolyte requirements.
These estimates are then used to guide the management of fluid and
electrolyte therapy.
Replacement of Fluid and Electrolyte Deficits

The deficit of fluid is calculated from the estimated degree of dehydration
determined from measured body weight loss or by clinical examination. The
rate and composition of initial fluid replacement depend on the severity of

dehydration. As a rule, dehydration of acute onset and short duration
requires more rapid correction. An exception to this rule is the case of
hypertonic dehydration, in which rapid expansion of body water may cause
brain swelling and convulsions.
The deficit of electrolytes is calculated as the difference between total body
solute expected before dehydration and that observed in the dehydrated
state (Table 21-5). It is common to replace half the water deficit over the
first 8 hours and the other half over the next 16 hours. The
P.369
sodium deficit is replaced over 24 hours; more rapid replacement is not
necessary and may risk overcorrection. If the potassium deficit is large, one
should replace it over a longer period (i.e., 48 to 72 hours) to allow
ascertainment of adequate renal function and to avoid the possible cardiac
effects associated with rapid potassium infusion. Initiation of potassium
replacement is best deferred until urine flow is established.
Maintenance Fluid and Electrolytes

IWL, urine, fecal water, and water retained in new tissues during growth
are the four components that one must consider in estimating the daily
maintenance water requirement. Fecal water loss is approximately 5 to 10
mL/kg/day (98,99). The water retained for growth is about 10 mL/kg/ day,
assuming a weight gain of 10 to 20 g/kg/day, 60% to 70% of which is water
(3). In the first week of life, fecal water loss is small and no water is
deposited in new tissues because growth has not yet begun. In fact, water
is lost from body tissues during the period of physiologic extracellular
dehydration. After growth begins, replacement of fecal and growth water
may require up to 20 g/kg/day, but this amount is small compared with the
insensible and urine water losses, the two major routes of water loss that
must be considered in estimating the water intake required to maintain the
desired water balance.
A small portion of the water used to replace these normal losses is derived
from the oxidation of metabolic fuels (i.e., carbohydrate, protein, and fat).
This water of oxidation consists of about 0.60 mL/g of carbohydrate
oxidized, 0.43 mL/g protein, and 1.07 mL/g fat (100). A newborn infant
usually produces 5 to 10 mL/kg/day as water of oxidation. This amount is

small enough to be neglected in most calculations but, for practical
purposes, can be considered to offset the normal fecal water loss of 5 to 10
mL/kg/day.
For a term infant under basal conditions, IWL is approximately 20
mL/kg/day (29). Urine volume depends on the excess of water intake over
losses by other routes (i.e., IWL, feces, growth), and urine concentration is
determined by the urine volume and renal solute load. The range of urinary
water loss within which the infant's immature kidneys can safely excrete the
total renal solute load is determined by the limits of urine concentration
(volume = solute load/urine concentration) (101). A renal solute load of 15
to 30 mOsm/kg/day would require urine volume of 50 to 100 mL/kg/day to
maintain an average urine concentration of 300 mOsm/L. This urine
concentration is near the middle of the range of urine osmolarity that can
be produced by the neonate's kidneys, and it allows a margin of safety for
over- or underestimation of other water requirements.
In the first days of life, a term infant receiving intravenous fluid and
electrolytes would need to excrete about 15 mOsm/kg/day, assuming that
both endogenous solute production and tissue deposition of solute are
negligible. The urine volume of 50 mL/kg/day plus the IWL of 20 mL/kg/day
yield a total maintenance water requirement of 70 mL/kg/day; this assumes
growth and fecal water to be small enough to be offset by the water of
oxidation. If one allows for a negative water balance of 10 mL/kg/day, the
true water requirement at birth is about 60 mL/kg/day. With increasing
postnatal age and enteral feedings, the renal solute load and fecal water
loss increase, and water is deposited in new tissues as growth begins. By
the second week of life, a growing term infant needs 120 to 150 mL/kg/day.
In premature infants, the maintenance water requirement is larger because
of higher IWL (33,35). Therefore, the IWL component of maintenance water
should be increased with decreasing birth weight or gestation. During the
first days of life, the renal solute load is less because little exogenous
solute is provided. If sodium chloride is administered at a rate of 2
mEq/kg/day (4 mOsm/kg/day) and a solute load of 8 mOsm/kg/day
resulting from tissue catabolism is assumed to be excreted (102), a urine
volume of only 40 mL/kg/day is required to excrete this solute with a urine
concentration of 300 mOsm/L. Thus, a small premature infant requires
about 80 mL/kg/day on day 1 (60 IWL + 40 urine -20 for negative balance).
The water requirement for this same infant would be about 150 mL/kg/day
in the second or third week (55 IWL + 85 urine + 10 feces + 10 growth - 10
oxidation). Very premature (less than 26 weeks of gestation) infants in the
first week of life may have considerably higher IWL, raising the total water
requirement to 200 or 300 mL/kg/day or even higher, especially if
maintained in dry air. The minimum water intake of premature infants is
also higher than that of term infants because of premature infant's slightly
lower urinary concentrating capacity (91,92). However, the aforementioned
urine volumes (40 to 100 mL/kg/day) were selected to avoid taxing this
limit of concentration and so are not influenced by this effect of immaturity.
The allowance for IWL should be increased by about 50% for infants under
radiant warmers (33,41,42,43,44) or receiving overhead phototherapy
(33,46). If both are used, the allowance for IWL should be increased by
approximately 100% (43). The effect of fiberoptic phototherapy blankets or
pads on IWL is not known but is probably less than that of overhead
phototherapy. The IWL of infants in incubators also is increased if body or
environmental temperature is too high (29,39,40). The IWL can be reduced
by increasing the ambient or inspired humidity (29,31) or by using certain
types of heat shields (32,44,52), plastic blankets (52,53,54) or chambers
(54,55), semipermeable membranes (56,57,58), or waterproof topical
agents such as paraffin (59,60) (Table 21-2).
The infant's maintenance requirements of sodium, potassium, and chloride
can be estimated by adding the dermal, urinary, and fecal losses to the
amounts retained in the body tissues during growth. The estimated
requirements for sodium, potassium, and chloride are each between 2 and 4
mEq/kg/day (103,104). Small premature infants may require additional
sodium because of increased urinary excretion (105,106,107), especially
during the second and third weeks of life. The magnitude of urinary sodium
excretion is inversely proportional to gestational age (Fig. 21-6) (85).
Figure 21-6 Urinary sodium excretion in infants from 27 to 40 weeks of

gestation. (From Siegel SR, Oh W. Renal function as a marker of human
fetal maturation. Acta Paediatr Scand 1976;65: 481-485, with permission.)
P.370
Ongoing Abnormal Losses of Fluid and Electrolytes

Ongoing abnormal losses must be replaced along with correction of
established deficits and provision of maintenance fluid and electrolytes.
Abnormal losses may occur with vomiting or diarrhea, ileostomy output, or
removal by aspiration of gastrointestinal, pleural, peritoneal, or
cerebrospinal fluid. The amount of extra water required can be determined
by carefully measuring the volume lost. The additional amounts of
electrolytes required can be estimated by measuring their concentrations in
an aliquot of fluid (Table 21-6).
Example of Fluid and Electrolyte Calculation

Consider a 3-kg infant who presents with 10% isotonic dehydration (i.e.,
serum sodium = 140 mEq/L). In the first 24 hours of therapy, the infant
should be given 27 mEq of sodium chloride in 600 mL of water as a
dextrose solution (Table 21-7). If the potassium deficit is to be corrected
over 72 hours, 13 mEq of potassium chloride (21/3 + 6) should be given

during the first 24 hours. This fluid could then be ordered as dextrose
solution with 45 mEq NaCl/L and 20 mEq KCl/L to be infused at 25 mL/hour.
If the infant also has significant metabolic acidosis, some or all of the
sodium could be given as sodium bicarbonate or sodium acetate.
TABLE 21-6 ELECTROLYTE CONTENT OF BODY FLUIDS
Sodium
Potassium
Chloride
(mEq/L)
(mEq/L)
(mEq/L)
20-80
5-20
100-150
Small intestine
100-140
5-15
90-120
Bile
120-140
5-15
90-120
45-135
3-15
20-120
10-90
10-80
10-110
130-150
2-5
110-130
Fluid Source
Stomach
Ileostomy
Diarrheal stool
Cerebrospinal
fluid
The initial glucose concentration is determined by the estimated water

requirement and the desired glucose intake, which is usually 5 to 8
mg/kg/minute. Smaller, more premature infants, because they require more
water and may tolerate less glucose, are usually begun on solutions
containing lower concentrations of glucose (e.g., 5 g/dL); larger term or
near-term infants require higher glucose concentrations (e.g., 10 g/dL).
The fluid and electrolyte requirements of very small premature infants vary
widely and are difficult to predict. Therefore, close monitoring of the fluid
and electrolyte balance of these infants is especially important so that any
imbalance can be detected as soon as possible.
Monitoring the Effectiveness of Fluid and

Electrolyte Therapy
During the course of parenteral fluid therapy, detailed and organized data
collection is necessary for monitoring the adequacy of fluid and electrolyte
intake. Data that should be collected and recorded regularly at designated
intervals include water and electrolyte intake by all routes, measurable
output of water, body weight change, and serum electrolyte concentrations;
in addition, clinical assessment should be made for the presence of
dehydration, edema, or acute water overload. A calibrated infusion pump
should be used to assure accurate administration of prescribed parenteral
fluids. Accurate measurement of urine volume is difficult in small infants. It
is possible to collect urine from male infants by placing the penis into a
collecting vessel that is taped to the abdomen. Urine is then aspirated from
the vessel with a catheter and syringe. Urine volume can also be estimated
by comparison of the weights of dry and wet diapers, provided the diaper is
weighed soon enough to avoid evaporation of the urine.
Inadequate fluid may be administered if the fluid maintenance requirement
is underestimated or if a preexisting
P.371
deficit or ongoing loss is neglected or underestimated. Insufficient water
intake leads to reduced urine volume and increased urine concentration,
and if these compensatory measures are inadequate, water is mobilized
from body stores to provide for obligatory insensible loss and to allow
solute excretion. This results in weight loss, clinical signs of dehydration,
metabolic acidosis, and hemoconcentration. As serum osmolarity rises,
neurologic sequelae of hypertonicity may occur. In severe cases, untreated
dehydration may lead to decreased circulating blood volume, acute renal
failure, and finally to death from cardiovascular collapse.
TABLE 21-7 CALCULATION OF FLUID AND ELECTROLYTE INTAKE FOR A

3-KG INFANT WITH 10% ISOTONIC DEHYDRATION
Water
Sodium
Potassium
(mL)
(mEq)
(mEq)
Deficit
300 a
21 b
21 b , c
Maintenance
300 d
Total
600
27
27 c
Total/kg
200
9c
Ongoing losses
Water deficit: 0.10 3 kg.
Electrolyte deficits calculated as in Table 21-5 (14 mEq/kg 3 kg
divided between sodium and potassium).

c
Potassium deficit should be replaced slowly over 48 to 72 hours.
Maintenance water requirement assumed to be 100 mL/kg/day.
Excessive water intake leads to increased excretion of dilute urine. If these

compensatory mechanisms are overtaxed, water will be retained in the
body, resulting in edema and weight gain. Rapid overhydration may produce
congestive heart failure and pulmonary edema, particularly in ill infants
with cardiopulmonary disorders. Even gradual daily administration of excess
water (i.e., volumes larger than can be readily eliminated by the kidneys)
may increase the risk of heart failure from PDA in premature infants (19).
If appropriate fluid therapy is being provided, body weight should be stable

or slowly increasing after the first week of life, and there should be no
evidence of dehydration or fluid overload. During the first week of life, loss
of as much as 15% of body weight is considered normal, provided urine
output is adequate and there is no acidosis or evidence of dehydration. This
physiologic weight loss is at least partly the result of postnatal
contraction of the extracellular water volume. Efforts to prevent this weight
loss are ill advised and may lead to fluid overload, edema, and significant
PDA. Plotting daily weights on a standard postnatal growth chart (97) may
help in detecting excessive weight loss or inappropriate weight gain.
Routine measurement of serum electrolyte concentrations is the best way to
monitor the body's electrolyte status and the adequacy or excess of
electrolyte intake. It is not necessary to add sodium to the parenterally
administered fluid during the first 24 hours of life; in fact, starting sodium
on the day of birth may delay the physiologic contraction of the
extracellular water compartment (108,109). On the other hand, adding
some sodium as bicarbonate or acetate facilitates the use of permissive
hypercapnia to reduce lung trauma while still maintaining acceptable
arterial pH. In any case, it is advisable to determine the serum electrolyte
concentrations soon after birth in infants who require parenteral fluids; if
the mother received sodium-free fluid during labor, the infant may be
hyponatremic (25,26,27,28) and require immediate addition of sodium to
the prescribed fluid. If not required earlier, sodium in the form of sodium
chloride or sodium acetate should be administered at a dose of 2 to 4
mEq/kg/day beginning on the second day of life. Use of the acetate salt
helps to correct the metabolic acidosis sometimes seen in young premature
infants as a result of renal immaturity; it also helps the kidneys in buffering
the respiratory acidosis resulting from permissive hypercapnia.
During the first week of life, infants are usually in negative sodium balance
as a result of the mobilization of sodium along with water from the
extracellular compartment. This negative balance should be allowed as long
as the serum sodium concentration remains normal. In the second and third
weeks of life, small premature infants may require more sodium to replace
the large amounts that are lost in the urine (105,106,107). It is important
to understand that most cases of hyponatremia or hypernatremia result
from excessive or inadequate intake of water.
Potassium supplementation (2 mEq/kg/day) may be started after the infant

urinates unless the serum potassium concentration is elevated. In most
infants, potassium should be added to the infused fluids by the second day.
In very small, critically ill premature infants, however, it is wise to wait
until the serum potassium concentration falls below 4 mEq/L before
administering potassium; these infants are at increased risk of
hyperkalemia due to catabolism and release of potassium from cells in
combination with decreased renal potassium excretion
(110,111,112,113,114). Another mechanism proposed for the hyperkalemia
often seen in extremely premature infants is the movement of potassium
from cells to the extracellular compartment as a result of decreased
sodium-potassium-ATPase activity in erythrocytes (115).
Figure 21-7 Plasma creatinine levels of premature infants during the first 3
months of life. (From Stonestreet BS, Oh W. Plasma creatinine levels in
low-birth-weight infants during the first three months of life. Pediatrics
1978;61:788-789, with permission.)
P.372
Hyperkalemia is the most common life-threatening electrolyte disturbance

in newborn infants. If an elevated serum potassium level is reported by the
laboratory, it should not be attributed to in vitro hemolysis without
verification of the result with another blood sample obtained with good
technique to minimize hemolysis. If the serum potassium concentration is
higher than 6 mEq/L, only potassium-free fluids should be administered. If
the serum potassium concentration is above 7 mEq/L, rectal administration
of sodium polystyrene sulfonate (Kayexalate) or another potassium-binding
resin should be considered. Binding resins, if used, should be given
carefully to limit the risk of colonic necrosis and perforation (116,117,118).
If a cardiac arrhythmia occurs in the presence of hyperkalemia, calcium,
bicarbonate, and insulin with glucose should be given to force potassium
into body cells. In small premature infants, hyperkalemic arrhythmias
include sinus bradycardiaespecially if it occurs without hypoxiaand
ventricular tachycardia.
Blood urea nitrogen and serum or plasma creatinine levels (Fig. 21-7) (119)
are useful for assessing renal function, although the urea nitrogen may also
be elevated with dehydration.
ACID-BASE BALANCE
The physiologic buffer systemprimarily bicarbonate and its weak acid
counterpart carbonic acidand the renal and respiratory compensatory
systems are the major mechanisms responsible for the maintenance of
normal acid-base equilibrium in the body fluids. Changes in the hydrogen
ion concentration in body fluids are governed by the Henderson-Hasselbach
equation:
in which 6.1 is the pK or dissociation constant for carbonic acid (120) and
[H 2 CO 3 ] is the concentration of carbonic acid. It can be seen from this
equation that increase or decrease in the bicarbonate (HCO 3 - ) concentration
results in metabolic alkalosis or metabolic acidosis, respectively. Because
H 2 CO 3 is interchangeably linked to the partial pressure of carbon dioxide
(PCO 2 ) under the influence of carbonic anhydrase, any alteration in PCO 2 in
body fluid also alters pH. Thus, hyperventilation, by reducing PCO 2 ,
produces respiratory alkalosis, and hypoventilation, by increasing PCO 2 ,
causes respiratory acidosis.
Metabolic alkalosis occurs with pyloric stenosis because persistent vomiting

results in the loss of hydrochloric acid and thus produces a relative excess
of bicarbonate. Meta-bolic acidosis is most commonly seen as a
consequence of lactic acid accumulation from anaerobic metabolism in
hypoxic infants. Respiratory alkalosis may occur in an infant who is
hyperventilated with a mechanical ventilator. Respiratory acidosis occurs as
a result of hypercapnia in infants with respiratory distress syndrome or
other pulmonary disease.
In all types of acid-base disturbance, compensation by the lungs or kidneys
occurs to restore the pH toward 7.4. Metabolic acid-base disorders are
corrected by change in ventilation, and respiratory disorders are
compensated for by change in renal bicarbonate excretion. If the
compensation is adequate to correct the pH to normal, the acid-base
disturbance is said to be compensated. For example, if an infant with
respiratory distress syndrome has a pH of 7.38, PCO 2 of 32 torr,
bicarbonate concentration of 18 mEq/L, and base excess of -5 mEq/L, the
acid-base status is called compensated metabolic acidosis. If the same
infant has a pH of 7.38, PCO 2 of 50 torr, bicarbonate concentration of 29
mEq/L, and base excess of +3 mEq/L, the acid-base status is compensated
respiratory acidosis.
FLUID AND ELECTROLYTE PROBLEMS

ASSOCIATED WITH SPECIFIC CLINICAL
CONDITIONS
Extreme Prematurity
Infants born at gestational ages younger than 26 weeks present special
problems in fluid and electrolyte management. These infants have large IWL
(36,37), in some cases more than 200 mL/kg/day if the infant is nursed
under a radiant warmer. The large IWL of extremely premature infants
results from their meager skin barrier to evaporation and their large ratio of
surface area to body weight. The average water requirement for an infant
with birth weight between 500 and 750 grams in the first week of life is
estimated to be about 170 mL/kg/day if the infant is in an incubator with
low ambient humidity and without phototherapy and perhaps 150 mL/kg/day
if moderate humidity is added to the incubator. The requirement would be
210 to 220 mL/kg/day if a radiant warmer or overhead phototherapy is used

and 250 to 270 mL/kg/day if both radiant warmer and photo therapy are
used (121). In the first week of life, the water requirement may be even
higher in infants born after only 22 or 23 weeks of gestation. In these
extremely premature infants, the cutaneous IWL and consequently the total
water requirement decrease toward the end of the first week
P.373
of life, as the stratum corneum matures and becomes less permeable to
water.
Extremely premature infants, unless hyponatremic from maternal
hypotonicity (25,26,27,28), should initially be started on electrolyte-free
solutions of glucose (5 g/dL) in water or solutions that provide sodium in a
dose of 2 to 3 mEq/kg/day as sodium acetate or sodium bicarbonate. The
latter bases compensate for respiratory acidosis while allowing hypercapnia
with the goal of minimizing lung trauma. Administering some or all of the
maintenance sodium as sodium acetate or sodium bicarbonate also serves
to mitigate the metabolic acidosis present in many extremely premature
infants as a result of renal immaturity and transient renal tubular acidosis.
If sodium is not begun initially, it should be added by the second day,
provided the serum sodium concentration is below 145 mEq/L.
Hyperkalemia is a very common problem in extremely premature infants
during the first week of life. This potentially dangerous condition results
from potassium release from catabolized cells in the presence of immature
distal renal tubular function (110,112,113). Hyperkalemia is exacerbated if
dehydration and oliguria occur as a result of inadequate water intake. No
potassium should be given to extremely premature infants until the serum
potassium concentration falls below 4 mEq/L. If the serum potassium
concentration exceeds 7 mEq/L, rectal administration of a potassiumbinding resin should be considered.
Respiratory Distress Syndrome and

Bronchopulmonary Dysplasia
The renal function of infants with respiratory distress syndrome (RDS) is
similar to that of infants of the same gestational age without respiratory
distress, provided their cardiorespiratory status is not significantly
compromised (122,123,124). If infants with RDS become hypoxic and

acidotic, however, they may have decreased glomerular filtration, reduced
renal blood flow, and lower renal bicarbonate threshold (123,124,125,126).
It has been observed that clinical improvement in infants with RDS is
accompanied by an increase in urine volume that occurs on the second and
third days of life (127) and, at the same time, by a rise in circulating levels
of atrial natriuretic peptide (13). It is not clear whether this diuresis causes
or is caused by the improving pulmonary function, or whether each
enhances the other. These associated events may, in fact, be causally
unrelated because this is the time when all infants, especially premature
infants, are in negative water balance (128,129) and experience contraction
of the extracellular water.
In addition to any changes in renal function that may occur with pulmonary
dysfunction resulting from severe RDS, positive-pressure ventilation may
also cause water retention through effects on renal function. Positivepressure ventilation can impair water and sodium excretion by mechanisms
that are not well defined but probably include increased aldosterone
secretion (130) and increased production of antidiuretic hormone
(131,132). It is also possible for impaired filtration and solute excretion to
result from decreased aortic pressure and renal perfusion if excessive mean
airway pressure is used during ventilatory support. Infants with RDS and
other pulmonary disorders may have increased secretion of antidiuretic
hormone, especially if they develop pneumothorax (64,133,134). When
carefully used, positive-pressure ventilation, either conventional or highfrequency ventilation, does not impair diuresis or cause water retention
(135).
PDA is a problem that affects premature infants; those with RDS are more
likely to develop cardiorespiratory signs related to PDA than those without
RDS. Infants receiving carefully limited intake of water are less likely to
develop significant PDA (19,21,22).
RDS is commonly associated with a combined respiratory and metabolic
acidosis resulting from hypercapnia and lactic acidemia. In severe RDS,
when the acidosis is primarily respiratory, assisted ventilation should be
instituted. If the acidosis is primarily metabolic, the underlying cause
should be identified and corrected; if this cannot be done, sodium
bicarbonate may be given to correct the acidosis. There is little evidence to
support the routine use of bicarbonate in infants with RDS

(21,136,137,138). Therefore, the current approach is to treat only
significant metabolic acidosis, generally if the pH is below 7.25. When
bicarbonate is used, the dose should be calculated from the following
equation:
The 0.5 value in the equation is the volume of distribution (i.e., bicarbonate
space), which is confined mostly to the extracellular fluid compartment (Fig.
21-3). There is some disagreement regarding the true bicarbonate space;
the reported values range from 0.3 to 0.6 L/kg. The 0.5 value in the
preceding equation applies to term infants; 0.6 L/kg is a more appropriate
value for premature infants because of their larger volume of extracellular
water (1).
The calculated bicarbonate dose should be diluted to a concentration of 0.5
mEq/mL and given intravenously at a rate no faster than 1 mEq/kg/minute.
A slow infusion over 30 to 60 minutes is preferable, especially for
premature infants, because it may reduce the risk of rapid osmolar changes
in the intravascular and interstitial fluid compartments (139), which have
been associated with intracranial hemorrhage in infants receiving
bicarbonate infusions (140).
Infants with RDS or BPD are sometimes treated with dexamethasone or
other corticosteroid hormones in an attempt to ameliorate the severity of
their lung disease. During dexamethasone therapy, infants may have
impaired glucose tolerance, growth failure, and elevated blood urea
nitrogen resulting from the catabolic effects of dexamethasone
(141,142,143). Dexamethasone also perturbs the balance of phosphorus
and potassium by increasing urinary excretion of these elements (144).
Perinatal Asphyxia
Infants with hypoxia or ischemia of the brain and kidneys during the
perinatal period may suffer brain or kidney
P.374
injury. Increased secretion of AVP often accompanies hypoxic
encephalopathy (64,145,146). Moreover, acute renal failure may result from
renal ischemia in these infants (147,148). Both conditions cause oliguria

and therefore decrease the need for exogenous water. After birth asphyxia,
it is advisable to restrict water intake in anticipation of possible increased
AVP secretion (also known as inappropriate antidiuretic hormone secretion)
or acute renal failure. During the first 24 hours of life, the water intake for
asphyxiated infants should be limited to IWL plus urine output minus about
20 mL/kg/day to allow for some physiologic contraction of the extracellular
water volume. If urine production is normal by the third postnatal day,
water intake can be restored to a normal level.
During the oliguric phase of acute renal failure, potassium should not be
administered unless the serum potassium concentration is less than 3.5
mEq/L. With acute renal failure resulting from hypoxia or ischemia, the
initial period of oliguria may be followed by a diuretic phase with polyuria.
If urine output increases and body weight falls below the expected level
(i.e., the weight before fluid retention began), the intake of water must be
increased to prevent dehydration. This diuretic phase may be accompanied
by large losses of sodium and other electrolytes, which must be replaced.
Replacement of the urinary sodium losses is facilitated by measuring the
volume and sodium concentration in an aliquot of urine from a timed
collection.
Central Nervous System Injury

Infants with brain injury from other causes, such as intracranial
hemorrhage or central nervous system infection, may also have oliguria and
water retention because of inappropriately increased secretion of AVP
(64,145).
Sepsis and Necrotizing Enterocolitis

Systematic reviews of randomized clinical trials of varying fluid intake in
premature infants have identified overhydration as a risk factor in the
pathogenesis of NEC (20,21,22). Infants with septicemia who also have
meningitis may develop inappropriate AVP secretion, reducing their water
requirement. Infants with septicemia or NEC may also develop shock from
endotoxin production or from hypovolemia due to loss of intravascular water
and protein to the interstitial and peritoneal spaces or to the intestinal
lumen, or from frank hemorrhage resulting from thrombocytopenia,

disseminated intravascular coagulopathy, or intestinal injury. In infants
with shock, it is essential to replace lost water and solutes in the form of
blood products or other solute-containing fluids. Even if properly treated,
septic shock may cause renal injury, which then further complicates fluid
and electrolyte management.
Pyloric Stenosis
With pyloric stenosis, water, electrolytes, and hydrogen ions are lost from
the stomach as a result of repeated vomiting of gastric contents. Infants
with severe vomiting caused by pyloric stenosis have elevated intracellular
concentrations of sodium and decreased concentrations of potassium.
Infants with pyloric stenosis are likely to be dehydrated and may have
hypochloremic metabolic alkalosis and hypokalemia. The alkalosis may
cause lethargy, hypoventilation, and, in severe cases, tetany.
Parenteral fluid therapy consists of replacing the deficits of water,
potassium, and chloride. The chloride should initially be given as sodium
chloride. Potassium chloride should be added after adequate urination has
been established. Specific treatment of the metabolic alkalosis with acidic
agents is not necessary. In most cases, control of vomiting, correction of
dehydration, and replacement of chloride and potassium deficits will restore
the blood acid-base status to normal.
Abdominal Wall Defects

Infants born with abdominal wall defects (gastroschisis or omphalocele)
have increased IWL from the exposed viscera and require extra water prior
to surgical repair. This is especially true of infants with gastroschisis
because there is no membrane covering the extruding organs. Once the
abdominal wall has been closed, evaporative water loss is no longer
increased, although there may still be abnormal loss of water into the
peritoneal cavity or interstitial fluid compartment.
Diarrhea
The principles of parenteral fluid therapy of diarrheal dehydration in the
newborn infant are similar to those applied to older infants and children.
Because of their limited renal concentrating ability, newborn infants are
quicker to develop severe dehydration, hypovolemia, and cardiovascular

collapse. Therefore, rapid establishment of intravenous access for vascular
expansion is of the utmost importance in newborn infants with moderate to
severe dehydration from diarrhea. After stabilization, fluid and electrolyte
deficits should be estimated as described earlier in this chapter. Water and
electrolytes should be given to correct established deficits, meet
maintenance requirements, and counteract ongoing losses.
Metabolic acidosis is a frequent finding in diarrheal dehydration. During
initial volume reexpansion, preexisting acidosis may worsen as the body
bicarbonate is further diluted with bicarbonate-free replacement fluids.
Prerenal azotemia is common with diarrheal dehydration. It usually corrects
spontaneously within several days as the infant is rehydrated.
Oral fluids and feeding should be withheld during recovery from diarrheal
dehydration. Oral rehydration has been used successfully in older children
with diarrhea, but there has been less experience with this technique in
newborn infants (149). For this reason, oral rehydration is not
recommended for newborn infants in areas where adequate personnel and
supplies are available
P.375
to maintain parenteral infusions. However, oral rehydration provides an
important alternate therapy when intravenous therapy is not feasible for
economic or technical reasons. The appropriate period of fasting for
diarrhea depends on the severity and duration of the diarrheal episode. As a
rule, a more severe, protracted bout requires a longer period of fasting.
How long diarrheal stools persist following the onset of fasting also helps to
determine the duration of fasting. Reintroduction of oral fluids should be
carried out with extreme care. Aggressive refeeding may precipitate a
recurrence of diarrhea or even protracted chronic diarrhea and
malabsorption with consequent growth failure.
Fluid and Electrolyte Management of Neonatal

Surgical Patients
If a condition that requires surgery (e.g., pyloric stenosis) results in
dehydration and electrolyte or acid-base disturbance, the infant should be
restored as close as possible to normal fluid, electrolyte, and acid-base

status before surgery. Otherwise, the risks of anesthesia and surgery might
be increased by dehydration, acidosis, alkalosis, or abnormal serum
potassium concentration.
During surgery, the prescribed fluid and electrolyte therapy should be
continued with the same fluid composition and rate of infusion unless
additional intraoperative losses require replacement. The anesthesiologist
and surgeons should be apprised of the plan for parenteral fluid therapy to
avoid errors resulting from lack of communication. The operative losses of
fluid and blood should be recorded and replaced either during surgery or
shortly thereafter.
During the initial postoperative period, some infants have reduced urine
output as a result of fluid loss from the vascular compartment or increased
secretion of AVP. Therefore, the fluid provided for maintenance may need to
be reduced during the immediate postoperative period. However, this
reduction may be offset by the extra fluid required to replace abnormal
operative and postoperative losses. A system of careful monitoring of fluid
and electrolyte balance is essential in the surgical patient, as in other ill
newborn infants.
TECHNICAL ASPECTS OF PARENTERAL FLUID

THERAPY
Blood Sampling
Sampling of capillary blood by heel puncture is a safe and commonly used
technique in newborn infants. Automated, spring-loaded lancets provide a
safe and simple way to create an incision of standardized length and depth.
Capillary blood is adequate for determination of serum electrolyte and blood
urea nitrogen concentrations because most of these values agree closely
with venous samples. Serum potassium concentration is slightly higher
when obtained by heel puncture, but the difference is minimized if the blood
flows freely from the puncture wound so that little squeezing is required.
Blood flow can be enhanced by prior warming of the foot. Because the
amount of blood that can be obtained by heel puncture is limited, the
neonatology service should be supported by a laboratory in close proximity
to the nurseries that is capable of performing the required analyses on

blood samples of small volume. Such laboratory support is also crucial to
minimize blood transfusions, which are necessitated in part by phlebotomy
blood losses (150).
Blood samples for blood gas and acid-base determination can also be
obtained by heel puncture. Their reliability is enhanced by warming the skin
to 44C, which reduces the arterial-capillary differences in blood gases and
pH; however, oxygen pressure (PO 2 ) measurements on samples obtained by
heel puncture tend to underestimate true arterial PO 2 , especially in the
hyperoxic range. Sufficient information for clinical management can usually
be obtained by using capillary acid-base and PCO 2 values in conjunction
with oxygen saturation values determined by continuous pulse oximetry.
However, indwelling umbilical or radial artery catheters often are used in
infants who are extremely ill and require continuous blood pressure
monitoring or high levels of oxygen and ventilatory support. Because
indwelling arterial catheters are associated with serious thromboembolic
complications, they should be removed as soon as possible. For occasional
sampling of arterial blood, direct puncture and aspiration from a peripheral
artery can be used. This procedure is fairly safe in skilled hands but can
cause serious complications, such as arterial obstruction or nerve injury.
Intravenous Fluid Infusion

Short catheters inserted into peripheral veins in a limb or the scalp can be
used for infusion of water, electrolytes, and nutrient solutions. Hypertonic
and vasoconstrictive solutions should be infused with caution into peripheral
veins. Peripheral infusion of solutions containing dextrose in concentrations
higher than 12.5 g/dL is generally contraindicated. Central venous access, if
available, is preferred for such infusions. Shaving scalp hair to expose veins
is distressing to many parents; the scalp should be used only if extremity
sites have been exhausted and only after the procedure has been discussed
with the parents. Intravenous sites must be carefully inspected on each
nursing shift for evidence of infection at the insertion site or for
extravasation of infusate that might injure subcutaneous tissues.
Long catheters of silicone elastomer or other materials can be inserted
percutaneously so that the tip lies centrally in a large vein within the
abdomen or chest. These catheters can be used for long-term intravenous

access with lower risks of infection and thrombosis than occur with
surgically inserted central vein catheters. Serious complications are rare
with percutaneous central vein catheters. However, infectious complications
are an ever-present threat. In addition, life-threatening complications, such
as
P.376
pleural effusion and cardiac tamponade, are possible (151).
Venous cutdown for vascular access is seldom necessary if skilled personnel
are available to insert and maintain standard intravenous catheters.
Surgical cutdown for venous access risks local infection or septicemia and
permanent venous obstruction. Moreover, catheters inserted by cutdown
usually do not last as long in newborn infants as they do in older children
and adults. For these reasons, venous cutdown should be avoided in
newborns whenever possible.
A catheter in the umbilical vein with its tip in the inferior vena cava may be
used for infusion of fluid, electrolytes, and nutrients during the first few
days, until potential infection can be treated and percutaneous central vein
access can be established if needed. Umbilical vein catheters ordinarily
should be removed within a few days of birth because of the risks of
infection, portal phlebitis, and liver damage. Umbilical vein catheters can
also be used for central venous pressure monitoring and exchange
transfusions. Except in unusual circumstances, an umbilical vein catheter
should be removed by the end of the first week of life, when the risks of
infection and thrombosis probably surpass the advantages of continued
catheter use.
With any route of parenteral fluid administration to newborn infants, it is
essential to use infusion pumps that can be precisely regulated and can
deliver fluids at an absolutely steady rate, even when very low flow rates
are used. A fluctuating infusion rate may cause erratic delivery of water,
glucose and other nutrients, or drugs.
REFERENCES
1. Friis-Hansen B. Changes in body water compartments during growth.
Acta Paediatr 1957;46(Suppl 110):1-68.
2. Friis-Hansen B. Body water compartments in children: changes during
growth and related changes in body composition. Pediatrics 1961;28:169181.
3. Ziegler EE, O'Donnell AM, Nelson SE, et al. Body composition of the
reference fetus. Growth 1976;40:329-341.
4. Forbes GB, Perley A. Estimation of total body sodium by isotopic dilution.
II. Studies on infants and children: an example of a constant differential
growth ratio. J Clin Invest 1951;30:566-574.
5. Cheek DB. Observations on total body chloride in children. Pediatrics
1954;14:5-10.
6. Romahn A, Burmeister W. [Body composition during the first two years of
life: analysis with the potassium 40 method]. Klin Pdiatr 1977;189:321327.
7. Shaffer SG, Bradt SK, Hall RT. Postnatal changes in total body water and
extracellular volume in the preterm infant with respiratory distress
syndrome. J Pediatr 1986;109:509-514.
8. Bauer K, Bovermann G, Roithmaier A, et al. Body composition, nutrition,
and fluid balance during the first two weeks of life in preterm neonates
weighing less than 1500 grams. J Pediatr 1991;118:615-620.
9. Hartnoll G, Btrmieux P, Modi N. Body water content of extremely
preterm infants at birth. Arch Dis Child Fetal Neonatal Ed 2000;83:F56-F59.
10. Cheek DB, Maddison TG, Malinek M, et al. Further observations on the
corrected bromide space of the neonate and investigation of water and
electrolyte status in infants born of diabetic mothers. Pediatrics
1961;28:861-869.
11. Kagan BM, Stanincova V, Felix NS, et al. Body composition of premature
infants: relation to nutrition. Am J Clin Nutr 1972; 25:1153-1164.
12. Heimler R, Doumas BT, Jendrzejczak BM, et al. Relationship between
nutrition, weight change, and fluid compartments in preterm infants during
the first week of life. J Pediatr 1993;122: 110-114.
13. Modi N, Btrmieux P, Midgley J, et al. Postnatal weight loss and

contraction of the extracellular compartment is triggered by atrial
natriuretic peptide. Early Hum Dev 2000;59:201-208.
14. Oh W, Oh MA, Lind J. Renal function and blood volume in newborn
infant related to placental transfusion. Acta Paediatr Scand 1966;55:197210.
15. Aperia A, Broberger O, Elinder G, et al. Postnatal development of renal
function in pre-term and full-term infants. Acta Paediatr Scand
1981;70:183-187.
16. Horster M. Embryonic epithelial membrane transporters. Am J Physiol
Renal Physiol 2000;279:F982-F996.
17. Cassady G, Milstead RR. Antipyrine space studies and cell water
estimates in infants of low birth weight. Pediatr Res 1971;5: 673-682.
18. Stonestreet BS, Bell EF, Warburton D, et al. Renal response in lowbirth-weight neonates. Results of prolonged intake of two different amounts
of fluid and sodium. Am J Dis Child 1983; 137:215-219.
19. Bell EF, Warburton D, Stonestreet BS, et al. Effect of fluid
administration on the development of symptomatic patent ductus arteriosus
and congestive heart failure in premature infants. N Engl J Med
1980;302:598-604.
20. Bell EF, Warburton D, Stonestreet BS, et al. High-volume fluid intake
predisposes premature infants to necrotising enterocolitis. Lancet
1979;2:90.
21. Bell EF. Fluid therapy. In: Sinclair JC, Bracken MB, eds. Effective care
of the newborn infant. Oxford: Oxford University Press, 1992:59-72.
22. Bell EF, Acarregui MJ. Restricted versus liberal water intake for
preventing morbidity and mortality in preterm infants. Cochrane Database
Syst Rev 2001;2:CD000503.
23. Brown ER, Stark A, Sosenko I, et al. Bronchopulmonary dysplasia:
possible relationship to pulmonary edema. J Pediatr 1978; 92:982-984.
24. Van Marter LJ, Leviton A, Allred EN, et al. Hydration during the first
days of life and the risk of bronchopulmonary dysplasia in low birth weight
infants. J Pediatr 1990;116:942-949.
25. Battaglia F, Prystowsky H, Smisson C, et al. Fetal blood studies. XIII.
The effect of the administration of fluids intravenously to mothers upon the
concentrations of water and electrolytes in plasma of human fetuses.

26. Tarnow-Mordi WO, Shaw JC, Liu D, et al. Iatrogenic hyponatraemia of
the newborn due to maternal fluid overload: a prospective study. Br Med J
1981;283:639-642.
27. Grylack LJ, Chu SS, Scanlon JW. Use of intravenous fluids before
cesarean section: effects on perinatal glucose, insulin, and sodium
homeostasis. Obstet Gynecol 1984;63:654-658.
28. Zimmer EZ, Goldstein I, Feldman E, et al. Maternal and newborn levels
of glucose, sodium and osmolality after preloading with three intravenous
solutions during elective cesarean sections. Eur J Obstet Gynecol Reprod
Biol 1986;23:61-65.
29. Hey EN, Katz G. Evaporative water loss in the new-born baby. J Physiol
1969;200:605-619.
30. Sulyok E, Jquier E, Prod'hom LS. Respiratory contribution to the
thermal balance of the newborn infant under various ambient conditions.
Pediatrics 1973;51:641-650.
31. Sosulski R, Polin RA, Baumgart S. Respiratory water loss and heat
balance in intubated infants receiving humidified air. J Pediatr
1983;103:307-310.
32. Fanaroff AA, Wald M, Gruber HS, et al. Insensible water loss in low birth
weight infants. Pediatrics 1972;50:236-245.
33. Wu PY, Hodgman JE. Insensible water loss in preterm infants: changes
with postnatal development and non-ionizing radiant energy. Pediatrics
1974;54:704-712.
34. Shaffer SG, Weismann DN. Fluid requirements in the preterm infant.
Clin Perinatol 1992;19:233-250.
35. Okken A, Jonxis JH, Rispens P, et al. Insensible water loss and
metabolic rate in low birthweight newborn infants. Pediatr Res
1979;13:1072-1075.
36. Hammarlund K, Sedin G. Transepidermal water loss in newborn infants.
III. Relation to gestational age. Acta Paediatr Scand 1979;68:795-801.
37. Costarino AT, Baumgart. Controversies in fluid and electrolyte therapy
for the premature infant. Clin Perinatol 1988;15:863- 878.
P.377
38. Hooper JM, Evans IW, Stapleton T. Resting pulmonary water loss in the
newborn infant. Pediatrics 1954;13:206-210.
39. Rutter N, Hull D. Response of term babies to a warm environment. Arch
Dis Child 1979;54:178-183.
40. Bell EF, Gray JC, Weinstein MR, et al. The effects of thermal
environment on heat balance and insensible water loss in low-birth-weight
41. Williams PR, Oh W. Effects of radiant warmer on insensible water loss in
newborn infants. Am J Dis Child 1974;128:511-514.
42. Jones RW, Rochefort MJ, Baum JD. Increased insensible water loss in
newborn infants nursed under radiant heaters. Br Med J 1976;2:1347-1350.
43. Bell EF, Neidich GA, Cashore WJ, et al. Combined effect of radiant
warmer and phototherapy on insensible water loss in low-birth-weight
44. Bell EF, Weinstein MR, Oh W. Heat balance in premature infants:
comparative effects of convectively heated incubator and radiant warmer,
with and without plastic heat shield. J Pediatr 1980; 96:460-465.
45. Kjartansson S, Arsan S, Hammarlund K, et al. Water loss from the skin
of term and preterm infants nursed under a radiant heater. Pediatr Res
1995;37:233-238.
46. Oh W, Karecki H. Phototherapy and insensible water loss in the newborn
infant. Am J Dis Child 1972;124:230-232.
47. Grnhagen DJ, de Boer MG, de Beaufort AJ, et al. Transepider-mal
water loss during halogen spotlight phototherapy in preterm infants. Pediatr
Res 2002;51:402-405.
48. Engle WD, Baumgart S, Schwartz JG, et al. Insensible water loss in the
critically ill neonate. Combined effct of radiant-warmer power and
phototherapy. Am J Dis Child 1981;135:516-520.
49. Day R. Respiratory metabolism in infancy and in childhood: XXVII.
Regulation of body temperature of premature infants. Am J Dis Child
1943;65:376-398.
50. Zweymller E, Preining O. The insensible water loss of the newborn
infant. Acta Paediatr Scand 1970;205(Suppl):1-29.
51. O'Brien D, Hansen JDL, Smith CA. Effect of supersaturated atmospheres

on insensible water loss in the newborn infant. Pediatrics 1954;13:126-132.
52. Baumgart S, Fox WW, Polin RA. Physiologic implications of two different
heat shields for infants under radiant warmers. J Pediatr 1982;100:787790.
53. Marks KH, Friedman Z, Maisels MJ. A simple device for reducing
insensible water loss in low-birth-weight infants. Pediatrics 1977;60:223226.
54. Baumgart S, Engle WD, Fox WW, et al. Effect of heat shielding on
convective and evaporative heat losses and on radiant heat transfer in the
premature infant. J Pediatr 1981;99:948-956.
55. Fitch CW, Korones SB. Heat shield reduces water loss. Arch Dis Child
1984;59:886-888.
56. Knauth A, Gordin M, McNelis W, et al. Semipermeable polyurethane
membrane as an artificial skin for the premature neonate. Pediatrics
1989;83:945-950.
57. Vernon HJ, Lane AT, Wischerath LJ, et al. Semipermeable dressing and
transepidermal water loss in premature infants. Pediatrics 1990;86:357362.
58. Mancini AJ, Sookdeo-Drost S, Madison KC, et al. Semipermeable
dressings improve epidermal barrier function in premature infants. Pediatr
Res 1994;36:306-314.
59. Rutter N, Hull D. Reduction of skin water loss in the newborn. I. Effect
of applying topical agents. Arch Dis Child 1981;56:669-672.
60. Nopper AJ, Horii KA, Sookdeo-Drost S, et al. Topical ointment therapy
benefits premature infants. J Pediatr 1996;128:660- 669.
61. Kjartansson S, Hammarlund K, Sedin G. Insensible water loss from the
skin during phototherapy in term and preterm infants. Acta Paediatr
1992;81:764-768.
62. Kjartansson S, Hammarlund K, Riesenfeld T, et al. Respiratory water
loss and oxygen consumption in newborn infants during phototherapy. Acta
Paediatr 1992;81:769-773.
63. Leung AK, McArthur RG, McMillan DD, et al. Circulating antidiuretic
hormone during labour and in the newborn. Acta Paediatr Scand
1980;69:505-510.
64. Wiriyathian S, Rosenfeld CR, Arant BS Jr, et al. Urinary arginine

vasopressin: pattern of excretion in the neonatal period. Pediatr Res
1986;20:103-108.
65. Marchini G, Stock S. Thirst and vasopressin secretion counteract
dehydration in newborn infants. J Pediatr 1997;130:736-739.
66. Rees L, Brook CGD, Shaw JC, et al. Hyponatraemia in the first week of
life in preterm infants. Part I. Arginine vasopressin secretion. Arch Dis Child
1984;59:414-422.
67. Siegel SR, Fisher DA, Oh W. Serum aldosterone concentrations related
to sodium balance in the newborn infant. Pediatrics 1974;53:410-413.
68. Dillon MJ, Rajani KB, Shah V, et al. Renin and aldosterone response in
human newborns to acute change in blood volume. Arch Dis Child
1978;53:461-467.
69. Aperia A, Broberger O, Herin P, et al. Sodium excretion in relation to
sodium intake and aldosterone excretion in newborn pre-term and full-term
infants. Acta Paediatr Scand 1979;68: 813-817.
70. Sulyok E, Nmeth M, Tnyi I, et al. Postnatal development of reninangiotensin-aldosterone system, RAAS, in relation to electrolyte balance in
premature infants. Pediatr Res 1979;13:817-820.
71. Kojima T, Fukuda Y, Hirata Y, et al. Effects of aldosterone and atrial
natriuretic peptide on water and electrolyte homeostasis of sick neonates.
Pediatr Res 1989;25:591-594.
72. David L, Anast CS. Calcium metabolism in newborn infants. The
interrelationship of parathyroid function and calcium, magnesium, and
phosphorus metabolism in normal, sick, and hypocalcemic infants. J Clin
Invest 1974;54:287-296.
73. David L, Salle B, Chopard P, et al. Studies on circulating
immunoreactive calcitonin in low birth weight infants during the first 48
hours of life. Helv Paediatr Acta 1977;32:39-48.
74. Hillman LS, Rojanasathit S, Slatopolsky E, et al. Serial measurements of
serum calcium, magnesium, parathyroid hormone, calcitonin, and 25hydroxy-vitamin D in premature and term infants during the first week of
life. Pediatr Res 1977;11:739-744.
75. David L, Salle BL, Putet G, et al. Serum immunoreactive calcitonin in
low birth weight infants. Description of early changes; effect of intravenous
calcium infusion; relationships with early changes in serum calcium,
phosphorus, magnesium, parathyroid hormone, and gastrin levels. Pediatr

Res 1981;15:803-808.
76. Smith FG, Sato T, Varille VA, et al. Atrial natriuretic factor during fetal
and postnatal life: a review. J Dev Physiol 1989;12:55-62.
77. Mercadier JJ, Zongazo MA, Wisnewsky C, et al. Atrial natriuretic factor
messenger ribonucleic acid and peptide in the human heart during ontogenic
development. Biochem Biophys Res Commun 1989;159:777-782.
78. Shaffer SG, Geer PG, Goetz KL. Elevated atrial natriuretic factor in
neonates with respiratory distress syndrome. J Pediatr 1986;109:10281033.
79. Liechty EA, Johnson MD, Myerberg DZ, et al. Daily sequential changes in
plasma atrial natriuretic factor concentrations in mechanically ventilated
low-birth-weight infants. Effect of surfactant replacement. Biol Neonate
1989;55:244-250.
80. Rozycki HJ, Baumgart S. Atrial natriuretic factor and postnatal diuresis
in respiratory distress syndrome. Arch Dis Child 1991;66:43-47.
81. Robillard JE, Weiner C. Atrial natriuretic factor in the human fetus:
effect of volume expansion. J Pediatr 1988;113:552-555.
82. Bierd TM, Kattwinkel J, Chevalier RL, et al. Interrelationship of atrial
natriuretic peptide, atrial volume, and renal function in premature infants. J
Pediatr 1990;116:753-759.
83. Tulassay T, Rascher W, Seyberth HW, et al. Role of atrial natriuretic
peptide in sodium homeostasis in premature infants. J Pediatr
1986;109:1023-1027.
84. Kojima T, Hirata Y, Fukuda Y, et al. Plasma atrial natriuretic peptide
and spontaneous diuresis in sick neonates. Arch Dis Child 1987;62:667-670.
85. Siegel SR, Oh W. Renal function as a marker of human fetal maturation.
Acta Paediatr Scand 1976;65:481-485.
86. Leake RD, Trygstad CW, Oh W. Inulin clearance in the newborn infant:
relationship to gestational and postnatal age. Pediatr Res 1976;10:759-762.
87. Arant BS Jr. Developmental patterns of renal functional maturation
compared in the human neonate. J Pediatr 1978;92:705-712.
88. Ross B, Cowett RM, Oh W. Renal functions of low birth weight infants
during the first two months of life. Pediatr Res 1977;11: 1162-1164.
89. Wilkins BH. Renal function in sick very low birthweight infants: 1.
Glomerular filtration rate. Arch Dis Child 1992;67: 1140-1145.
90. Wilkins BH. Renal function in sick very low birthweight infants: 2. Urea
and creatinine excretion. Arch Dis Child 1992;67:1146-1153.
P.378
91. Hansen JD, Smith CA. Effects of withholding fluid in the immediate
postnatal period. Pediatrics 1953;12:99-113.
92. Calcagno PL, Rubin MI, Weintraub DH. Studies on the renal
concentrating and diluting mechanisms in the premature infant. J Clin
Invest 1954;33:91-96.
93. McCance RA, Naylor NJ, Widdowson EM. The response of infants to a
large dose of water. Arch Dis Child 1954;29:104-109.
94. Leake RD, Zakauddin S, Trygstad CW, et al. The effects of large volume
intravenous fluid infusion on neonatal renal function. J Pediatr
1976;89:968-972.
95. Aperia A, Herin P, Lundin S, et al. Regulation of renal water excretion in
newborn full-term infants. Acta Paediatr Scand 1984;73:717-721.
96. Brosius KK, Ritter DA, Kenny JD. Postnatal growth curve of the infant
with extremely low birth weight who was fed enterally. Pediatrics
1984;74:778-782.
97. Shaffer SG, Quimiro CL, Anderson JV, et al. Postnatal weight changes in
low birth weight infants. Pediatrics 1987;79:702- 705.
98. Lemoh JN, Brooke OG. Frequency and weight of normal stools in
infancy. Arch Dis Child 1979;54:719-720.
99. Patrick CH, Pittard WB. Stool water loss in very-low-birth-weight
neonates. Clin Pediatr (Phila) 1988;27:144-146.
100. Williams GS, Klenk EL, Winters RW. Acute renal failure in pediatrics.
In: Winters RW, ed. The body fluids in pediatrics, medical, surgical, and
neonatal disorders of acid-base status, hydration, and oxygenation. Boston:
Little, Brown and Company, 1973: 523-557.
101. Gamble JL, Butler AM. Measurement of the renal water requirement.
Trans Assoc Am Phys 1944;58:157-161.
102. Sinclair JC, Driscoll JM Jr, Heird WC, et al. Supportive management of
the sick neonate. Parenteral calories, water, and electrolytes. Pediatr Clin
North Am 1970;17:863-893.
103. Ziegler EE. Feeding the low birth weight infant. In: Gellis SS, Kagan
BM, eds. Current pediatric therapy, 13th ed. Philadelphia: WB Saunders,
1990:713-716.
104. American Academy of Pediatrics Committee on Nutrition. Nutritional
needs of the preterm infant. In: Kleinman RE, ed. Pediatric nutrition
handbook, 5th ed. Elk Grove Village, Ill: American Academy of Pediatrics,
2004:23-54.
105. Roy RN, Chance GW, Radde IC, et al. Late hyponatremia in very low
birthweight infants (<1.3 kilograms). Pediatr Res 1976;10: 526-531.
106. Engelke SC, Shah BL, Vasan U, et al. Sodium balance in very lowbirth-weight infants. J Pediatr 1978;93:837-841.
107. Al-Dahhan J, Haycock GB, Chantler C, et al. Sodium homeostasis in
term and preterm neonates. I. Renal aspects. Arch Dis Child 1983;58:335342.
108. Costarino AT Jr, Gruskay JA, Corcoran L, et al. Sodium restriction
versus daily maintenance replacement in very low birth weight premature
neonates: a randomized, blind therapeutic trial. J Pediatr 1992;120:99-106.
109. Hartnoll G, Btrmieux P, Modi N. Randomized controlled trial of
postnatal sodium supplementation on body composition in 25 to 30 week
gestational age infants. Arch Dis Child Fetal Neonatal Ed 2000;82:F24-F28.
110. Gruskay J, Costarino AT, Polin RA, et al. Nonoliguric hyperkalemia in
the premature infant weighing less than 1000 grams. J Pediatr
1988;113:381-386.
111. Brion LP, Schwartz GJ, Campbell D, et al. Early hyperkalaemia in very
low birthweight infants in the absence of oliguria. Arch Dis Child
1989;64:270-272.
112. Shaffer SG, Kilbride HW, Hayen LK, et al. Hyperkalemia in very low
birth weight infants. J Pediatr 1992;121:275-279.
113. Sato K, Kondo T, Iwao H, et al. Internal potassium shift in premature
infants: cause of nonoliguric hyperkalemia. J Pediatr 1995;126:109-113.
114. Lorenz JM, Kleinman LI, Markarian K. Potassium metabolism in
extremely low birth weight infants in the first week of life. J Pediatr
1997;131:81-86.
115. Stefano JL, Norman ME, Morales MC, et al. Decreased erythrocyte Na + ,
K + -ATPase activity associated with cellular potassium loss in extremely low
birth weight infants with nonoliguric hyperkalemia. J Pediatr 1993;122:276
Chapter 22
Nutrition
Michael K. Georgieff
The provision of nutrition to term and preterm newborn infants remains one of the
most important aspects of neonatal care. With increasing survival rates among
sick newborns, the nourishment of full-term and preterm infants has assumed an
increasingly greater role in the neonatal intensive care unit in the past 25 years.
Great strides have been made in understanding neonatal nutritional physiology
and pathophysiology in these years, allowing physicians to more precisely
estimate the nutritional needs of the infants in their care. Knowledge of newborn
infants' nutritional requirements and of their neurological, gastrointestinal, and
metabolic capabilities is a prerequisite to informed decision making about
nutritional therapy in the nursery. It is also important to understand the tools
available for assessment of neonatal nutritional status to judge the success or
failure of nutritional therapies.
The goal of nutritional therapy in the term neonate is to ensure a successful
growth transition from the fetal to the postnatal period. In the preterm infant, the
goal has been to continue the process of intrauterine growth in what is now an
extrauterine environment until 40 weeks postconception and to foster catch-up
growth and nutrient accretion in the postdischarge period. Until lately, the goal for
the growing preterm infant has been to match the third trimester intrauterine
rates of weight gain, linear growth, and brain growth. Even if these rates are
successfully attained, the body composition of the preterm infant raised in an
extrauterine environment nevertheless remains remarkably different than that of
the same postconceptional-age infant who has remained in utero (1,2). Current
efforts are aimed at understanding the metabolic processes that determine the
body composition of the preterm infant. Additionally, the preterm infant is still
likely to be well below the standard gestational growth curves at discharge (3)
because of nutrient deficits that have accrued during the prolonged period of
neonatal illness (4). The effect of illness on neonatal metabolism and nutritional
requirements is being recognized (5). Conditions such as bronchopulmonary
dysplasia (BPD), congestive heart failure (CHF), acute respiratory distress,
intrauterine growth retardation, and sepsis (and their treatments) have negative
effects on neonatal energy, protein, and mineral and vitamin requirements. These
conditions may also affect the digestive and absorptive capacities of the neonate.
This chapter reviews the nutritional requirements, digestive capabilities, and
expected growth of term and preterm infants. These factors will be addressed in
the context of the three time phases of neonatal nutritional development:
transition, premie growth, and postdischarge. The chapter also provides an

overview of various nutrient delivery systems, both enteral and parenteral.
Finally, the chapter covers techniques of nutritional assessment and offers
suggestions for appropriate monitoring of neonatal nutritional status.
NUTRITIONAL CAPABILITIES OF THE NEWBORN

INFANT
The ability to suck and swallow a meal in a coordinated fashion and then process
those nutrients for utilization by the body may be one of the most complex
developmental tasks facing the newborn infant (6). Success depends on a certain
amount of neurological, digestive, absorptive and metabolic maturity. The term
infant is quite mature in these respects. However, the preterm infant's physiology
is progressively more immature as a function of decreased gestational age.
Neurological Maturity
The neurologically intact term infant is able to suck and swallow in a coordinated
fashion within minutes of birth (7). In the preterm infant, the sucking reflex is
strong at the limit of viability (23 weeks) and likely prior to that age (8). However,
the ability to coordinate the suck reflex with swallowing to ensure that food is
propelled into the gastrointestinal tract rather than the airway matures at
approximately 34 weeks gestation (9). To a great extent, this coordinated suck
and swallow reflex appears to be postconceptional-age mediated; that is, it does
not appear that practice can stimulate the infant to become more mature at an
earlier postconceptional age. Nevertheless,
P.381
the age at which the infant matures varies widely, and some preterm infants are
able to suck and swallow in a coordinated manner by 32 weeks postconception.
Motility in the gastrointestinal tract is also dependent on neurological maturation
(10). For example, the esophagus shows a very discoordinated pattern of
peristalsis at 24 weeks gestation, with weak peristaltic waves beginning
sporadically and propagating either rostrally or caudally (11). By term, the pattern
has matured into a coordinated pattern that propels food downward to the
stomach (12). The lower esophageal sphincter of the very preterm infant is
tenuous and provides little barrier to gastroesophageal reflux (GER). GER disease
in the preterm infant can be associated with apnea and bradycardia, aspiration
syndromes, and feeding intolerance. At term, although GER remains demonstrable
in most infants, it is generally not the potentially life-threatening problem seen at
earlier gestational ages.
The stomach also undergoes maturation during the third trimester. The preterm
infant's stomach does not coordinately wring the stomach from antrum to
pylorus, frequently being subject to periods of antiperistalsis, which in turn

promotes GER disease (13). Additionally, pyloric function is different in the
preterm compared to the term infant. Gastric emptying is longer in the preterm
infant, although stomach volume is smaller (14). The small stomach capacity
frequently results in the need to feed small volumes to preterm infants on a more
frequent schedule. The prolonged gastric emptying time, however, causes
retention of the food in the stomach and subsequent obtainment of preaspirates
prior to the next feeding. Since the presence of preaspirates may also signify an
ileus associated with more severe diseases such as necrotizing enterocolitis (NEC),
feedings are often discontinued with this sign.
The maturation of small intestinal motility has been extensively studied by
Berseth and her colleagues (15,16,17,18). They have demonstrated an orderly
progression of peristaltic frequency, amplitude, and duration with increasing
gestational age. They have also demonstrated a two-hour periodicity to peristalsis
in very preterm infants (18). This finding may influence decisions about feeding
schedules. The lack of a coordinated motility pattern in the very preterm infant
makes it more likely that they will present with signs of feeding intolerance,
usually marked by abdominal distention. When compared to adults, neonates
retain food in their small intestine for proportionately longer periods and in their
colon for shorter periods. The newborn's ability to modulate stool water and
electrolyte content is immature compared to adults.
In summary, multiple neuromaturational factors work against the preterm infant's
ability to enterally feed as successfully as the term infant (9). These immaturities
have an impact on feeding management, as will be discussed in a later section.
Digestive and Absorptive Capabilities

The newborn infant does not have the capability to digest and absorb nutrients
from a complex diet (9). Fortunately, the term human infant has a ready source of
nutrition in the form of human milk (7). Human milk is remarkably adapted to fit
the digestive capacities of the term newborn and to meet the infant's nutritional
needs for at least the first 6 months of life (7). Recent studies have also shown
the nutritional value of feeding human milk to preterm infants; however, for
infants weighing more than 1,500 g, human milk needs to be fortified (19).
The functional immaturities of the gastrointestinal, hepatic, and renal systems in
the newborn have an impact on the delivery of all classes of nutrients:
macronutrients, minerals, trace elements, and vitamins.
Protein
Protein digestion begins in the stomach with the action of pepsin on the intact
protein (20). Pepsin is activated by acid hydrolysis of its precursor molecule,
pepsinogen. The newborn is capable of creating an acidic stomach environment by

1 week of age (21); thus, pepsin activation is thought to be intact. Dietary protein
is then acted upon by pancreatic peptidases released into the duodenum. These
enzymes include trypsin, chymotrypsin, carboxypeptidases A and B, and elastase,
and are amino acid selective with respect to cleavage sites, resulting in peptides
of relatively small length. The peptides are subsequently cleaved once more by
peptidases located in the intestinal mucosal cells, absorbed as amino acids or
dipeptides, and transported to the liver. Protein digestion and absorption in adults
is very efficient; up to 95% of a protein load can be fully digested. Although term
and preterm infants have relatively low concentrations of chymotrypsin, the
carboxypeptidases, and elastase (22), they nevertheless achieve more than 80%
protein digestion.
Fat
The efficiency of fat digestion in the neonate has been a controversial topic. Data
from the 1970s and 1980s suggested that fat is the most poorly digested
macronutrient in the neonate (23). Whereas adults will absorb close to 95% of a
fat meal and term infants absorb 85% to 90%, early studies indicated that
preterm infants absorb as little as 50%, depending on the type of fat presented to
them (24). The perceived functional immaturity of fat digestion in the preterm
infant led to modification of fat blends in the formulas used in preterm infants.
Fat digestion in the neonate begins in the stomach with the action of a lipase
secreted in the mouth (lingual lipase) or by the gastric mucosa (gastric lipase)
(25). The two lipases are identical, function ideally at acid pH, work primarily on
medium-chain triglycerides (MCT), and do not require bile salts. Hamosh has
estimated that this enzyme may be responsible for up to 50% of fat digestion in
the newborn (26). Infants fed human milk have the additional benefit of a lipase
secreted into the milk by the mother (27). This lipase is found in all carnivores
(but not herbivores) and functions more like pancreatic or intestinal
P.382
lipases found in adults. It works primarily on long-chain triglycerides at a neutral
pH, as is found in the intestine, and requires bile salts. This lipase may be
responsible for the digestion of up to 20% of dietary fat (28). These two lipases
are referred to as the compensatory lipases of the newborn and function in place
of pancreatic and intestinal lipases seen in more mature humans (29).
Long-chain fatty acids are dependent on bile salts for proper micellization and
uptake into the intestinal lymphatics. From there, the micelles are carried to the
venous system via the thoracic duct, ultimately destined for the liver. Mediumchain fatty acids do not require micellization and can be directly absorbed into the
blood stream. The bile acid, and hence bile salt, pools of the preterm newborn are
low, thus restricting the fat-absorption capacity of the infant. Prenatal
administration of glucocorticoids to the mother can mature the fetal bile salt pool
in the preterm infant less than 34 weeks gestation to the level of the term infant
(30). Without such priming, however, the preterm infant has significant
impairment of fat absorption (including fat-soluble vitamins) prior to 34 weeks
gestation. The fat blend in preterm infant formulas designed for infants less than
34 weeks gestation has been significantly modified to optimize fat absorption.
These formulas contain a higher percentage of MCT and higher vitamin A, D, and
E levels than formulas manufactured for term infants.
Carbohydrate
Like fats, carbohydrates can present a significant digestive challenge. The neonate
has a limited ability to digest complex carbohydrates because of relatively small
amounts of pancreatic amylase (31). Thus, beikost in the form of cereal rarely
makes up a significant portion of the infant's diet until after 4 months of age. The
term and preterm newborn readily uses glucose, which can be delivered either
parenterally or enterally. Intestinal glucose uptake is seen as early as ten weeks
gestation, long before the fetus is viable (32). However, provision of all
carbohydrate calories as glucose would result in the neonatal gut being exposed to
a hyperosmolar solution with a high potential for mucosal damage.
The primary carbohydrate found in mammalian milk is the disaccharide lactose.
Like other disaccharides (sucrose, maltose, isomaltose), enzymatic cleavage by a
disaccharidase must occur before the monosaccharides can be absorbed. In the
case of lactose, glucose and galactose are produced by the action of lactase. The
disaccharidases sucrase and maltase appear very early in gestation and appear to
be inducible enzymes (33). In contrast, lactase begins to appear at 24 weeks
gestation and rises in concentration very slowly until term. It does not appear to
be a particularly inducible enzyme (34). The preterm infant is thus functionally
somewhat lactose intolerant and will have typical symptoms of gas formation,
diarrhea, and acidic stools characteristic of lactose malabsorption when fed high
doses of lactose. Positive hydrogen breath tests have been documented in
preterm infants following lactose challenges (35).
Preterm infant formulas have lower lactose contents than term formulas for this
reason. Up to 60% of carbohydrate calories in preterm infant formulas are derived
from linear glucose polymers, which produce a lower osmolar load than the
equivalent number of individual glucose molecules. The enzyme required to digest
glucose polymers (glucoamylase) is present from 24 weeks gestation (36). The
lower lactose content is also present in the premature discharge formulas,
although it is likely that the preterm infant is fully mature with respect to lactose
absorption at the time of discharge (37).
NUTRIENT REQUIREMENTS FOR TERM AND
PRETERM INFANTS
Estimation of nutrient requirements is an inexact process, particularly when the
goal is unclear. To date, the goal has been to achieve the same growth rates and
body composition as the reference infant; the healthy breastfed infant serves as
the gold standard for the term infant. Never-theless, it is clear that breastfed
babies have different growth rates and body compositions than formula-fed
infants (38). Human milk composition varies greatly among mothers, and the
length of time that it remains sufficient for all the nutrient needs of the infant is
not uniform. Breastfed infants may have lower iron stores (39) and be at greater
risk for vitamin D deficiency than formula- fed infants (40).
Determining the ideal growth for the infant born before term is far more
problematic. Indeed, the ideal growth rate and body composition of the healthy
preterm infant remain unknown and are likely to be different from his or her
gestationally age-matched fetal counterpart. Until recently, the daily and weekly
accretion rates of various nutrients in the preterm infant have been modeled on in
utero accretion rates of these nutrients in gestationally age-matched fetuses. The
reference fetus described by Widdowson and again by Ziegler has served as the
benchmark by which neonatal nutritionists judge fetal growth and body
composition (2,41). Nevertheless, energy requirements are likely to be different in
a 28-week-gestational-age newborn exposed to the thermal stresses of
extrauterine life than in a 28-week fetus comfortably surrounded by amniotic fluid.
The rates of weight gain, linear growth, and head growth between the ages of 24
and 36 weeks gestation can be calculated from the standard growth curves
generated from infants born prematurely (42,43). It must be recognized that the
data used to generate these plots are necessarily cross-sectional and thus need
smoothing to create the resemblance of a curve. Additionally, since premature
birth is an abnormal event and up to 30% of very-low-birth-weight (VLBW) infants
are small for dates (most likely as a result of the pregnancy failing over time), the
reliability of newborn data to assess the growth velocity of healthy fetuses
P.383
is suspect. Nevertheless, these curves are used extensively as guideposts for
neonatal growth of the preterm infant. On average, these curves predict that the
preterm infant should gain 10 to 15 g/kg body weight each day, grow 0.75 to 1.0
cm per week linearly, and demonstrate 0.75 cm per week of head growth. These
values had been utilized to calculate the energy and protein needs of the preterm
infant until recently. More accurate ultrasonographic techniques have been used
to measure fetal growth in healthy pregnancies. These studies suggest that the
rate of weight gain is closer to 18 to 20 g/kg body weight per day (44).
The nutrient requirements of term and preterm infants can be calculated based on
fetal reference figures, balance studies, serum nutrient values, or a combination
of these.
Energy Requirements
Energy requirements must take into account the amount and caloric density of the
solution ingested, the route of administration (enteral versus parenteral), the
amount lost in stool or urine, and the energy requirements in the body (e.g.,
basal metabolic rate, cost of growth, energy cost of food processing by the body)
(45). Many of these are now measurable, and reasonable estimates of energy
requirements to maintain optimal growth velocities can be made for both term
and preterm infants.
Energy is predominantly derived from carbohydrates and fat in the diet, which
provide 4 and 9 kcal/g, respectively. The infant fed human milk receives calories
predominantly from fat (46), whereas the formula-fed infant receives calories
more evenly distributed between fat and carbohydrate (47). The calories derived
from these sources are used first to maintain the total energy need of the infant,
which consists of the basal metabolic rate, the thermic effect of feeding, and
physical activity. Energy intake beyond this baseline is stored and recorded as
weight gain. Protein is not normally utilized as an energy source, unless the total
energy intake is less than the total energy expenditure of the infant. In those
cases, certain amino acids can be deaminated and shunted into the gluconeogenic
pathways to provide approximately 4 kcal/g of protein (48).
Energy requirements can be affected by numerous factors, including the route of
delivery and the disease state. Energy requirements are lower when infants are
fed parenterally as opposed to enterally because no energy is excreted in the
stool. Thus, the term infant who normally requires 100 kcal/kg/day enterally may
be fed 90 kcal/kg/day parenterally. Diseases that increase energy needs include
CHF (49), BPD (50), acute respiratory disease (51), and overwhelming sepsis
(52). Diseases that decrease energy needs include hypoxic-ischemic
encephalopathy and degenerative neurological conditions in which there is paucity
of physical movement.
Term Infants
Healthy breastfed term infants show adequate growth on as little as 85 to 100
kcal/kg body weight per day during the first four months of life (53). Formula-fed
infants have higher energy requirements (100-110 kcal/kg), most likely as a
result of a lower efficiency of digestion and absorption of fat (54). The presence of
a lipase in human milk increases the digestibility of its fats.
Preterm Infants
Preterm infants have higher energy requirements than term infants because of a
higher resting energy expenditure and greater stool losses as a result of immature
absorptive capacities (55,56) (Appendix J-1). Whereas the resting energy

expenditure of the term infant is 45 to 50 kcal/kg/day, the preterm infant less
than 34 weeks gestation consumes 50 to 60 kcal/kg/day (55,56). Stool losses
vary between 10% and 40% of intake, depending on the diet. For example, a diet
in which the carbohydrate is 100% lactose and the fats are predominantly longchain triglycerides will promote more stool losses because of the low levels of
lactase and the small bile salt pools in the premature infant. Replacement of up to
50% of the lactose with glucose polymers and between 10% and 40% of the fat
with medium-chain triglycerides (MCT) appears to reduce malabsorption to
approximately 10% (57). The preterm infant will need an additional 50 to 60 kcal/
kg/day beyond the daily energy expenditure and the loss of energy in the stool to
maintain growth along the intrauterine growth curve (10 to 15 g/kg/day). Thus,
barring any excess needs from diseases that increase oxygen consumption or
from malabsorption, the preterm infant will gain weight adequately on
approximately 120 kcal/kg/day. Assuming that 2.5 kcal are needed to achieve 1
gram of weight gain, this upward adjustment of expected weight gain would
require feeding an additional 10 to 15 kcal/kg body weight per day to the preterm
infant. Thus, it is likely that energy intake on the order of 130 to 135 kcal/kg/day
is a more reasonable target for the premature infant than the previous
recommended intake of 120 kcal/kg/day (57).
Energy Sources
Carbohydrates
Newborn infants are highly dependent on a source of glucose for normal brain
metabolism (58). The primary source of glucose in the term infant is lactose in
human milk and cow-milk formulas. Soy-based formulas provide glucose from the
metabolism of dietary sucrose or glucose polymers. Preterm infants also receive
glucose, initially as dextrose in parenteral solutions, but subsequently enterally
from lactose or glucose polymers. Galactose is also important to the newborn, as
it is needed for glycogen storage (59). The newborn infant typically utilizes
between 4 and 8 mg/kg/minute of glucose (60). This figure is commonly used as
the glucose infusion rate for parenteral nutrition. Because of their low glycogen
stores and poorer gluconeogenic capacities, preterm infants are more prone to
hypoglycemia than term infants (61). Higher rates of
P.384
glucose delivery (up to 15 mg/kg/minute) may be required in growth-retarded
infants and in infants of diabetic mothers to maintain normal glucose
concentrations.
Intravenous dextrose infusion rates up to 12.5 mg/kg/ minute are commonly used
in preterm infants to promote catch-up weight gain. Beyond this rate, a cost/
benefit analysis must be made. Although faster rates of weight gain can be
achieved on higher glucose infusion rates (especially if the serum glucose is
controlled with exogenous insulin infusion) (62), a higher metabolic rate and a
shift in the respiratory quotient will also occur. Thus, a higher oxygen
consumption rate coupled with proportionately more carbon dioxide generated by
the cells may significantly affect serum carbon dioxide and ventilatory
requirements. In one study, infants who received glucose and insulin remained on
the respirator an average of 13 days longer than their counterparts given lower
glucose infusion rates (63). Moreover, the increased growth rates demonstrated
with high glucose infusion rates are as a result of fatty weight gain, without any
increase in linear or brain growth (62). The overall metabolic cost of glucose
infusion rates > 12.5 mg/kg/minute must be weighed against the benefit of
increased rates of nonlean weight gain.
Fats
Lipids constitute the other major energy source for neonates. Certain fatty acids,
such as linoleic (omega-6, 18:2) and linolenic (omega-3, 18:3), are essential in
the diet, and their absence will produce deficiency syndromes characterized by
growth failure and skin rash (64). Although the full syndrome is rare, lower
essential fatty acid concentrations are seen within one week of discontinuing lipid
intake. Infants receiving parenteral nutrition or on a fat-restricted enteral diet
require 0.5 mg/kg/day of an intravenous fat blend containing these fatty acids at
least three times per week to prevent deficiency. The American Academy of
Pediatrics (AAP) has recommended that 3% of total energy intake in infants
should be in the form of linoleic acid (65).
Daily fat intake varies greatly based on the method of delivery (enteral vs.
parenteral) and the dietary source (human milk vs. formula). Enterally fed term
infants consume approximately 5 to 6 g/kg/day of fat, whereas parenterally fed
infants rarely receive greater than 4 g/kg/day, largely because of concerns about
toxicity. Infants receiving human milk (especially human milk expressed by
mothers who have delivered preterm) may receive up to 7 g/kg/day.
Infants fed human milk receive a unique blend of fats that has not been precisely
replicated in infant formulas. Cow-milk fat is generally not well tolerated by
newborn infants, forcing formula manufacturers to use vegetable oils as
substitutes. The spectrum of fatty acids found in palm, palm-olein, corn, and
coconut oils are distinctly different from human-milk fats.
The role of omega fatty acids such as docosahexaenoic acid (DHA) and
arachidonic acid (ARA) in the infant diet continues to be a subject of intense
research (57,66). These fatty acids are products of an elongation pathway from
linoleic acid and are important in cell membrane structure, in cell-signaling
cascades and in myelination (67). The synthetic pathways may be immature in
preterm infants, and for some undetermined period of time after birth in term
infants (68). Sources of these fatty acids include the placenta and human milk. In
contrast, cow-milk fat and vegetable oil do not contain these compounds. A
number of studies addressed whether these particular fatty acids are essential in
the preterm and term infant (69,70,71,72). Because the preterm infant may be
less capable of synthesizing the compounds and would have received them in
utero, the European Society for Pediatric Gastroenterology and Nutrition has
recommended that a source of these fatty acids be added to preterm infant
formula (73). Evidence of their efficacy includes studies that demonstrate better
visual acuity, more mature electroretinograms, and short-term gains in general
neurodevelopment (74,75,76). Studies of the longer-term growth and
developmental outcomes of infants who have been supplemented with these fatty
acids are in their early stages but suggest continued positive effects on the visual
system at one year of age (77). It remains unclear whether potential early
neurodevelopmental gains are sustained beyond the first year (78). A review
assessing the role of long-chain polyunsaturated fatty acid (LC-PUFA)
supplementation on neurodevelopment concluded that the studies remain too
underpowered to support a positive long-term effect (79). The lack of a consistent
effect may be as a result of suboptimal dosing of the compounds. Additionally,
any time a single component of human milk can be isolated and added to cowmilk-based infant formula, an important consideration is whether the component
exerts its nutritional effect individually or in consort with other compounds (80).
The Food and Drug Administration (FDA) in the United States has recently
approved a fungal source of DHA and ARA and its incorporation into infant formula
under a provision termed Generally Recognized as Safe (GRAS). This
designation confirms that the FDA has no questions of the manufacturers
regarding the safety of the source and stability of these compounds when used in
the intended matrix (e.g., infant formula). The GRAS designation is used to
indicate that the ingredient in question has been in the food chain of humans and,
based on prior usage or experimental evidence reviewed by an expert scientific
panel, poses no safety concern. The GRAS determination does not assess efficacy
claims; indeed, claims of efficacy for a new ingredient added to infant formula
would be subjected to a more thorough examination process by the FDA, not
unlike that required of new drugs. The major formula manufacturers in the United
States have now added DHA and ARA to their term, preterm hospital, and preterm
discharge formulas without making efficacy claims. The formulas appear to be as
safe as formulas without added DHA and ARA. It is likely that formulas without the
added fats will be phased out over time.
Carnitine is another compound involved in fat metabolism in the neonate.
Although carnitine deficiency is rare in
P.385
the enterally fed infant because of high levels in human milk and supplementation
of formulas, it remains a concern in infants receiving long-term exclusive

parenteral nutrition in which carnitine has not been supplemented (81). Carnitine
supplements can be added to total parenteral nutrition (TPN), and this is
recommended in infants who receive TPN for greater than 3 weeks (82).
Protein Requirements
Protein requirements in humans are determined by a number of factors including
protein quality and quantity, the amount of energy delivered, and the protein
nutritional status of the subject (20). The latter is influenced by the degree of
previous malnutrition, by the rate of catch-up growth, and potentially by
inflammatory processes. The sick newborn infant is exposed to many of these
influences. Additionally, they have a high basal requirement for protein accretion
based on in utero nitrogen accretion rates (2,41). Adequate energy intake is
important to promote optimal protein utilization, with a nonprotein calorie-togram nitrogen ratio of 200:1 considered ideal. Overall protein intake in the
neonate is ultimately limited to about 4 to 4.5 g/kg/day because of the inability of
the immature kidney to excrete titratable acid, blood urea nitrogen (BUN), and
ammonium ion (83). The renal excretion limitations are proportional to the degree
of prematurity.
Protein requirements, in general, and branched-chain amino acid needs, in
particular, are increased in adults with physiological instability as a result of septic
or surgical illness (84). Recently, the possibility that similar changes might occur
in sick neonates has been preliminarily investigated (51,52,85). Neither acute
respiratory disease nor sepsis nor surgical ligation of the patent ductus arteriosus
(PDA) results in increased protein requirements (51,52,85). At this time,
increasing protein delivery routinely on the basis of illness or physiological
instability is not recommended. Conversely, practitioners frequently limit nutrient
delivery during illness out of concern that high loads may be metabolically taxing.
Recent studies demonstrate that up to 3 g of protein/kg body weight can be
administered daily to sick preterm infants beginning in the first 24 hours after
birth (86).
Term Infants
The full-term breastfed infant grows adequately and maintains normal serum and
somatic i.e., muscle protein status on as little as 1.5 g/kg/day of protein. Although
the protein content is low (1.1%), the quality of human milk protein is excellent
because the spectrum of amino acids provides a unique match for the amino
acid needs of the newborn. The protein content is predominantly lactalbumin, as
opposed to casein, which makes for smaller curds and easier digestibility.
Additionally, human milk is replete with nondietary nitrogen sources including
nucleotides, which may enhance the immune system (87); immunoglobulins and
other antimicrobial factors, which help protect the gut epithelium (88,89); growth
factors, which stimulate intestinal growth (90); and enzymes (e.g., lipases), which
aid digestion.
The term infant fed a cow-milk or soy-based infant formula requires a greater
protein delivery rate, most likely to compensate for the less-than-ideal protein
quality. Thus, the infant on cow-milk formula typically requires 2.14 g/ kg/day and
the infant on soy formula up to 2.7 g/kg/day of protein (47). Cow-milk protein is
predominantly casein, although a number of cow-milk-based formulas are
modified to be whey predominant. The soy formulas also promote adequate
growth of lean body mass. However, these formulas contain a smaller percentage
of available nitrogen as essential or semi-essential amino acids (91).
Protein can also be delivered to the term infant by way of protein hydrolysate or
individual amino acid formulas. These formulas are specifically designed to
decrease the exposure of the infant to potentially antigenic cow- or soy-milk
proteins. By hydrolyzing the cow-milk-based protein such that greater than 90%
of the proteins have a molecular weight of 1,250 Daltons or less, allergic disease
as a result of cow-milk allergy can be treated or potentially prophylaxed. These
formulas provide approximately 2.8 g/kg/day of protein at an energy delivery of
100 kcal/kg/day.
Preterm infants
Recommendations for protein intake in preterm infants follow many of the same
parameters as in term infants. However, the needs of the preterm infant appear
to be greater than the term infant. Early studies suggested that the most rapid
weight gain and most efficient energy utilization was achieved with protein intakes
of 3 to 5 g/kg/day (92,93). Kashyap and associates refined these goals when they
demonstrated that weight gain and nitrogen retention were greatest in healthy 32week-gestational-age infants fed 3.9 g/kg/day of protein (94). They also
attempted to define the optimal energy-to-protein ratio which promoted growth,
finding that approximately 30 kcal were necessary for each gram of protein
delivered. In a later study by the same group, Schulze and associates proposed
that preterm infants tolerate 3.6 g/kg/day of protein with an energy intake of at
least 120 kcal/kg/day (95). Heird has emphasized that increasing protein delivery
requires increased energy delivery, and vice versa (96). His conclusions include:
The low-birth-weight (LBW) infant who can take feeds soon after birth
requires a protein intake of at least 2.8 g/kg/day.
Infants who do not receive protein in the first few days of life lose at least
1% of their endogenous protein stores daily. (These findings are consistent
with research stating that preterm neonates were in better nutritional status
if amino acids were added to dextrose solutions in the first days of life [97].)
Infants with delayed protein intake subsequently require higher protein

intakes to correct for early losses.
P.386
Denne has used stable isotopes of nitrogen to study the protein requirements and
distribution (synthesis versus breakdown) in extremely-low-birth-weight (ELBW)
infants (98). His calculations indicate that an average daily intake of 3.2 g/kg of
protein is necessary to counter the negative nitrogen balance of neonatal illness
and to match the expected in utero protein accretion rate. Ziegler's data in stable,
growing premature infants also confirms that enteral protein delivery less than 3.5
g/kg/day results in suboptimal growth (99).
Besides total protein delivery, recent studies have considered which amino acids
may limit protein accretion in the preterm infant. The terms essential and
nonessential amino acids have been replaced in the neonatal lexicon by
indispensable or limiting and dispensable (100) because they are more
descriptive of the effects of amino acids on protein metabolism. Threonine and
lysine are clearly indispensable because they cannot be synthesized de novo from
products of carbon intermediary metabolism. Heird suggests that these two amino
acids may currently be the limiting amino acids in TPN solutions.
Finally, illnesses or medications that increase protein turnover or muscle
breakdown will have an influence on protein delivery. Van Goudoever and
associates demonstrated that the steroids used for the treatment of BPD cause
negative nitrogen balance by increasing the rate of protein breakdown but have
little effect on protein synthesis (101).
Many preterm infants receive protein initially as part of a regimen of parenteral
nutrition. Intravenous amino acid solutions have advanced to the point of being
specifically formulated for preterm infants. These amino acid solutions are
designed to normalize the plasma amino acid profile of the healthy infant,
promoting levels similar to those of a one-month-old breastfed infant (102).
Anderson and associates demonstrated that dextrose solutions with amino acids
given early in life promote better nutritional status than dextrose solutions without
protein (97). Denne et al. have shown that newborn preterm infants respond to
parenteral nutrition with an acute increase in protein synthesis and a decrease in
proteolysis (103). Thus, it appears that amino acid delivery in the first days of life
is critical.
TABLE 22-1 NUTRIENT AND MINERAL CONTENT OF PRETERM MILK
Element
Days 37
Day 21
Days 2942 Days 5798
Protein (g/dL)
3.24 0.31 1.83 0.14 1.31 1.81 1.8 0.07
Lactose (g/dL)
Fat (g/dL)
5.96 0.2 6.49 0.21 0.12

1.63 0.23 3.68 0.4
Energy (kcal/dL)
Sodium (mEq/dL)
51.4 2.4
2.66 0.3
65.6 4.3
1.3 0.18
Chloride (mEq/dL)
3.16 0.3
1.70.17
Potassium (mEq/dL)
Calcium (mg/dL)
1.74 0.07 1.630.09

20.326.3 20.4 1.5
0.1 1.1
1.1 0.1
24.626.2 31.5 1.3
1.7
2.2
0.76 0.09 0.55 0.05
Phosphorous (mg/dL)
9.514.6
0.7
14.9 1.3
13.3 0.3
Magnesium (mg/dL)
2.8 .1
2.4 0.1
4.9 0.1
From Gross SJ, David RJ, Bauman L, et al. Nutritional composition of milk
produced by mothers delivering preterm. J Pediatr 1980;96:641644;

Schanler RJ, Oh W. Composition of breast milk obtained from mothers of
premature infants as compared with breast milk obtained from donors. J
Pediatr 1980;96:679-681; Feeley RM, Eitenmiller RR, Jones JB, et al.
Calcium, phosphorus and magnesium contents of human milk during early
lactation. J Pediatr Gastroenterol Nutr 1983;2:262-267, with permission.
The recommended amount of amino acids to be delivered and the rate of
advancement remain to be determined, and are influenced by the infant's
condition. Based on the estimates of protein loss by Heird and associates (96), it
is prudent to begin amino acids on day one. The BUN and acid-base status of the
infant must be followed. Older studies in sick neonates suggest that 1.5 g/kg/day
of amino acids on day two would achieve neutral nitrogen balance (51). However,
more recent data from Thureen and associates indicate that preterm infants can
be safely started on 3.0 g/kg/day on day one (86). Ultimately, the goal is to
deliver between 3.0 and 4.0 g/kg/day of parenteral amino acids. Some newborn
intensive care units carry stock solutions of dextrose with added protein to be
started upon admission to the nursery before parenteral nutrition from the
pharmacy can be ordered and obtained. The amount of protein is designed to
deliver 2 g/kg/day when the fluid solution is prescribed at 75 cc/kg/day.
Preterm infants who begin enteral feeds in the first days after birth will be
relatively protein restricted because they are typically given human milk or low
amounts of preterm infant formula. Although human milk from mothers delivering
preterm is initially higher in protein (Table 22-1) compared to term human milk
(104,105,106), the content is still relatively low and the infant will likely receive
low volumes. Fortification with a human milk fortifier addresses this low protein
issue. Preterm infant formula has a high concentration of protein (3.1 to 3.6
g/120 kcal), but it is unlikely that the infant will achieve that type of delivery in
the first week. With the trend toward increased protein delivery to the preterm
infant, Ziegler and others have stressed the importance of supplying adequate
energy intake (99,107).
P.387
Mineral And Element Requirements

Mineral requirements in newborns are influenced by the immature status of the
kidney, prematurity, and medications that affect mineral metabolism (108). In
general, preterm infants require higher amounts of minerals than term infants.
The daily needs of some minerals (e.g., sodium, potassium chloride) are
determined by measuring serum levels, although others (e.g., calcium,
phosphorus) are estimated from in utero accretion rates (2,41). A more
comprehensive discussion of these needs can be found in other chapters in this
volume (e.g., Fluid/Electrolyte, Bone Mineralization) and in other sources (109).
The following discussion focuses on selected minerals, comparing requirements in
the preterm infant to the term infant and emphasizing the effects of diseases and
their treatments on mineral status.
Sodium, Potassium, and Chloride

The typical term infant requires 1 to 3 mEq/kg/day of sodium (108). Breastfed
infants have a lower sodium intake than formula-fed infants, although formula
manufacturers have reduced the sodium content of cow-milk formula to levels
comparable to human milk. Preterm infants have higher sodium requirements and
may become hyponatremic on human milk (19,110). Baseline sodium
requirements range from 2 to 4 mEq/kg day in the preterm infant. The higher
sodium requirement with lower gestational age is due predominantly to the
immaturity of the proximal renal tubule in the small preterm infant. Treatment of
these infants with salt-wasting diuretics, e.g., furosemide, hydrochlorothiazide
may increase the needs to 10 to 13 mEq/kg/day.
Potassium requirements in the term infant generally range between 1 and 2 mEq/
kg/day. Because potassium is also reabsorbed at the proximal tubule, the preterm
infant needs 2 to 4 mEq/kg/day (110). Diuretics such as furosemide, metolazone,
and hydrochlorothiazide can increase requirements to 8 mEq/kg/day.
Sodium and potassium are provided in human milk and infant formulas to satisfy
the daily requirement for the normal infant. However, excessive demands from
prior or ongoing losses may require supplementation. Supplements are typically in
the form of chloride salts (NaCl or KCl) and are dosed based on calculations of
maintenance requirements plus deficits. Depending on the acid-base status of the
infant receiving parenteral nutrition, sodium and potassium can be added as
chloride or as acetate.
Calcium, Magnesium, and Phosphorus

The term infant who has not experienced intrauterine growth retardation has wellmineralized bones. Analysis of the reference fetus demonstrated that the average
fetus accretes 80% to 90% of its calcium during the last trimester (2). If there
has been no interruption of the process, the average term infant will grow well
and remain well mineralized on a diet that provides approximately 40 to 60 mg/kg/
day of calcium and 20 to 30 mg/kg/day of phosphorus. Human milk is an excellent
source for this rate of delivery. Infant formulas have higher concentrations of
calcium than human milk, but there is little evidence that term infants who are
formula fed are better mineralized.
LBW infants (intrauterine growth retarded or preterm) have significantly lower
calcium and phosphorus content than term, appropriate-for-dates infants. Daily
enteral calcium and phosphorus requirements in the preterm infant are estimated
at 120 to 230 and 60 to 140 mg/kg/day, respectively (111). This rate of daily
accretion can only be achieved enterally. The two minerals become insoluble in
typical TPN solutions when added at concentrations that would provide more than
60 mg/kg/day of calcium and 30 mg/kg/day of phosphorus (at typical fluid
delivery of 150 cc/kg/day). Human milk is relatively low in calcium and
phosphorus and thus must be supplemented with a fortifier to achieve adequate
mineralization in the less than 1,500 g infant (112). Preterm infant formula
contains enough calcium and phosphorus to deliver the recommended daily
amount and some excess to begin to replace previously acquired deficits. This is
important because most preterm infants do not tolerate full enteral feedings
immediately after birth. Preterm infants will invariably need intakes greater than
the projected daily in utero requirement to avoid becoming osteopenic since
parenteral nutrition results in negative calcium and phosphorus balance (4).
Because serum calcium levels will always be maintained at the expense of the
bone, following calcium levels as a way of monitoring status is not useful. Instead,
indicators of bone activity (serum alkaline phosphatase) or urinary phosphorus
excretion are used to monitor long-term calcium status.
As with calcium, the rates of magnesium accretion in utero are high during the
third trimester. Thus, the magnesium requirement for infants born preterm are
greater than those for term infants. Human milk and preterm infant formula
appear to support normal magnesium levels. Magnesium delivery in parenteral
nutrition is titrated based on serum magnesium levels, which should be
monitored. The daily enteral magnesium intake should be 10 to 26 mg/kg/day
(57) although the advisable parenteral intake should be 4.3 to 7.2 mg/kg/day
(111).
Iron
The majority of total body iron found in the term infant is accreted during the
third trimester. The fetus maintains a constant total body iron content of 75 mg/
kg during the last trimester, increasing from 35 to 40 mg at 24 weeks gestation to
225 mg at term (113). Preterm delivery results in disruption of this process, and
premature infants are therefore born with lower iron stores than term infants.
Small-for-dates infants are frequently born with low iron stores, presumably
because of decreased placental iron transport (114). Infants of diabetic mothers
are born with low stores
P.388
because much of the fetal iron is in the expanded red cell mass (114). These
infants also appear to be low in total body iron, most likely as a result of altered
transport of iron by the diabetic placenta, such that the increased iron need of the
infants of diabetic mothers exceeds the placental transport capacity (116,117).
Term Infants
The appropriate-for-dates newborn infant has sufficient iron stores to last four to
six months; the small-for-dates infant has closer to a 2-month supply (118,119).
In the absence of adequate dietary iron, these stores are mobilized for
hemoglobin synthesis in the rapidly expanding blood volume of the growing infant.
An adequate source of iron generally maintains iron stores until the infant begins
to obtain iron from other dietary sources in the second 6 months of life. The
estimated daily iron requirement for the term infant is 1 mg/kg/day (118,119).
The major source of iron for the healthy, term infant is dietary, either through
human milk or infant formula fortified with iron (119). Although human milk has a
low iron content (0.3 mg/L) compared to either iron-fortified infant formula (1012 mg/L) or low-iron formula (4.5 mg/L), the iron is much more bioavailable as
a result of proteins such as lactoferrin (120). Greater than 50% of the iron in
human milk is absorbed, compared to only 4% to 12% of formula iron (121). The
rate of iron deficiency in breastfed infants before six months is relatively low,
although few methodologically sound studies have been performed (122). After 6
months, iron deficiency rates of 20% to 30% have been recorded in breastfed
infants (123), although it has been unclear whether the infants in these studies
were exclusively breastfed. Innis and associates showed an iron-deficiency anemia
rate of 15% in 8-month-old breastfed infants (124). Given this, it may be wise to
screen the breastfed infant's iron status at 6 months.
The FDA defines low-iron formulas as those containing less than 6.7 mg/L. Prior
to changes in the iron content of low-iron formula at the end of the millennium,
the rate of iron deficiency in infants fed exclusively a low-iron formula containing
less than 2 mg/L was unacceptably high, with rates between 28% and 38% (123).
This compares with a rate of less than 5% in infants fed iron-fortified formula (12
mg/L) during the first 6 months of life. Indeed, the introduction of iron fortification
of formulas in the early 1970s represents one of the most effective public health
campaigns in this country. Recent studies indicate that infants fed formulas
containing either 4 or 7 mg/L remain iron sufficient (125). No advantage in iron
status is conferred to infants consuming formulas containing 8 as compared to 12
mg/L (126). Low-iron formulas have been reconstituted during the last 4 years
such that they all contain at least 4 mg of iron/L, thus meeting the AAP
recommendations (119). They continue to carry the designation low iron
because of the FDA standard.
Low-iron formula continues to account for 9% to 30% of elective, e.g., non WIC
formula sales. The reasons appear to involve the unfounded perception that iron
in formula causes gastrointestinal symptoms such as colic, diarrhea, constipation,
and GER. Double-blind studies have failed to support these claims (127,128).
Preterm Infants
The preterm infant that is not growth retarded begins extra-uterine life with the
same iron stores per kilogram body weight as the term infant (approximately 12
mg/kg). However, the preterm infant is exposed to several stressors that perturb
iron balance, with the result that by the time of hospital discharge, the infant may
be iron deficient (129,130) or iron overloaded (131,132). The range of iron status
of the preterm infant at 40 weeks postconception appears to be far wider than the
term infant, although systematic studies are lacking.
The preterm infant frequently goes into negative iron balance because of blood
lost during phlebotomy although sick, coupled with a rapid growth rate (and
expansion of the red cell mass) during the convalescent period. In the past,
phlebotomy losses were replaced by red cell transfusions, but criteria for
transfusion have become more stringent because of concerns of exposure to
infectious agents (133). Recombinant human erythropoietin has been used to
stimulate endogenous red cell production in place of red cell transfusion, and
produces additional negative stress on neonatal iron balance (134). Since 3.4 mg
of iron are necessary to synthesize 1 g of hemoglobin, the therapeutic use of
recombinant erythropoietin significantly taxes the already low iron stores of the
preterm infant. Because of these factors, the daily enteral iron requirement for the
preterm infant who does not receive recombinant erythropoietin is 2 to 4 mg/kg/
day, with the greater requirement for the more preterm infant (108). Infants who
receive recombinant erythropoietin require at least 6 mg/kg/day of iron (108).
The issue of when to begin iron supplementation in the preterm infant is
controversial. Iron is necessary for normal growth and development of all tissues,
including the brain. A rich literature supports the hypothesis that early iron
deficiency results in neurodevelopmental sequelae at the time of the deficiency
and persists well after iron has been repleted (135). Iron deficiency in infants as
young as 6 months of age slows nerve conduction (136), a finding that persists
even after repletion of iron (137). Nevertheless, iron is also a very potent oxidant
stressor since it catalyzes the Fenton reaction to produce reactive oxygen species.
Since preterm infants have immature antioxidant systems, there is concern that
free iron (i.e., in excess of the total iron-binding capacity) can exacerbate
diseases that may be related etiologically to oxidative stress, including BPD,
necrotizing enterocolitis, neuronal injury, and retinopathy of prematurity (ROP)
(137,138,139,140,141). Although definitive studies on the role of iron in these
diseases of prematurity have not been performed, it is clear that infants who are
multiply transfused with packed red blood cells and those who receive parenteral
iron are at risk for having free circulating iron and increased markers of oxidative
stress
P.389
(131,132,142). Because of these concerns, parenteral iron should be used very
sparingly. Since infants are born with adequate iron stores, there is no need to
begin iron supplementation in a sick, nongrowing preterm infant. Therefore,
enteral iron supplementation should not be started before two weeks postnatal
age (108,138). Conversely, delaying iron supplementation until after two months
confers a very high risk of iron deficiency in the postdischarge period (129).
TABLE 22-2 RECOMMENDED MICROMINERAL INTAKES FOR PRETERM
INFANTSd
Transitional Period (014 days)
Stable/Postdischarge Periods
Enteral(g/ Parenteral(g/kg/ Enteral(g/ Parenteral(g/kg/

kg/day)
day)
kg/day)
day)
Zn
500800
150
Cu
120
0, 20
120150
20
Se
1.3
0, 1.3
1.33.0
1.52.0
Cr
0.05
0, 0.05
0.10.5
0.050.2
Mo
0.3
0.3
0.25
Mn
0.75
0, 0.75
0.757.5
1.0
1127
0, 1.0
3060
1.0
1,000
b
400
b
Postdischarge supplement of 0.5 mg/kg/day for infants fed human milk.
Should be withheld when hepatic cholestasis is present.
For long-term total parentarel nutrition only.
From Reifen RM, Zlotkin SH. Microminerals. In: Tsang RC, Lucas A, Vauy
R, et al., eds. Nutritional needs of the preterm infant. Baltimore: Williams
& Wilkins, 1993:195207; Greene H, Hambridge K, Schanler R, et al.
Guidelines for the use of vitamins, trace elements, calcium, magnesium,
and phosphorus in infants and children receiving total parenteral nutrition:
report of the Subcommittee on Pediatric Parenteral Nutrient Requirements
from the Committee on Clinical Practice Issues of the American Society for
Clinical Nutrition. Am J Clin Nutr 1988;48: 1324-1342, with permission.
Generally, human milk should be used whenever possible in preterm infants.
However, because of the low iron content of human milk and the rapid growth
rate of these infants, iron supplementation is highly recommended. Additionally,
those who receive recombinant human erythropoietin should be supplemented
earlier with iron sulfate to have an adequate erythropoietic response. It is clear
that iron must be available to see a sustained erythropoietic response with
recombinant erythropoietin treatment (143). Preterm infant formulas are iron
fortified and should provide adequate amounts to the larger preterm infant.
However, preterm infants less than 30 weeks gestation may well need enteral iron
supplementation in addition to their preterm formula to bring their total dose
closer to 4 mg/kg/day.
Trace Elements
Ten trace elements are nutritionally essential for the human: zinc, copper,
selenium, chromium, manganese, molybdenum, cobalt, fluoride, iodine, and iron
(144). Zlotkin and associates have written an excellent review of trace element
requirements in newborns (144). Most trace elements are accreted during the last
trimester. Thus, the term infant is fully replete and needs modest dietary intake of
these elements. Both human milk and infant formula ensure adequate intakes.
The preterm infant or the term infant on prolonged TPN would rapidly go into
negative balance of any of these elements if not provided with an exogenous
source. While on TPN, infants should receive neonatal trace elements (Table 222). Preterm infant formulas and preterm human milk appear to supply adequate
amounts of trace elements to the enterally fed premature infant (145).
Selenium is a potent antioxidant. Preterm infants have lower selenium stores than
term infants (146), and this has been proposed as an etiology for diseases such as
BPD and ROP (147). Studies linking selenium insufficiency with these diseases
have not been persuasive (148), but the general consensus is that selenium
status should be supported in the preterm infant (149). Selenium is not found in
commercially available products and must be added separately to TPN at the rate
of 2 g/kg/day (149). Iodine is not added to TPN, but adequate amounts of iodine
are absorbed through the infant's skin from iodine solutions applied topically
(144). Nevertheless, a dose of 1 g/kg/day has been recommended for infants
who are on TPN for more than 6 weeks (149).
Vitamins
Vitamin requirements in newborn infants can be most easily conceptualized by
considering water- and fat-soluble vitamins separately. An extensive review of all
of the vitamins and their deficiencies is beyond the scope of this chapter and the
reader is referred to sources dedicated to this subject (150). This section will deal
primarily with vitamins that are of particular relevance to neonates and to those
with a specific risk for deficiency.
P.390
Water-Soluble Vitamins
Term newborns are rarely deficient of water-soluble vitamins in the B group
(150). As with all humans, neonates need a daily source of vitamin C and folate.
These are provided in adequate concentrations in human milk, infant formulas and
multivitamin preparations added to parenteral nutrition. The AAP has stated that
term breastfed infants do not need supplemental water-soluble vitamins during
the first six months unless there are extenuating circumstances (7). Preterm
infants also do not appear to need supplemental water-soluble vitamins once they
are taking an adequate amount of formula or fortified human milk. The minimum
amount of enteral feeds needed to maintain vitamin sufficiency varies among the
formulas and the human milk fortifiers available. Infants receiving Enfamil
Premature Formula or human milk fortified with Enfamil Human Milk Fortifier need
no vitamin supplementation if their intake exceeds 150 mL/day. Those receiving
Similac Special Care Formula or human milk fortified with Natural Care must
exceed 300 mL/day to remain vitamin sufficient (151).
Fat-Soluble Vitamins
Fat-soluble vitamin deficiencies are also rarely a problem for term, healthy
newborns fed human milk or infant formula. Nevertheless, certain groups of
infants are at risk for vitamin D deficiency (150). These include breastfed infants
whose mothers are vitamin D deficient as a result of their diet (vegan) or whose
mothers completely protect their own skin from sunlight. Their infants must also
be exposed to less than 30 minutes of sunlight per day to be at greatest risk. Most
reports of rickets in breastfed infants in these circumstances have been in farnorthern climates, although in the United States the problem has been seen as far
south as San Diego and North Carolina (152,153). The AAP has recently
recommended that all infants receive 200 IU of vitamin D daily. For infants
consuming less than 500 mL of infant formula per day (including all breastfed
infants), the least expensive and easiest way to achieve this goal is through a trivitamin preparation.
Virtually all infants receive vitamin K in the delivery room to prevent hemorrhagic
disease of the newborn. The prevalence of this condition is very low, but the
neurological consequences are so disastrous and preventable that the current
recommendation is to continue to give vitamin K at birth. Once a gut flora has
been established in the first 2 postnatal days, vitamin K deficiency is exceedingly
rare (154). However, infants who receive broad-spectrum antibiotics, which
markedly reduce the intestinal flora, should be supplemented with vitamin K at

least twice per week (154).
Whereas infants are not wholly dependent on dietary sources for vitamin D (it can
be synthesized de novo from sterol precursors by 1 week of age) and vitamin K
(supplied by gut bacteria), vitamins E and A must be provided in the diet. The
term infant who consumes human milk or infant formula receives an adequate
amount of each, assuming that there are no impediments to fat absorption, such
as cystic fibrosis or short bowel syndrome. In infants with those conditions, a
water-soluble A and E preparation should be utilized and serum levels monitored.
There are significant issues with fat-soluble vitamins, particularly vitamins A and
E, in preterm infants less than 34 weeks gestation because of their relatively poor
digestion of fats. As with term infants, vitamin K and, most likely, vitamin D are
not a major problem, although preterm infant formulas are supplemented with
more vitamin D than are term formulas. Even the most premature infants are
capable of synthesizing the active vitamin D metabolite by one week of age (155).
Vitamin A deficiency in growing mammals results in significant tissue fibrosis,
particularly of the lung and liver. In 1925, Wolbach and Howe noted that lambs
born to vitamin A-deficient mothers had significant lung disease that, in
retrospect, bears remarkable similarity to BPD (156). These animals had
squamous metaplasia, loss of cilia, and loss of ciliary motility. Several
investigators have noted that preterm infants have low vitamin A levels and
hepatic stores (157). Moreover, preterm infants with respiratory distress
syndrome who had low cord-blood vitamin A concentrations are more likely to
develop BPD (158). Additionally, both prenatal and postnatal steroid
administration increased serum vitamin A levels in preterm infants although
preventing and treating BPD, respectively (159, 160). Two small but wellcontrolled studies assessed whether prophylactically administering 2,000 IU of
vitamin A intramuscularly every other day from birth in <1500 g infants prevented
BPD. Shenai and associates found a significant reduction of BPD in their treated
infants, from 85% to 45%, with a concomitant rise in serum vitamin A levels
(161). The serum retinol levels of placebo-treated infants remained abnormally
low (<20 g/dL). In contrast, Pearson et al., utilizing the same protocol, found no
difference in BPD between the vitamin A- and placebo-treated groups (46% vs.
44%) despite a significant rise in serum retinol concentrations in the treated
group (162). Of note was that retinol levels remained above 20 g/dL in the
placebo group. The unifying conclusion of these two studies is that vitamin A
deficiency, as seen in the control group of Shenai and associates is indeed a risk
factor for BPD, but not all infants less than 1,500 g are vitamin A deficient in the
first month of life. Further supplementation of infants with adequate vitamin A
status does not appear to further reduce the risk of BPD according to Pearson's
data (162). More recently, the National Institute of Child Health and Human
Development network addressed this issue in a multi-center, randomized trial
(163). There was a small but statistically significant reduction in the rate of BPD in
the vitamin A-treated group. The data were not analyzed as a function of initial
serum retinol concentration. At this time, treatment with vitamin A for all infants
at risk for BPD is not unreasonable. The risks are relatively small, but include
three-time-weekly intramuscular injections for 4 weeks in infants with little muscle
mass and the relatively remote possibility of vitamin A toxicity. Vitamin A levels
should be followed
P.391
weekly to monitor for toxicity and efficacy since the early studies suggest the
greatest benefit occurs in infants with low initial vitamin A levels. Another
approach is to check vitamin A levels in preterm infants at risk for BPD at birth
and treat only those with low serum concentrations, realizing that serum
concentrations are not the most reliable measure of tissue vitamin A status.
Ever since Oski and associates reported on hemolytic anemia caused by vitamin E
deficiency, studies have assessed vitamin E sufficiency in the context of oxidative
stress (164). Clearly, phospholipid membranes are at high risk for oxidative
stresses; and if not adequately protected by circulating antioxidants such as
vitamin E, selenium and superoxide dismutase will be damaged, with subsequent
cell death. Thus, it was hoped that vitamin E supplementation of the preterm
infant who has an immature antioxidant system might prevent or ameliorate
established BPD or ROP. Studies along those lines have been a disappointment;
after initial reports that vitamin E supplementation at birth prevented BPD
(165,166,167,168), subsequent studies have not been able to duplicate the effect
(168). Similarly, a meta-analysis of trials of vitamin E supplementation to prevent
the occurrence or progression of ROP has not shown a significant effect (169).
Moreover, high vitamin E levels following supplementation appear to increase the
risk of sepsis and NEC (170).
Preterm infant formulas are supplemented with vitamins E and A. For most infants
fed the preterm infant formula with higher vitamin E and A concentrations, serum
levels remain in the normal range. However, routine assessment of these levels in
the high-risk infant less than 1,500 g may be prudent, since it is likely that the
deficiency state is not advantageous to the growing infant. The consensus panel
recommendations for daily vitamin intake via parenteral nutrition were published
in 1993 (171) and were adapted from a special report of the subcommittee on
pediatric parenteral nutrient requirements (149).
The Effect Of Neonatal Illness on Nutritional

Requirements
Most studies of neonatal nutritional requirements have dealt with defining the
needs of the healthy growing term or preterm infant. Nevertheless, adults and
older children who are ill undergo profound changes in metabolism based on the
type and degree of illness (172,173). Cerra and associates have investigated the
independent effects of surgery, trauma, and sepsis on adult metabolism and have
found consistent changes in protein-energy requirements (173). Each incident
increases cellular oxygen consumption and promotes more negative nitrogen
balance; sepsis has the most profound effects. Cytokines such as tumor necrosis
factor (TNF-alpha), interleukin-6 (IL-6) and interleukin-1 (IL-1) appear to be
important mediators of the response (174). These are elevated in preterm and
term infants with sepsis (175). The adult studies suggest that energy and amino
acid delivery must be significantly modified in sick patients. In particular, these
patients appear to require higher energy delivery and more protein to remain in
neutral or positive nitrogen balance. Special amino acid solutions that are rich in
branched-chain amino acids are utilized to support nitrogen balance (176). Fewer
studies have assessed these issues in preterm and term newborns. Nevertheless,
some of the metabolic effects of acute lung disease, chronic lung disease, CHF,
and sepsis have been studied (5). The conclusions of these studies support the
concept that simply supplying the nutrients normally required by the healthy
newborn will not be sufficient for infants with these illnesses.
Overall Approach Based on Stage of Illnessin

Preterm Infants
In 1995, the Canadian Pediatric Society proposed a novel approach to neonatal
nutrition, primarily designed for premature infants but adaptable for term infants
(177). This approach acknowledged that nutritional needs of premature infants
change based on the phase of their illnesses. The initial transition phase is
characterized as the time of neonatal illness and physiological instability. Although
defined as the first ten postnatal days, the timing is flexible based on the duration
of neonatal illness. The sick premature infant is likely to be relatively insulin
resistant and to have increased circulating counter-regulatory (gluconeogenic)
hormones, including cortisol and glucagon. Vasopressor therapy with dopamine,
dobutamine, or epinephrine may have effects similar to the endogenous
hormones. These factors, combined with the release of cytokines during illness
(177), place the infant in a catabolic rather than anabolic state. Under such
circumstances, growth factors are down-regulated (178) and growth is unlikely.
Premature infants typically show weight loss and lack of head growth during their
period of illness. The nutritional strategy for this first phase is incompletely
defined. Principles gleaned from the adult critical care literature would suggest
that energy delivery should at least meet resting energy expenditure to prevent
further glycogen, muscle, and fat breakdown for gluconeogenesis. Once resting
energy expenditure has been met in the stable, growing infant (approximately 60
kcal/kg/d), excess calories are channeled for growth, with a cost of about 2.5
calories per gram of weight gain. However, in the insulin-resistant sick infant with
down-regulated growth factors, these extra calories are likely to increase the
cellular metabolic rate, generating more carbon dioxide but not resulting in
growth. Carbohydrates, with their respiratory quotient of 1.0, are more likely than
fat to create more carbon dioxide. The source of nutrition is usually parenteral
with minimal enteral or trophic feedings. Administration of protein appears to be
beneficial during this period to reduce the degree of negative nitrogen balance.
Recent data of Thureen and associates demonstrating the safety and efficacy of 3
grams of protein/kg body weight on day one supports this concept (86). The use
of specialized protein blends to replace particular amino acids lost during illness
has not yet been studied extensively.
P.392
The in-hospital growth period is characterized by physiological stability and an
anabolic state. The goal during this period is to match intrauterine growth rates
and mineral accretion. For the preterm infants less than 34 weeks postconception,
nutrient delivery should be adjusted to take into account digestive and absorptive
immaturities. Fortification of preterm human milk and the use of premature infant
formulas typically address these issues.
Preterm infants in the postdischarge phase are also anabolic and growing.
Compared to the term infant, their physiology is characterized by a mature
absorptive and digestive system. Unlike the term infant, however, these infants
have accrued large energy, protein, and mineral deficits (3,4), and their growth is
frequently below the 5th percentile for their age adjusted for prematurity.
Premature infant discharge formula and fortification/supplementation of human
milk after hospital discharge are indicated. The premature infant discharge
formulas have increased energy, protein, calcium, phosphorus, iron, vitamin A,
and vitamin D compared to term infant formulas. Infants on these formulas have
more rapid catch-up growth and mineralization than premature infants-fed term
formula (179).
In summary, a tri-phasic system seeks to customize nutritional delivery for the
preterm infant based on physiology and nutrient needs. Future directions may
include better definition of specific nutrient needs (e.g., amino acids, growth
factors) for each phase and the development of noninflammatory formulas to be
delivered enterally during transition or periods of medical instability.
Specific Disease Effects on Nutrient Needs

Whether acute lung disease, such as hyaline membrane disease, increases the
oxygen consumption of the infant in direct proportion to the degree of respiratory
illness remains debated (51,86,98). Unlike in adults, however, nitrogen balance in
newborns appears to be unaffected by respiratory illness. The mean protein
requirement to maintain neutral nitrogen balance during respiratory disease is 1.5
to 2.0 g/kg/day. Severe respiratory illness is associated with a higher incidence of
hypocalcemia and hypoglycemia. There appears to be no indication to increase
protein delivery beyond what would normally be given.

BPD increases the resting energy expenditure by up to 30% (50), and infants with
BPD will require energy intakes up to 150 kcal/kg/day to grow adequately. Much
of the growth failure in BPD occurs in the first month, after which the rate of
weight gain can be quite similar (although at a lower percentile) to infants without
BPD (180). Protein requirements have not been extensively studied in infants with
BPD, but those who are treated with steroids have increased muscle breakdown
and a more negative nitrogen balance (101). Clearly, malnutrition plays an
important role in the genesis of the disease and the rate of recovery (181).
Sodium and potassium requirements are increased in infants with BPD treated
with salt-wasting diuretics, occasionally requiring supplementation. Calcium
balance is more tenuous as a result of more prolonged periods of parenteral
nutrition, renal calcium wasting as a result of calciuric diuretics (furosemide), and
bone demineralization as a result of corticosteroid therapy. A relationship between
vitamin A deficiency and BPD has been documented (161,163); thus, it is
important to monitor and support vitamin A status in infants with BPD.
CHF has a profound effect on resting energy expenditure equal to that of BPD.
Infants with CHF as a result of structural heart disease may require up to 150 kcal/
kg/day on the basis of both an increased metabolic rate and malabsorption as a
result of intestinal edema. The protein requirement of infants with CHF has not
been studied, although it is clear many fail to thrive and have reduced muscle
mass. It is prudent to increase the protein delivery commensurate with the
increase in energy delivery, maintaining a 25 to 30 kcal:g protein ratio. Diuretics
used to treat CHF cause the same sodium, potassium, and calcium deficiencies
seen in infants with chronic lung disease. Furthermore, the use of citrated blood
products to replace losses perioperatively can cause profound hypocalcemia.
Persistent cyanosis (See color plate) causes an additional nutritional stress by
increasing the infant's need for iron. Since many of these infants have secondary
polycythemia, there must be enough iron in the diet to support augmented
erythropoiesis. Failure of a cyanotic infant to maintain an elevated hemoglobin
may be as a result of iron deficiency, which can be screened for with a ferritin
concentration and by assessing the red cell indices for microcytosis.
The effect of sepsis on neonatal nutritional status has not been well evaluated.
Infants rarely present with the overwhelming multi-system organ failure that
adults routinely have with sepsis. This may be as a result of an incomplete
cytokine response. Septic infants have increased TNF and IL-6 levels, but their
levels are not nearly as high as in adults (52). Furthermore, neonatal sepsis does
not result in a profoundly negative nitrogen balance. Sepsis does increase oxygen
consumptionalthough this appears to be a nonspecific response to illness, as it is
seen in other nonseptic states. It appears that sepsis increases energy but not
protein needs in the neonate.
NUTRIENT DELIVERY
Almost all term infants and many preterm infants more than 33 weeks gestation
will feed orally on demand immediately after birth. Breastfed infants should be
offered the breast within 30 minutes of delivery. However, ill term infants and
preterm infants who are not physiologically mature or who are unstable will
require alternate forms of nutrient delivery. The first decision revolves around
whether the infant is stable enough to be fed enterally or if parenteral nutrition is
indicated. If long-term parenteral nutrition is anticipated, decisions will need to be
made whether a central catheter should be placed or whether the nutrients should
be given through a peripheral vein. If the infant is to be enterally gavage fed, the
practitioner has multiple options with respect to where the gavage tube is
P.393
placed, whether it remains indwelling or is replaced after each feed, and whether
the feedings are by continuous drip or bolus.
Parenteral Nutrition
Indications
Parenteral nutrition is indicated in all infants in which enteral nutrition is
contraindicated or delivers less than 75% of total protein and energy
requirements. Although parenteral nutrition has become a more refined nutritional
tool with fewer complications over the past decade, the enteral route remains the
preferred way to nourish babies. For all practical purposes, infants on ventilators
in the acute stage of their diseases are rarely enterally fed nutritionally
meaningful amounts. It is also not appropriate to simply provide these infants
with a dextrose and electrolyte solution. Anderson et al. demonstrated that the
addition of amino acids to dextrose solutions shortly after birth improves the
nutritional status of infants (97). Earlier initiation of parenteral nutrition was one
factor associated with higher weight, length, and OFC percentiles at discharge and
better long-term developmental outcome (182). Therefore, it is appropriate to
begin parenteral nutrition for infants within 24 hours of birth. The trend in the
past five years has been to begin feedings earlier in preterm infants to promote
ongoing maturity of the intestinal tract, to avoid villous atrophy as a result of
disuse, and to kindle gut hormone activity (183). Thus, infants who still require
moderate respiratory support will receive trophic feeds, but these feedings are
hypocaloric, and parenteral nutrition is indicated in these infants.
Absolute indications for parenteral nutrition include surgical lesions such as
omphalocele, gastroschisis, intestinal tract atresias (e.g., tracheoesophageal
atresia, duodenal atresia, ileal atresia), meconium peritonitis, diaphragmatic
hernia, short bowel syndrome, and Hirschsprung's disease. Medical indications
include NEC (or feeding intolerance), meconium ileus, ileus as a result of

generalized illness, infants on extracorporeal membrane oxygen (ECMO) therapy,
and preterm infants on slowly advancing feedings.
Routes of Delivery
The decision whether to supply parenteral nutrition centrally or peripherally
requires weighing the benefits versus the risks. Parenteral nutrition administered
through a central line allows for greater energy delivery because solutions with
dextrose concentrations more than 12.5% can be administered. Dextrose
concentrations of that magnitude and calcium infusions are poorly tolerated by
peripheral veins and carry a high rate of venous sclerosis (184). Skin sloughs are
likely to occur if the solution extravasates from the vein. For the same reasons,
many intensive care nurseries will not allow or will limit the amount of calcium to
be run through a peripheral venous line. This practice is sound, but effectively
limits the amount of calcium and phosphorus that can be delivered to an infant
who is already at great risk for osteopenia.
The risks of central TPN relate primarily to the risk of central venous line
placement and maintenance. Umbilical venous catheters placed at birth have
traditionally been used as the primary central catheter, but the incidence of
venous thromboses is high (185). Clots can be detected as early as 24 hours after
catheter placement. The clots are frequently infected with Staphylococcus
epidermidis, which has become the most common pathogen isolated in the ELBW
infant after seven days of age (186). Equally concerning is the high rate of
Candida septicemia seen with high dextrose delivery and high serum
concentrations (187).
One risk of peripheral TPN is undernutrition. The infant receiving maximal
concentrations of dextrose (D12.5%), amino acids (3.0 g/kg/day), and
intravenous fat (3.5 g/kg/day) at an average fluid rate of 150 cc/kg/day will
receive approximately 95 nonprotein kcal/kg/day. Although this amount of intake
meets the daily resting energy expenditure of the premature infant (65 kcal/kg/
day), there are insufficient extra calories to sustain weight gain at 15 to 18 g/kg/
day. Thus, long-term peripheral TPN will result in preterm infants slowly falling
away from the growth curve. Calcium delivery will also be constrained, either
because of an absolute contraindication (in some nurseries) or because of
osmolarity issues. Each day on peripheral TPN results in a larger deficit calcium
balance and a higher risk of osteopenia of prematurity.
Infants who are not expected to tolerate oral feedings within a week of starting
parenteral nutrition should have a central line placed and be maintained on central
parenteral nutrition. The choice of venous access also involves weighing the risks
and benefits. Surgical placement of an anchored catheter (e.g., Broviak) is a
riskier procedure than placement of a peripherally inserted central catheter
(PICC). However, it is likely that the Broviak, placed under sterile operating room
conditions, will last longer. Additionally, the choice of lines (single lumen vs.
double lumen, differences in gauges) is greater with surgically placed lines, and
frequently blood can be drawn from one of the ports for laboratory monitoring.
Conversely, the PICC lines are easily placed in the unit, can be as small as 27
gauge, and are silastic (which are less prone to clotting and infection). Their
disadvantage is that they generally cannot be used for blood drawing. In our unit
we try to remove all umbilical venous catheters after the infant has been
stabilized following delivery room resuscitation and attempt to place a
percutaneously inserted central catheters (PICC) line within the first 24 to 48
hours if the infant is expected to be on parenteral nutrition. Because of the low
incidence of infection with these lines, we typically do not use peripheral
parenteral nutrition.
Maintaining patency of the lines is important for the success of parenteral
nutrition. Most centers use heparin in TPN solutions to keep central lines patent
and reduce the formation of a fibrin sheath around the catheters. These
P.394
fibrin sheaths are most likely to be the site of infectious agents. In order to
maintain patency, our unit utilizes the following protocol for PICC lines: for flow
rates more than 7 cc/hour, no heparin is used; for flow between 2 and 7 cc/hr,
0.25 U of heparin/cc is added; for flow rates more than 2 cc/hour, 0.5 U of
heparin/cc is added. We do not heparin-lock our PICC lines and run them with a
minimum rate of 0.5 cc/hour with 1 U heparin/cc of solution. We do heparin-lock
surgically placed lines, administering 10 Units of heparin in 1 cc of solution, given
every 12 hours.
Catheter occlusions are usually treated with removal of the line because so many
of the clots are infected. Nevertheless, catheter occlusions that occur without
signs of sepsis (e.g., if the line was inadvertently shut off) can be treated with
urokinase (5000 U/mL). The amount of urokinase solution should approximate the
internal volume of the catheter (0.2 to 0.5 mL). After instillation, the solution
should be allowed to dwell in the catheter for 30 minutes. If two attempts at
clearing the line fail or if a positive blood culture has been obtained from the
clotted line, the catheter should be removed (188).
Nutritional Management
Parenteral nutrition should be started within 24 hours of delivery, since dextrose
solutions alone cannot meet the resting energy requirements or the protein
requirements of the neonate.
Dextrose delivery should typically begin between 4 and 6 mg/kg/minute and be
advanced as tolerated. Extremely preterm infants are frequently glucose
intolerant because of relative insulin hypoactivity and poor peripheral glucose
utilization. Although their energy needs are higher because of higher basal
metabolic rates and higher brain-liver weight ratios, they frequently develop
hyperglycemia and glycosuria. These are serious complications that must be
treated immediately. Persistent glycosuria will result in a large free-water diuresis,
intravascular dehydration, hypernatremia, and azotemia. Persistent hyperglycemia
is a significant risk factor for fungal infection. Dextrose delivery can be slowly
advanced based on how well the infant tolerates this. Typically we do not
administer more than 12.5 g/kg/day because of the significant effect on the
respiratory quotient. However, others have advocated rates up to 20 g/kg/day,
aided by the administration of insulin to maintain normoglycemia (62). As stated
earlier, I do not advocate this approach because the weight gain is predominantly
fat rather than lean body mass, and the metabolic cost of fat synthesis from
glucose is high in terms of both oxygen consumption and carbon dioxide
production. On the other hand, insulin is very useful in treating the hyperglycemia
seen in ELBW infants in the first week of life, in which glucose intolerance may
necessitate decreasing dextrose delivery to unacceptably low rates (<4 mg/kg/
minute).
Protein in the form of amino acid solutions designed for newborns should be
administered within the first 24 hours. There are few contraindications to early
protein delivery, and there is evidence that amino acid solutions improve nitrogen
balance (52,86,97,98,189). At energy intakes above resting energy expenditure
(65 kcal/kg/day), the main determinant of positive nitrogen balance is the
nitrogen intake (190). The goal is to achieve in utero nitrogen accretion rates
although compensating for nitrogen losses as a result of illness. This appears to
be possible with amino acid delivery rates of 2.7 to 4.0 g/kg/day (86,98,191).
Although protein requirements may be higher as a result of prior malnutrition, to
diseases that increase nitrogen turnover, or to catch-up growth, it is rarely
practical to give more than 4.0 g/kg/day of parenteral amino acids because of
increasing BUN concentrations. Recent studies demonstrate that administration of
amino acids is safe for all infants on day one to two. Most infants can be safely
started on at least 2 g/kg/day and advanced by 1 g/kg/day to a maximum of 4.0
g/kg/day, thus ensuring that they will be on full protein delivery within 48 hours.
Very unstable preterm infants and those with renal insufficiency as a result of
indomethacin administration, surgery, a patent ductus arteriosus, or shock may
need to be advanced more slowly. Monitoring the BUN allows the practitioner to
decide on a daily basis whether the protein delivery can increase. A rising BUN is
an indication that the infant is not clearing nitrogen waste and that the rate of
amino acid infusion should not be increased. Initially, when amino acid solutions
intended for adults were given to infants, significant complications occurred
because these solutions did not meet their metabolic needs. In the mid-1980s,
improved solutions were introduced that added the semi-essential amino acids
taurine, water-soluble tyrosine, and L-cysteine. Potentially toxic amino acids such
as phenylalanine and glycine were reduced. These newer solutions promote a
more normal serum amino acid profile (103), better nitrogen retention and weight
gain (190,191,192,193), and lower rates of cholestasis (194). The lower rates of
cholestasis may be as a result of the addition of taurine, which may also be
important in neuronal development (194,195,196). Nevertheless, the neonatal
parenteral amino acid solutions are not perfectly formulated, and investigations
continue to determine whether threonine, lysine, and glutamine are limiting amino
acids (100). There is no evidence that specialized solutions such as HepatAmine,
BranchAmine or NephrAmine are indicated in newborns and the spectrum of
amino acids found in them may dangerously imbalance an infant's serum amino
acid profile.
Intravenous fats provide a low-volume source of calories and shift cellular
metabolism toward less carbon dioxide production, perhaps improving the
respiratory load of the infant. They can be utilized within the first three days of life
and are important in preventing essential fatty acid deficiency (197). Close
monitoring of serum triglyceride levels is important during intravenous fat
therapy. Intravenous fat solutions can be started at a delivery rate of 1 g/kg/day
and advanced to a maximum of 4 g/kg/day. Total fat calories should be less than
60% of the diet and typically are in the 30% to 40% range. Like amino acids,
intravenous fats can be advanced by 1 g/kg/day if tolerated. Since fat
incorporation into cells is dependent on insulin, fat intolerance in
P.395
VLBW infants is more likely to be manifested by hypertriglyceridemia or,
interestingly, hyperglycemia, requiring a slower rate of advancement (0.5 g/kg/
day) or interruption of fat delivery. Infants with birth weights less than 1,250 g
and gestational ages less than 30 weeks may need to have their fat dose held at 1
g/kg/day until their hyperbilirubinemia begins to resolve. This group of infants
seems to be at greatest risk for intravenous lipids exacerbating hyperbilirubinemia
(see TPN complications). Fat emulsions are predominantly 20% solutions and are
generally infused over no fewer than 16 hours to allow for metabolic clearing. It is
important both to run them separately from other solutions, so as not to disturb
the stability of the emulsion, and to cover the solution from light, to decrease
breakdown. The solutions can be joined with the amino-acid-containing solution
with a Y-connector near the infusion point on the infant.
Since infants initially undergo a free-water diuresis before a salt diuresis, sodium
needs remain low until after day three of life (198). Thereafter, sodium and
potassium requirements increase rapidly and serum concentrations should be
monitored at least daily although infants are on intravenous solutions. Table 22-3
lists the approximate electrolyte requirements for premature infants in the face of
no extraneous losses, such as those incurred by renal failure or diuretic therapy.
However, requirements may approach 10 mEq/kg/day for each if there are
excessive urinary losses. Chloride is the usual anion for both sodium and
potassium; however, these cations also can be given as acetates, allowing for fine
tuning of acid-base balance. Amino acid solutions have an inherent chloride and
acetate load (e.g., TrophAmine contains 1 mEq of acetate for every gram of amino
acid).
Calcium and phosphorus are the most difficult minerals to maintain in positive
balance in the preterm infant because of the large requirements for adequate
mineralization, excessive losses as a result of calciuric diuretics and steroids, and
the limited solubility of these nutrients in TPN (199). A calcium-to-phosphorus
ratio of 1.7:2.0 appears to be optimal for mineralization (200). Because of
solubility issues, calcium concentrations more than 16.6 mEq/L with a
concomitant phosphorus concentration of 8.3 mM are rarely obtained. In an infant
receiving 150 cc/kg/day, these values are equivalent to a calcium delivery of 50
mg/kg/day and a phosphorus delivery of 25 mg/kg/dayfar less than the in utero
accretion rate. Strategies to increase calcium retention and bone mineralization
have been largely unsuccessful, but have included infusing calcium in one line and
phosphorus in another and alternate infusions of higher doses of the two minerals
(201,202). Monitoring of serum phosphorus and calcium levels is important.
Infants are prone to hypocalcemia in the first 72 hours as a result of transient
hypoparathyroidism and to hypophosphatemia. Both calcium and phosphorus
should be added early during TPN therapy. Calcium delivery without phosphorus
delivery should be avoided because of the likelihood of hypophosphatemia. This
complication tends to occur in the first 72 to 96 hours because of the focus on the
diagnosis and treatment of neonatal hypocalcemia. More acidic TPN solutions
appear less likely to cause calcium-phosphorous precipitation (203,204).
TABLE 22-3 DAILY REQUIREMENTS OF TOTAL PARENTERAL NUTRITION
Nutrient
Protein
Fat emulsion
Calories
H20
Na
K
Ca
P
Mg
Multivitamins (e.g., MVI Pediatric)
Requirement
2.53.5 g/kg
24 g/kg (max 3.5 g/kg in infants
<2.5 kg)
90110 kcal/kg or as needed
125150 mL/kg or as needed
34 mEq/kg
23 mEq/kg
50100 mg/kg, depending on size of
infant
11.5 mM/kg
0.51 mEq/kg
10 mL (40%/kg/d)
Infants on TPN receive 0.2 mL/kg body weight of a neonatal trace element
solution that supplies 0.02 mg/kg of copper, 0.3 mg/kg of zinc, 5 g/kg of
manganese, and 0.17 g/kg of chromium. This supplement should be added with
initiation of TPN and given daily. Selenium should be added after 2 weeks of TPN
(149). Although we do not routinely measure zinc, copper, chromium,
manganese, or selenium levels in infants on TPN, the practitioner should be aware
that preterm infants in particular have low stores of these trace elements and that
deficiencies have been described (205,206,207). Water- and fat-soluble vitamins
are added as a pediatric multi-vitamin solution to match the recommended
parenteral dosing guidelines (149,171). This supplement should be added at
initiation of TPN and given daily.
Complications of TPN
Administration of parenteral nutrition remains an inexact science. Because it is not
the normal mode of nutritional delivery, it is not surprising that complications
occur. For the most part, complications can be divided into those associated with
catheters and those related to the nutrients themselves. As discussed above,
centrally placed catheters are prone to thrombosis and infection. Thrombi can
occur in the right atrium or in the veins. Clotting in the superior vena cava is of
particular concern. Superior vena cava syndrome with or without hydrocephalus
can result. Occasionally, back pressure from the clot on the thoracic duct will
cause a chylothorax. Additionally, nonseptic complications such as skin sloughs
can occur. Erosion of catheters through vessels or cardiac walls has caused pleural
effusions, pericardial effusions, and endothelial damage. Improper technique of
catheter insertion can result in pneumothorax or nerve injuries, although improper
handling of fluids can result in air or fat embolisms. It is important for all
personnel involved in parenteral nutrition
P.396
therapy, including nurses, pharmacists, and physicians, to be aware of the
complications and the techniques to avoid them.
Intravenous lipids have been associated with hypoxia, pulmonary hypertension,
hyperbilirubinemia, and infection (208). Infants with respiratory disease have
minimally lower PaO2 values when given intravenous lipids, most likely because
lipids can uncouple hypoxic vasoconstriction (209). Normally, to optimize
ventilation/perfusion matching, the pulmonary vasculature supplying a poorly
oxygenated alveolar area will constrict. This effect is reduced by the infusion of
lipids, most likely moderated by serotonin. Similarly, higher pulmonary arterial
pressures are seen in neonatal lambs infused with pharmacological doses of lipids
(6 mg/kg over 1-4 hours) (210). Finally, trials that have assessed whether early
administration of intravenous lipids causes chronic lung disease (211,212) have
had mixed results. Overall, given the profound and early onset of growth failure in
infants with severe lung disease, it seems prudent to start small amounts of lipids
early in life.
Free fatty acids can displace bilirubin from albumin-binding sites, prompting some
practitioners to limit the dose of lipids to very small preterm infants. A study of
infants weighing 670 to 3360 g demonstrated adequate albumin binding of
bilirubin and no effect on serum bilirubin levels (213,214). There are no reports of
fat emulsions increasing the incidence of kernicterus. Theoretically, chylomicrons

can be taken up by the reticuloendothelial system and interfere with fighting
infection. Fat emulsions are also good media for fungi, including Candida albicans
and Malassezia furfur (215). Whether these risks clinically outweigh the benefits of
higher energy intake for small preterm infants has not been studied. At this time it
is likely that intravenous lipids improve the survival of infants through better
growth.
Parenteral amino acids also are associated with toxicity. Excessive amino acid
delivery will lead to increased serum BUN and ammonia levels as a result of the
newborn infant's relatively immature renal and hepatic status. Parenteral amino
acids also have been associated with cholestasis, although the mechanism
remains unknown (216). Infants at greatest risk are those who receive TPN for
more than 3 weeks, those with intestinal disease, particularly NEC, those with
sepsis, and those who remain NPO. Small amounts of trophic feeds reduce the
prevalence of cholestasis, most likely by stimulating bile flow via cholecystokinin.
Reduced bile flow from prolonged TPN is frequently associated with gallstones. In
rare cases, prolonged TPN with no enteral intake will lead to cirrhosis.
Aluminum toxicity is worth considering in infants who have been on TPN for more
than 3 weeks. The largest contamination comes from the calcium and phosphorus
salts that are added (217). The risk to the neonate is two-fold. Aluminum
accumulates in the bones of infants on TPN in which it is avidly taken up because
of underlying osteopenia of prematurity (218). Of greater concern is the possibility
that aluminum will cross the blood-brain barrier and induce an acute or chronic
encephalopathy, as has been described in adult patients (219). Reduced renal
capacity for excreting aluminum appears to be a necessary setting for this to
occur, but it is not unusual for preterm infants to have a significant measure of
renal insufficiency after treatment with indomethacin (220). Some have proposed
that aluminum toxicity may be a factor in the poorer neurodevelopment of
preterm infants. Since the body has no need for aluminum, manufacturers are
being pressured to reduce the aluminum content of their solutions (221).
Monitoring TPN Efficacy and Toxicity

Administering TPN is an inexact science still in the process of evolution. Matching
nutrient delivery to the infant's needs presupposes knowledge of how a particular
disease or its treatment affects the infant's metabolism and, therefore, nutritional
requirements. Careful monitoring of growth is indicated for any infant on TPN or
partial parenteral nutrition. Weight should be measured daily and length and head
circumference weekly. Standards have been published for arm muscle area and
arm fat area for preterm and term infants, and these areas can be measured
using a tape measure and skin caliper (180,222). Protein status can be assessed
in two ways: somatic protein deposition (arm muscle area) and serum protein
synthesis (serum albumin and prealbumin). The former provides a longitudinal
view of protein accretion, although the latter reflects a more rapidly turned-over
pool of protein. Assessment of serum concentrations of proteins with short halflives such as prealbumin has been shown to reflect recent protein intake and to
predict future weight gain (223). Prealbumin, also known as transthyretin, has a
half-life of 1.9 days and can be measured once or twice per week to yield useful
nutritional information. If the serum concentration remains stable or increases,
one can expect that the infant is in reasonable nitrogen balance and will gain
weight subsequently (224). A decrease of more than 10% from the previous
measurement suggests relative protein-energy malnutrition and the need for a
higher intake. Like most rapidly turned-over proteins, prealbumin acts as an acutephase reactant and will rise rapidly with stress, infection, and glucocorticosteroid
administration, rendering it useless as a nutritional marker. Serum albumin, which
has a half-life of 21 days, can be monitored every two to four weeks.
It is important to monitor infants on parenteral nutrition because of the toxicities
associated with its administration. Table 22-4 provides guidelines for nutritional
monitoring of infants on TPN. At the least, infants on TPN should have a set of
electrolytes and a serum glucose checked daily. Serum glucose concentrations
more than 110 mg/dL are an indication not to increase the glucose infusion rate;
concentrations more than 150 mg/dL are an indication to reduce the rate. Serum
triglyceride concentrations should be checked at least twice per week, or more
frequently if the infant is showing signs of lipid intolerance. ELBW infants and
infants with sepsis are especially
P.397
prone to hypertriglyceridemia, even if they have tolerated intravenous fat
previously. Methylxanthines and glucocorticosteroids increase the likelihood of
both glucose and fat intolerance in preterm infants. A triglyceride level more than
150 mg/dL measured with the infant off lipid infusion is an indicator of impending
intolerance and lipid doses should not be increased. A triglyceride concentration
more than 200 mg/dL is considered a sign of intolerance and the lipid dose should
be decreased. Since persistent hypertriglyceridemia represents a risk to the
pulmonary system, daily serum triglyceride levels should be monitored in the
infant who exhibits intolerance.
TABLE 22-4 SUGGESTED MONITORING FOR TOTAL PARENTERAL
NUTRITION
Variable
First Week
Later
Growth
Weight
Length and head
Daily
Weekly
Daily
Weekly
circumference
Chemistry
Na, K, Cl, CO2
Glucose (Chemstrip bG)
Triglycerides
Ca (ionized Ca is most
accurate)
P
Daily until stable Daily

Daily
Daily
With each
increase in IL
Twice weekly when stable

Weekly
Daily until stable

Initially
Twice weekly for first

week,
then weekly
Albumin
Prealbumin
Initially
Initially
Monthly
Weekly (biweekly) in
infants <1000 g
Alkaline phosphatase
Bilirubin
Initially
Initially
Weekly
Every 4 wk or PRN
Mg
Ammonia
Gamma GT
Initially
As needed
Initially
Weekly
As needed
Weekly
Alanine aminotransferase
Amino acids
Zinc
Serum osmolarity
As needed
As needed
Monthly
Monthly
Monthly
Weekly
Vitamin A (if infant is <

1,300 g)
Hematology
Complete blood count
Type and screen
Urinalysis
Sugar
Protein
Specific gravity
Initially
Weekly
Weekly while
supplemented
Initially
Initially
At least weekly
Each void
Each void
Each void
Each shift
Each shift
Each shift
Calcium status must be monitored carefully in the first days of postnatal life
because hypocalcemia is commonly seen in ill newborns. Preterm infants, growthretarded infants, and infants of diabetic mothers appear particularly prone to
hypocalcemia. Infants receiving large amounts of citrated blood products, such as
those who are postoperative, who are on ECMO or who have disseminated
intravascular coagulopathy, will require large amounts of calcium. Similarly,

maintenance of normophosphatemia is important for normal metabolism.
Therefore, serum calcium, phosphorus, and magnesium should be monitored daily
in the first week of life, or until stable, and then weekly thereafter (Table 22-4).
Bone mineralization is problematic for the preterm infant on long-term parenteral
nutrition; therefore, close monitoring is indicated. Unfortunately, this can be quite
difficult. Although osteopenia of prematurity is predominantly as a result of
deficient intakes of calcium and phosphorus, the serum levels of these minerals
will be maintained at the expense of the bones. Thus, serial measurements of
calcium and phosphorus are not useful in monitoring this complication. Serum
alkaline phosphatase is an indirect measure of osteopenia because its level will
increase with the bone remodeling that takes place to supply the serum calcium
pool. The level should be monitored weekly, particularly in preterm and growthretarded infants. The level may be difficult to interpret, since the alkaline
phosphatase will rise with cholestatic liver disease (a complication of parenteral
nutrition itself) and with intestinal injury (such as NEC). Fraction-ating the alkaline
phosphatase level into its bone and nonbone components can be done, but may
take weeks depending on whether the laboratory has the capability to perform the
fractionation. As with monitoring of the prealbumin concentration, the most
important aspect of the alkaline phosphatase to follow is the trend. Rising alkaline
phosphatase levels generally mean aggressive bone remodeling and an increased
risk of osteopenia.
P.398
Strategies to increase calcium and phosphorus delivery should be considered.
The incidence and severity of hepatic toxicity from parenteral nutrition has been
on the decline with the introduction of more specialized neonatal amino acid
solutions. The toxicity is typically cholestatic in nature, with an initial rise in serum
bile acids followed by an increase in the direct bilirubin, alkaline phosphatase, and
gamma glutamyl transferase. Transaminase elevations are seen only in very
severe cases. The liver toxicity was initially thought to be as a result of
intravenous lipids (fatty liver), but has now quite conclusively been associated
with amino acid solutions. Although these solutions have been refined (192,193),
the problem persists and its cause is unknown. Total and direct bilirubin
concentrations will typically be monitored in all newborns in the first week of life.
Infants on prolonged TPN should have direct bilirubin measured weekly. If it is
elevated, the remaining liver function tests should be assayed and followed
weekly.
Trace minerals are rarely deficient in infants on TPN because of supplementation.
Nevertheless, the importance of maintaining normal zinc status for growth and
protein utilization (225) makes it wise to monitor the serum zinc concentration
monthly, particularly if the infant is not growing adequately or has physical signs
of zinc deficiency.
With the exception of vitamins E and A, vitamin status generally need not be
checked in infants on TPN. Most vitamin assays are cumbersome and are a poor
reflection of total body load. Serum vitamin E and A levels also do not necessarily
reflect total body stores. Nevertheless, the association of low serum retinol
(circulating vitamin A) levels with an increased risk for BPD in the VLBW infant
suggests that monitoring may be appropriate (161,163). An initial measurement
in all infants less than 1500 g with respiratory disease should indicate the degree
of risk. Infants with levels less than 20 g/dL should be supplemented and their
levels followed weekly. The methodologies for assaying vitamin A (high
performance liquid chromatography or fluorometry) are the same as for vitamin E
and the values for both can be done simultaneously. As with vitamin A, it is
important to keep vitamin E concentrations in the normal range, as an insufficient
concentration has been associated with anemia (164) and perhaps ROP (165),
although toxic levels increase the risk of sepsis and NEC (170).
Enteral Nutrition
Oral Feeding
The goal for virtually all infants prior to discharge from the hospital is full oral
feedings, preferably by breast. Oral feedings come naturally to infants born at
term, but can be a significant task for those born at less than 34 weeks gestation,
those with significant central nervous system disease, and those with anatomical
abnormalities that prevent oral feedings.
Oral feedings should be initiated within half an hour of birth by placing the infant
to the mother's breast. Infants who are breastfed will have a different sucking
motion than those who are bottle fed. Thus, it is important that artificial nipples
(and probably pacifiers) not be introduced as the infant is establishing
breastfeeding (7). The oral pattern associated with breastfeeding is typically well
established within two weeks, although a substantial number of infants who are
then supplemented with bottles will demonstrate nipple confusion and may give
up on breastfeeding. The healthy breastfed infant has no need for supplemental
water, juice, or formula (7). Breastfeeding can be supported in the delivery
hospital by training all of the staff to encourage mothers to breast-feed and to
provide the necessary environment to promote breastfeeding. This includes
allowing the mother to nurse within half an hour of delivery, having the infant
room with the mother, and having the mother learn the cues of her infant's
hunger. Hospitals can contribute by eliminating policies about supplementing
breastfed babies and by supplying charts that assess the infant's feeding and
hydration status (7).
Oral feedings can be more problematic for the healthy premature infant. These
infants rarely show any interest in oral feeding until approximately 32 weeks
gestation and rarely have a mature, safe feeding pattern until 34 weeks gestation
(9). Coordination of sucking, swallowing, and breathing is most difficult; the issue
is predominantly one of inappropriate swallow-respiration interface rather than
suck-swallow interaction (226). There is little evidence that practice helps the
gestationally immature infant to feed orally sooner. Nevertheless, Meier has
reported that breastfed premature infants have longer periods of sucking with
fewer obstructive apnea and desaturation spells than comparably sized bottle-fed
infants (227). This may relate to the more metered rate of milk flow. It is
important to note that preterm infants are frequently exposed to pacifiers to
stimulate nonnutritive sucking, which improves gastric motility and likely
increases the flow of important gastrointestinal hormones (228,229,230). It is
unclear whether this nonnutritive sucking at an earlier postconceptional age
affects the success of breastfeeding at 34 weeks gestation.
A strong case can be made for feeding breast milk to the preterm infant, either by
gavage tube or by breastfeeding, because of its superior performance with respect
to immune status and neurodevelopment, among other advantages (19,227,231).
In order to successfully breast-feed the premature infant, the mother needs to be
available to begin the process as the infant nears 33 weeks gestation. Before that
point, it is important that she maintain her milk supply. The intensive care nursery
can help by providing a place to nurse, an electric breast pump, storage
containers, and a freezer for storing the milk. An organized program with an
informed leader is quite useful in timing the introduction of actual breastfeeding
and in overseeing the progress made by the individual infant. With such a
program, more than 60% of preterm infants whose mothers desire to nurse can
successfully breast-feed at the time of discharge.
Preterm infants who are bottle fed also require close observation as they
transition from gavage to nipple feeds. There is an energy cost to bottle feeding.
Gavage feedings
P.399
require between 4% and 17% less energy to process, and excessive oral feedings
may tire an infant and reduce weight gain velocity (232,233). Typically, attempts
at bottling should begin around 33 weeks' gestation with one feeding per day. If
the infant shows no interest or has significant obstructive apnea, it may be
prudent to wait several days before attempting again. The frequency of feedings
can be increased as the infant shows more aptitude. Once the infant has advanced
to full oral feedings, it is important to see whether weight gain can be maintained
on an ad libitum on-demand schedule prior to discharge. Consistency in feeding
personnel can improve the infant's performance, and in the best of all worlds,
having the mother give most of the feedings is ideal. Adopting a cue-based
feeding program, in which personnel pay attention to the feeding cues exhibited
by the infant, may reduce the number of episodes in which the infant refuses to
feed because of fatigue and aversion.
Oral aversion is a significant problem in infants who have been NPO or on

ventilators for long periods of time. Symptoms include aversive behaviors such as
tongue-thrusting, head-turning, pooling of milk in the mouth, and occasionally
breath-holding apneic spells. A barium swallowing study with fluoroscopy can help
identify whether the problem is anatomical, or is as a result of discoordination,
immaturity, or neurological pathology. For infants with severe cases, the worst
course is to force oral feedings. The involvement of an occupational or speech
therapist can be invaluable in desensitizing the oral area.
Gavage Feeding
Gavage feedings are indicated for infants who can be fed enterally but not orally.
For the most part, this approach is used in premature infants who are
neurologically immature and the full expectation is that they will feed orally.
Infants who will not be candidates for oral feeds either because of anatomical or
neurological conditions can have gastrostomy tubes placed. Gavage feedings are
most frequently accomplished by placing a naso- or orogastric tube and bolusing
feedings intermittently. Some practitioners prefer to place an indwelling
transpyloric tube to reduce aspirates and to ensure nutrient delivery. Infants can
receive feeds by continuous drip and by bolus.
Oro- or nasogastric tube feedings can be initiated using a soft silastic 5-French or
8-French catheter. The tube is most commonly placed into the stomach prior to a
feeding and the contents of the stomach aspirated to ensure there are no
residuals from the previous feeding. The feeding is allowed to run in by gravity,
although in some infants with very slow gastric emptying the feeding can be
titrated in over one hour. The tube is typically withdrawn rapidly after the feeding,
although there is some increased risk of the infant vomiting in response to this
stimulus. A long-term indwelling catheter can be placed, but this type of catheter
may lose flexibility over time and increase the risk of stomach perforation.
Gastric gavage feedings can be given on a schedule between every one and four
hours. Typically, smaller infants do not tolerate excessive stomach distention with
large-volume feedings and may exhibit respiratory compromise. They may need
to be fed small amounts on a more frequent schedule. Infants less than 1,000 g
can be fed on a bolus schedule of every one to two hours or with continuous-drip
feedings. This approach may reduce oxygen consumption and total energy
expenditure, and potentially contribute to faster rates of weight gain. Infants may
be fed on this schedule until 1,250 to 1,500 g, after which every-3-hour feeds are
more appropriate. Nevertheless, larger infants who are not tolerating bolus
feedings, those who remain on ventilators, or those with severe apnea and
bradycardia may require drip feedings. Term infants who require gavage feedings
may do best on an every-four-hour schedule.
Tube placement and maintenance may cause significant problems in the infant.
Tubes can be malpositioned in the airway instead of in the stomach. With the
placement of any new tube it is important to document its position by auscultation
and by checking the pH of aspirated stomach contents. Placement of the tube can
cause significant vagal stimulation that results in apnea or bradycardia. The
presence of an indwelling tube can cause apnea and bradycardia either by
excessive vagal stimulation or, more commonly, by upper airway obstruction.
Although nasogastric tubes are more stable, they appear to cause more problems
with airway obstruction. Gastric and esophageal perforations are rare but must be
considered if there is a significant change in the infant's behavior or physical exam.
Gavage feedings can also be given through a transpyloric tube. The advantages of
this type of feeding include ensured nutrient delivery and a smaller chance of GER
and aspiration pneumonia. There are significant mechanical and nutritional
disadvantages to this approach. The mechanical problems include the difficulty of
placing the tube, although this becomes easier with practice. In order to place the
tube, the infant is turned with his or her right side down and the tube is inserted
into the stomach with small amounts of injected air. As the infant remains in the
right-side-down position, the tube has a reasonable chance of advancing through
the pylorus into the duodenum. The tube has reached the duodenum when bilestained fluid is returned or when the pH of the aspirated fluid changes from acidic
(pH 3) to alkaline (pH 5-7). Infants on histamine-2 blocking agents cannot be
assessed in this way. The procedure can also be done in the radiology suite under
fluoroscopy with a weighted tube. The position of the tube is confirmed on x-ray.
Frequently, the tip of the tube will curl back on itself or simply move back into the
stomach, and the process will need to be repeated. Although rare, the most
devastating complication of transpyloric feedings is intestinal perforation and
peritonitis.
Transpyloric feedings also pose significant nutritional risks (233,234,235).
Bypassing the stomach decreases fat digestion and absorption, since up to 50% of
fat processing takes place in the stomach by the lingual and gastric lipase
enzymes. Additionally, secretion of gut hormones such as cholecystokinin and
gastrin are dependent in part on stomach distention by a meal. Potassium
accretion may
P.400
be impaired. Bacterial colonization of the normally sterile intestine may be a
significant risk since the normal mechanism by which the stomach acid kills
bacteria has been bypassed.
TABLE 22-5 ORAL FEEDING SCHEDULE FOR THE LOW-BIRTH-WEIGHT INFANT

Time
First feeding
Substance
1,000 g
Amount
Full-strength
12 mL/kg
human milk or
Frequency
1,0011,5000 g
Amount
Frequency
1,5012,000 g
Amount
Frequency
> 2000 g
Amount
Frequency
12 h or
13 mL/kg
continuous drip
2h
34 mL/kg
23 h
10 mL/kg
3h
(full strength)
2h
2h
Increase 2
mL every
23 h
Increase 5
mL every
1/4 strength
formula
Subsequent
feedings, 12
Formula or full- Increase 1

strength
mL every
72 h
human milk
Final feeding
Full-strength
schedule, 150
formula or
mL/kg
Total time to
full feeds
human milk
Increase 1
mL every
other feeding
to maximum
other feeding
to maximum
other feeding
to maximum
other feeding
to maximum
of 5 mL
of 20 mL 20
28 mL
of 15 mL 28
37 mL
of 20 mL
1015 mL
2h
23 h
3h
3750 mL,
3h
34 h
then ad
libitum
1014 d or
more for
710 d
57 d
35 d
infants <705 g
a
Supplemental intravenous fluids should be given to fulfill requirements of 140160 mL/kg and caloric requirements of90130 cal/kg
Initiation of gavage feedings through any of the tubes mentioned above requires a
careful assessment of the infant. The stable infant more than 1500 g birth weight
can typically be fed within hours of birth, although if the infant is less than 35
weeks gestation it is prudent to advance the strength and volume of feedings in a
proscribed manner. Table 22-5 provides a sample of feeding schedules in stable
infants based on birth weight. Advancement at a rate of 20 cc/kg body weight per
day appears to be safe as long as the infant shows no signs of feeding intolerance.
Interestingly, a meta-analysis of randomized or quasi-randomized trials of rapid
versus slow rates of advancement revealed that more rapid rates were associated
with a shorter time to regain birth weight and to achieve full enteral feedings
without an increase in morbidity (236). Infants with birth weights more than
1,500 g can be started on every-3-hour feedings; infants between 1,000 and
1,500 g on every-two-hour feedings; and infants less than 1,000 g on every-onehour, two-hour, or continuous-drip feedings. It must be remembered that
although low-volume feedings are better tolerated from a respiratory standpoint,
the gastric emptying time of the preterm infant is often between 60 and 90
minutes. Therefore, it is likely that gastric aspirates will be present in an infant fed
every hour or by continuous drip. In the infant fed every two hours or less
frequently, gastric aspirates should be less than 2 cc/kg body weight. Aspirates
greater than that amount may be indicative of an ileus as a result of feeding
intolerance or impending NEC. A thorough evaluation including an abdominal
examination is indicated before resuming feedings. The availability and ease of
administration of parenteral nutrition makes a strong argument for being
conservative with feeding advancement in preterm infants.
Trophic Feeds
Slow advancement of feedings is recommended in an infant who has been ill and
likely had an ileus. The trend in the last ten years has been to start with trophic
feeds in infants who in the past would otherwise have remained NPO. Trophic
feedings are defined as continuous-drip feedings at 1 cc/hour or less. Studies of
VLBW infants begun on trophic feedings in the first week of life have shown a
lower incidence of feeding intolerance and NEC, a more mature gastrointestinal
tract, and a shorter duration of time to regain birth weight
(237,238,239,240,241). Animal studies demonstrate that early feedings prevent
involution of the gut villi and loss of intestinal enzymes normally seen as few as
three days after beginning intravenous feedings (242). Trophic feedings can be
considered more as oral medication than as true feedings because little is gained
nutritionally from them. Trophic feedings have not been studied in infants less
than 800 g birth weight, and it is unclear whether the benefits, if any, of early
feedings in these infants would outweigh the risks. Most practitioners agree that
these infants should not receive feedings although they are unstable. However,
mechanical ventilation or the presence of an umbilical arterial catheter per se is
not absolute contraindication to initiating feedings.
WHAT TO FEED NEWBORN INFANTS

What one decides to feed infants is dependent on understanding the
developmental physiology of the newborn
P.401
gastrointestinal tract, the requirements of the infant for normal growth and body
composition, and the available mechanisms of nutrient delivery. It is not
surprising that term infants will thrive on different amounts and types of foods
than will preterm infants, and that allowances in both groups must be made for
the effect of illness on nutrient requirements.
Term Infants
Human Milk
Human milk is species-specific food for human beings (7). As such, it represents
the best choice of food for the newborn infant. Substitute feedings, usually made
from an animal-milk base, have been available for hundreds of years and have
been highly refined in the past century. Nevertheless, no manufactured food can
match the content of human milk for several reasons. Human milk is delivered
fresh and has no shelf life. This simple property allows live cells, growth factors,
enzymes, and immune factors to remain intact and active. Formulas, which are
designed to have a shelf life of one to two years (depending on the type of
formulation), do not incorporate most of these factors because they would be
unstable and would degrade over time. The factors found in human milk are
thought to be responsible for many of its immunological and developmental
advantages. Human milk is always at the correct temperature and requires no
sterilization.
Approximately 69% of women in the United States elect to breast-feed their
infants (243). This figure has remained relatively stable during the past five years
and represents a rise from the nadir of 40% in the 1950s. It falls short of the goal
of 75% set by the Healthy People 2000 initiative sponsored by the National
Institutes of Health and endorsed by the AAP (7). The obstacles to improving the
rate of initiation of breastfeeding include physician apathy or misinformation
(244), insufficient prenatal breastfeeding education (245), and the lack of a
perception of breastfeeding as culturally normal (246). By 6 months of age, only
33% of infants are breastfed even though human milk is nutritionally sufficient for
infants through the first six months (243). This figure falls far short of the Healthy
People 2000 goal of 50%. The decrement is due primarily to failure to maintain a
milk supply in the first days after birth and discontinuation of breastfeeding upon
the mother's return to work, typically at 6 or 12 weeks postpartum. The former
relates to hospital and office practices that encourage formula feeding or are, at
best, ambivalent to breastfeeding. For example, early hospital discharges
combined with lack of timely routine follow-up care and postpartum health visits
contribute to this early loss (247,248). The late dropout relates to the fact that
many mothers work and many workplaces are not equipped to support the mother
to maintain her milk supply (249,250,251). The AAP reaffirmed its support of
breastfeeding and provided recommendations to improve the initiation and
retention rates (7).
There are few absolute contraindications to breastfeeding. Infants with
galactosemia should not be breastfed (252), nor should infants whose mothers
are using illegal drugs (253). Mothers with active tuberculosis and mothers in first
world countries who have human immunodeficiency virus (HIV) should also not
breast-feed (254,255). Mothers who are taking certain medications (e.g.,
amethopterin, bromocriptine, cimetidine, clemastine, cyclophosphamide,
ergotamine, gold salts, methimazole, phenindione, thiouracil) should not breastfeed. Complete lists of maternal medications that contraindicate breastfeeding are
available (256,257). Temporary disorders, such as maternal mastitis or
engorgement, are not contraindications to breastfeeding.
Human milk is nutritionally complete for most term infants for the first 6 months
of life. Its primary carbohydrate is lactose. The protein content is low (1.1%) but
the amino acid spectrum is well matched for the human infants' needs, and the
predominance of lactalbumin ensures a low curd tension. The fat content of
human milk is high and may approach 55% of total calories. The fat blend is
unique and has been difficult to imitate in formula. In particular, the presence of
certain omega fatty acids (DHA and ARA) may be important for optimal retinal
and neurological development (258). Human milk is relatively low in sodium and
osmolality. Not surprisingly, gastric emptying is rapid with human milk, making it
ideal for infants with slow gastrointestinal motility as a result of illness. Human
milk is relatively low in iron content but high in iron bioavailablity. The vast
majority of term infants exclusively fed human milk will remain iron sufficient,
although their stores at 6 months may be lower than infants fed ironsupplemented formula (124). Human milk may have a low-vitamin D content,
particularly in women consuming a diet low in vitamin D and having low exposure
to sunshine. Their infants are at risk of developing rickets if they too are not
exposed to sunshine or given a vitamin D supplement. The AAP now recommends
that all breastfed infants receive 200 IU of vitamin D daily.
Epidemiological studies provide evidence for the advantages of human milk over
formulas, including better immune status, fewer infections (259,260,261,262),
greater psychological benefits, more rapid neurodevelopment (263), protection
from chronic childhood diseases (264,265,266), protection for the mother from
certain diseases (267,268), and a lower rate of allergic disease (269,270). The
reader is referred to the AAP statement on breastfeeding for a more complete
review of these benefits (7).
Infant Formula
Many women choose formula feeding instead of breastfeeding for their infants.
Infant formulas promote excellent growth and development when used as an
alternative to breastfeeding. They should be given for the first year (271).
Formula manufacturers are continuously attempting to improve their products,
with the goal of matching human milk composition or performance. Most infant
formulas
P.402
are cow-milk based and are formulated at 20 calories per ounce. Alternatives
include soy-based formula and elemental formulas.
Carbohydrates provide approximately 40% to 45% of the calories in formula. The
most commonly used cow-milk-based formulas contain lactose as the primary
carbohydrate, whereas the soy formulas contain either sucrose or glucose
polymers.
The protein in formula provides approximately 10% of the total calories. Cow-milk
protein is casein predominant, which is reported to have a higher curd tension
than whey. Formula manufacturers have increasingly processed the cow-milk
protein to make the formulas whey predominant with the whey-to-casein ratio
approaching 60:40. The ratio in human milk is 70:30 (19). One formula in the
United States has hydrolyzed the whey. Soy formulas contain soy proteins, which
also support normal linear growth and muscle accretion. The protein content of
soy formulas is higher than that of cow-milk formula. Soy formulas contain phytic
acid, which may bind divalent cations (Ca, Mg) in the formula. For this reason, the
calcium content of soy formulas is greater than that of cow-milk formulas. Both
bone mineralization and linear bone growth in term infants fed soy formulas
appear to be adequate.
Fat constitutes 40% to 55% of calories in infant formula and is usually a blend of
vegetable oils, such as corn, coconut, soy, or palm-olein. Vegetable oils are added
to cow-milk-based formulas because babies do not tolerate butterfat well. The fat
blends are generally well tolerated, although infants will malabsorb up to 1 g/kg/
day of ingested fat in the first 10 days of life (272). This malabsorption is lower
than what is observed with whole milk (2 g/kg/day) or evaporated milk (1 to 2 g/
kg/day). Recent research has focused on whether LC-PUFAs, such as DHA and
ARA, are essential in the diets of newborns. Human milk contains these fatty acids
whereas cow milk does not. Newborn infants have a relatively limited ability to
synthesize these fats at birth, although the rates of maturation of the enzymatic
pathways (elongation and desaturation) in the postnatal period are only now
being elucidated. The content of DHA in human milk decreases rapidly after 44
weeks postconception, yet infants maintain adequate DHA levels, suggesting that
the synthetic process is intact near that age (273). Addition of DHA to term infant
formula has yielded mixed results with respect to growth and neurodevelopment
(69,70,72,77). Those studies that have shown a positive effect on early retinal
development or neurodevelopment have failed to demonstrate long-term or
permanent benefits. From a safety standpoint, the FDA has determined that it has
no questions whether the addition of LC-PUFAs derived from fungal sources are
GRAS (generally recognized as safe), as claimed by the manufacturer. Based on
this safety designation, major formula manufacturers in the United States have
added DHA and ARA to their term infant formulas, although formula without the
added PUFAs remain available.
Substantial alterations need to be made to whole cow milk to create a formula
that a newborn infant will tolerate and thrive on. Whole cow milk is highly
osmolar, low in calcium, high in phosphorus, low in vitamins A and D, and very
low in bioavailable iron. Significant adjusting of all of these nutrients, in addition
to the protein and fat manipulations, is necessary before an infant formula is safe
for newborns.
Soy formulas are indicated for infants with galactosemia or lactase deficiency,
infants whose mothers choose a vegetarian diet for their family, and infants with
documented IgE-mediated allergy to cow-milk protein (274). On the other hand,
there is no evidence that soy formula prevents atopic disease. Soy formulas do
not relieve colic and are not indicated for premature infants (see below).
Elemental and casein hydrolysate formulas continue to make up a larger part of
the infant formula market despite their very high cost and poor taste. Their main
use has been in the treatment and prevention of allergy because 90% of the
protein fragments are less than 1250 Daltons molecular weight. These low
molecular weight fragments are less antigenic than cow-milk protein. In spite of
this, anaphylaxis to these formulas has been reported (275,276). Additionally, the
rate of true cow-milk protein allergy in newborns is less than 3%. Whey
hydrolysate formulas are similar in cost to standard term infant formulas but are
not as finely hydrolyzed as the casein hydrolysates and thus may present more of
an antigenic challenge. Their ability to prevent and treat cow-milk allergy is
debatable. All hydrolysate formulas promote adequate growth and nitrogen
retention. Hydrolysates are not indicated for refeeding infants after gastroenteritis
or for treating colic. They are more osmolar than standard cow-milk or soy
formulas and thus possess a potential risk to the intestinal epithelium, particularly
in the preterm infant.
Preterm Infants
Human Milk
Extensive research has assessed the adequacy and desirability of human milk
feedings in the preterm infant. The reader is referred to a recent review of the
subject (19). This research is predicated on the argument that human milk is the
ideal food for the term neonate and that the immunological, gastrointestinal
trophic, and psychological aspects are even more relevant to the preterm infant.
When beginning human milk feedings in the preterm neonate, one must ask
whether human milk is a good match for the preterm infant's nutritional
requirements.
Mothers who deliver preterm produce a milk that has a higher protein content,
higher caloric density, higher calcium content, and higher sodium content than
milk from mothers who deliver at term (103,104,105,106). To a certain extent
these higher concentrations match the increased needs for these nutrients in
preterm infants. The composition of preterm human milk changes during the first
month postnatally and becomes more like term human milk thereafter. Table 22-1
demonstrates the change in content of preterm human milk in the first months of
life.
P.403
Human milk provides multiple nutritional advantages for the LBW infant (19). The
carbohydrate composition is predominantly lactose, but also includes
oligosaccharides that are important for intestinal host defenses (277). These
oligosaccharides may play a role in protecting the human-milk-fed premature
infant from NEC (19).
The fat blend of preterm human milk is unique and allows up to 95% absorption of
dietary fat. This is due in part to the presence of lipases in human milk, but also
apparently as a result of the fat blend. Preterm human milk also has detectable
concentrations of omega-3 and omega-6 fatty acids. These fatty acids, particularly
DHA, are important constituents of phospholipid membranes in the brain (278)
and are normally delivered transplacentally. They are not found in cow milk, but
are currently added to preterm infant formula in the United States. They do not
appear to be readily synthesized by the preterm infant from linoleic and linolenic
acid precursors and are thus considered by some to be semi-essential. Studies of
infants who receive a source of these fatty acids either from human milk or in
preterm infant formula suggest better visual acuity (71,74,279). Studies of longterm developmental outcome continue to determine whether any early
advantages fade over time.
The protein content of human milk is predominantly whey, as opposed to the
casein predominance of whole cow milk. Although preterm infant formulas are
whey predominant, there are important differences in the proteins that make up
the whey. The main human milk whey protein is alpha-lactalbumin, as opposed to
-lactalbumin in cow milk. Additionally, only human milk has significant
concentrations of important proteins involved in host defense, such as lactoferrin
and secretory IgA, in the whey fraction. These proteins may contribute to the
observed protective effect that human milk has on the occurrence of NEC. There is
evidence that these proteins act at a local (280) and systemic (281) level. Their
effects may be in combination with a more benign fecal flora (282). Human milk
contains multiple growth factors (19), including erythropoietin (283) and
epidermal growth factor (284).
In spite of these advantages, feeding human milk to preterm infants poses several
nutritional problems, particularly for the infant less than 1500 g. Preterm infants
fed unsupplemented human milk have slow growth rates and higher rates of
hyponatremia and osteopenia (285,286,287,288,289). These findings suggest
that despite the altered content of preterm human milk, there is still not enough
energy, protein, calcium, phosphorus and sodium to sustain adequate growth and
bone mineralization. There is concern that some of the energy loss occurs when
fat separates from human milk (290) and adheres to delivery tubing and storage
containers.
Rather than abandon human-milk feedings, the solutions to these nutritional
inadequacies include preventing losses by using short tubing lengths and
employing a syringe and pump, maintaining the syringe upright (19).
Most importantly, human milk delivered to all infants less than 1,500 g should be
fortified with commercial products that increase the caloric, protein, sodium, and
calcium density of the milk (19). Two preformulated powder products are
currently available. Each promotes better growth and bone mineralization than
unsupplemented preterm human milk when fortified to a presumed caloric density
of 24 kcal/ounce.
There is great variability in the milk expressed by mothers delivering preterm.
Therefore, monitoring of nutritional status is critically important in preterm infants
fed fortified human milk. In particular, growth rates, serum sodium
concentrations, and bone mineralization status (serum alkaline phosphatase
concentration, urinary excretion of phosphorus) must be assessed with regularity
in these infants. Inadequate weight gain (<15 g/kg/day consistently over one
week) can be treated by giving the infant more hind milk in the diet (19).
Persistent increases in serum alkaline phosphatase concentrations despite
fortification may necessitate adding some feedings of preterm infant formula.
Care must be taken in handling human milk to protect its important nutritional
and immunological advantages. Fresh human milk is best, but is often impractical,
particularly if the mother lives out of town. Fresh milk can be kept refrigerated up
to 24 hours, but must then be frozen. Although live cells are destroyed by deep
freezing (291), proteins remain largely intact. Suboptimal freezing results in fat
breakdown. Rewarming frozen human milk can be dangerous as microwaving
heats milk unevenly and can cause esophageal or gastric burns (292). It is more
prudent to thaw an aliquot of milk for the entire shift or day and dispense it once
it has been warmed in a water bath.
Initiating, advancing, and maintaining human milk feedings in the preterm infant
who cannot take oral feeds can be accomplished in many ways. Unlike preterm
infant formula, human milk does not need to be diluted since gastric aspirates are
less of a problem with human milk because of better gastric emptying.
Preterm Infant Formulas

The development of formulas specifically designed for the preterm infant
represented an important advancement in the nutrition of these infants. Prior to
the introduction of these formulas in the late 1970s through mid-1980s, preterm
infants were fed various formulations intended for infants with very different
intestinal maturity, nutrient assimilation capability, and nutritional requirements.
The science that went into developing preterm infant formulas carefully measured
the nutrient needs of the preterm infant (described above) and the digestive and
absorptive capabilities. When these two factors were considered together, a
unique formulation for preterm infants evolved. For the most part, the preterm
infant formulas are designed with the physiology of the less than 34 week
gestational age infant in mind. Infants more than or equal to 34 weeks whose
mothers choose not to breast-feed can be started on term infant formulas. If they
show signs of intolerance (usually diarrhea, excessive gas, abdominal distention),
P.404
relative lactase insufficiency as a result of immature intestinal development should
be suspected, and a preterm formula can be used.
Two preterm infant formulas are available in the United States, and their
nutritional constituents are remarkably similar. The carbohydrate source for both
is a combination of lactose and glucose polymers. The lactose content is reduced
compared to term infant formulas because of the relatively lower lactase
concentration found in the preterm intestine. Glucose polymers are easily digested
and are low-osmolar.
The protein source is cow milk that has been made whey predominant. The
concentration of protein is quite high, delivering up to 3.6 g/kg/day when the
formula is fed at a typical volume of 150 cc/kg/day. This high rate of delivery is
designed to match the intrauterine accretion of nitrogen and follows the guidelines
established by Heird and associates (96). Protein intakes at this rate maintain
reasonable muscle mass accretion and support normal serum albumin and
prealbumin concentrations.
As in term formulas, the fat blend is derived from vegetable oils. However,
preterm infant formulas contain between 10% and 50% of the fat content as MCT.
The necessity of MCT remains controversial (57). Addition of MCT was stimulated
by the finding that lingual and gastric lipases are particularly effective at
hydrolyzing fatty acids of this length and because long-chain fatty acids require an
adequate bile salt pool for absorption. As discussed previously, preterm infants
have low bile salt pools, which contributes to their higher fat malabsorption rate.
Excessive MCT are not indicated, as they are poorly utilized for fat storage. They
are an excellent source of energy, with the excess being excreted in the form of
dicarboxylic acids (293). As a result of the GRAS determination of LC-PUFAs in
infant formula, manufacturers of preterm infant formula are now adding DHA and
ARA to their products.
The sodium and potassium contents of preterm infant formulas are higher than
term formulas to compensate for renal tubular immaturity. Levels of trace
elements are likewise higher. The preterm infant formulas contain the most
calcium and phosphorus of any formula available. The current formulations, when
fed at a volume of 150 cc/kg/day, will provide approximately 225 mg/kg/day of
calcium and 110 mg/kg/day of phosphorus. This is well in excess of intrauterine
accretion rates, allowing these formulas to be utilized to provide catch-up bone
mineralization for those infants who have been on prolonged parenteral nutrition
or dilute formulas. In spite of this high content, most premature infants less than
1,500 g have evidence of osteopenia of prematurity at the time of discharge. The
bones of VLBW infants are frequently demineralized at discharge (294). Preterm
infant formulas have recently been supplemented with iron in recognition of the
fact that preterm infants are born with low iron stores compared to term infants,
and that a rapid expansion of the red cell mass when catch-up growth ensues
places a large stress on maintaining iron balance.
The preterm infant formulas are replete with water- and fat-soluble vitamins. Both
formulations have higher vitamin D, E, and A concentrations compared to term
formulas because of the poor fat absorption in preterm infants and the concern
about the consequences of deficiency states in the infants. Studies assessing
vitamin A and E levels in preterm infants fed the preterm infant formula with
higher vitamin A and E concentrations demonstrated that additional
supplementation with vitamins is not necessary once the infant is consuming at
least 150 cc. Infants on the product that has a lower concentration may need
additional supplementation. In either case, serum vitamin A and E levels should
be followed weekly in preterm infants less than 1,500 g birth weight.
Techniques for the initiation, advancement, and maintenance of preterm infant
formula feedings vary widely. The formula manufacturers have recommended
initiating feedings with dilute (12 kcal/ounce) formula. Most infants will be on
parenteral nutrition although their feedings are advanced. Although opinions vary
greatly regarding whether formula volume or strength should be increased first,
one should keep in mind that 1 cc of fully advanced peripheral parenteral nutrition
(D12.5%, 3.0 g/kg/day of amino acids, 3.5 g/kg/day of lipids) is equivalent to
approximately three-quarter-strength formula. Thus, volume-for-volume
substitution of TPN with half-strength formula will dilute the caloric delivery to the
infant, although substitution with full-strength formula will advance caloric intake.
Other Formulas
Although a large number of other formulas have been used for preterm infants,
none are specifically designed to meet the nutritional needs of these infants. Any
potential advantage of these formulas must be weighed against some fairly
serious side effects. For example, soy formulas were used extensively in the late
1970s and early 1980s for preterm infants because they do not contain lactose
and because of the concern that the preterm infant's intestine was particularly
permeable to translocation of antigenic milk proteins (295). However, calcium
absorption from soy formulas is very poor because the phytates in soy bind
divalent cations. The incidence of osteopenia and rickets in preterm infants who
are fed soy formulas is too high to justify recommending these products for this
population (296).
Similarly, the possibility of using elemental or casein hydrolysate formulas for
preterm infants has been suggested. The attractiveness of these formulas stems
from their more elemental nature, thus presenting less of a digestive challenge to
the immature preterm intestine. Unfortunately, these formulas are a poor
nutritional match for the preterm infant from a fat-soluble vitamin and mineral
standpoint. The vitamin E and A contents of the hydrolysates are one-quarter to
one-half that of premature infant formula. The significantly lower vitamin D levels,
lower calcium levels, and poor calcium-to-phosphorus ratio (1.4:1) place the
preterm infant at high risk for
P.405
osteopenia of prematurity. Finally, the osmolarity of these formulas ranges from
290 to 330 mOsm/L at 20 kcal/ounce, 25% higher than the preterm infant
formulas, which have osmolarities of 210 to 220 mOsm/L at 20 kcal/ounce and
250 to 270 mOsm/L at 24 kcal/ounce. Hyperosmolarity is associated with a higher
risk of NEC in preterm infants. As currently formulated, elemental or casein
hydrolysate formulas are not recommended for routine use in preterm infants.
Preterm Discharge Formulas

Follow-up formulas for preterm infants have recently been introduced. Prior to
their introduction, the preterm infant was usually switched to a formula designed
for term infants prior to discharge from the hospital. This made sense from a
digestive standpoint because most intestinal capacities are similar to term by 34
weeks postconception and preterm infants rarely leave the hospital prior to that
age. Nevertheless, this practice did not take into account the large deficits in
muscle stores, fat stores, and bone mineralization that occur in many of these
infants (3,4), nor did they take into account the high rates of growth in preterm
infants in the first year. Follow-up formulas represent a hybrid between preterm
and term infant formulas. The ones marketed in the United States are powders
designed to be diluted to 22 cal/ounce, but able to be prepared at various
concentrations. At the 22-kcal/ounce concentration, they have a 50% higher

calcium and vitamin D content than cow-milk-based formula for term infants
(297). The carbohydrate content is a blend of lactose and glucose polymers, and
the fat blends contain MCT oil in a manner similar to preterm infant formulas.
There is less vitamin A and less sodium than in preterm infant formulas, but more
vitamin A and D than in term formulas. The recommendation has been to give
these formulas for at least the first six months postdischarge, although the
manufacturers state that the formulas are safe for the entire first postnatal year.
These formulas promote better growth rates and mineral status than does term
formula in the preterm infant at follow-up (179). This is an important concept
because of the large nutritional deficits that are accrued by the preterm infant in
the neonatal intensive care unit (NICU) and the association between poor inhospital growth and neurodevelopmental disability (182,298). The protein-energy
deficit at hospital discharge at 25 grams of protein and 1000 kcal per kg of body
weight (e.g., 50 grams of protein and 2000 kcal of energy for the average 2 kg
premature infant at discharge) (4).
An alternative approach has been to continue to feed the discharged patient
premature infant formula (299). Two problems arise with this solution: the
formulas are not available commercially, and the improved fat digestive capacity
of the infant after 34 weeks postconception raises the possibility of excessive
vitamin A absorption. Moreover, the preterm infant formulas were designed for
the special physiology of the less than 1,500 g, less than 34 weekpostconceptional age infant.
NUTRITIONAL MONITORING
Any plan to nourish newborn infants should include plans for monitoring the
nutritional status. For the healthy term infant, periodic plotting of the infant's
weight, length, and head circumference on a standard growth curve is sufficient
(Appendix I). The assessment of these parameters at birth provides a metric of
the quality of fetal growth and also provides a starting point for postnatal
monitoring. Small-for-dates infants should be assessed for signs and symptoms of
intrauterine growth retardation. Signs of intrauterine wasting include small weightfor-length and mid-arm circumference:head circumference ratio (300,301), and
may be seen in many small-for-dates infants and some appropriate-for-dates
infants. It is important to plot newborns on a population-appropriate growth
curve. For example, the growth curves published by Lubchenco in the 1960s are
widely used, but were generated from a predominantly inner-city population born
at high altitude (43) (Appendix D and I). Both factors tend to be associated with
less intrauterine growth. It would not be appropriate to plot an infant of
Scandinavian heritage (in which birth weights are on average 11 oz higher than in
the United States [302,303]) born at sea level on the Lubchenco curves as it
would overestimate the rate of macrosomia. The current debate is whether
breastfed infants should be plotted on their own unique curve rather than on one
generated from a mixed or formula-fed population (53).
Similarly, appropriate curves must be used to judge the growth and nutritional
health of preterm infants. The IHDP growth curves (Infant Health and
Development Program, Ross Laboratories) have separate charts for VLBW and
LBW infants and for boys and girls (304) (Appendix I). These curves can be used
for the first two postnatal years. The weight gain velocity of the VLBW infant is not
as rapid as the LBW infant or the term infant, and little catch-up growth occurs in
the first year after discharge (305). Catch-up growth has been reported to occur
during late childhood (306).
The importance of monitoring protein-energy status in the hospitalized newborn
cannot be overemphasized. It is state of the art for NICUs with substantial
numbers of nutritionally at-risk infants to have nutrition support services that
review the infants' status at least weekly and provide nutritional
recommendations. Ideally, these nutrition support teams should include a
registered dietician, a doctor of pharmacy, and a physician. All should have a
background or additional training in neonatal nutritional principles. Studies
support the positive effect of such teams on nutritional status of the infants at
discharge (182).
Daily weights and weekly length and head circumference measurements should be
routinely performed and charted. The effect of manipulating protein-energy
delivery should be reflected with delays in the rate of weight gain. Interpretation
of protein-energy status from weight measurements can be complicated by fluid
retention or dehydration. Length measurements are the least reliable because of
the difficulty in obtaining reproducible numbers. Assessments of energy
requirements can also be made
P.406
by indirect calorimetry to estimate resting energy expenditure. These
measurements require special equipment and provide only a brief (usually 20
min) glimpse into energy utilization. The daily energy expenditure is extrapolated
from the short-term measurement with the potential errors introduced by the
extrapolation. Stable isotope techniques such as double-labeled water are the
province of research institutions and are not used for clinical monitoring. Similarly,
dual photon absorptiometry x-ray (DEXA) has been used in research studies to
assess fat and lean body mass. A more practical assessment of the infant's
relative fat status can be achieved with skinfold measurements and calculation of
the arm fat area (180).
Protein status can be assessed by measurements of somatic or serum proteins or
the serum BUN and creatinine concentrations in the absence of renal disease. The
BUN will reflect recent nitrogen intake although the creatinine will index muscle
mass. Low values are valid screening markers of poor protein status. Somatic
protein status is best reflected in measurements of peripheral muscles, usually in
the arm. The arm muscle area is calculated from the arm circumference and the
skinfold thickness (307). The somatic muscle pool turns over relatively slowly and
serial measurements, like those of length, do not provide acute information with
respect to recent nutritional manipulations. Serum proteins have various half-lives
and thus give differential time information. Serum prealbumin (transthyretin)
concentrations reflect recent protein intake and predict subsequent weight gain
velocity (224). The half-life of the protein is 1.9 days; therefore, a weekly
assessment of the serum prealbumin is useful. Serum albumin has a half-life of 10
to 21 days, can be used as a marker of chronic protein status, and can be
assessed monthly if needed. It is not very responsive to recent manipulations in
protein delivery. Serum transferrin concentrations, used extensively in older
children and adults because of its half-life of 10 days, is not useful in premature
infants either for assessing recent intake or for predicting weight gain (308). It
likely reflects a combination of iron and protein status and thus can be difficult to
interpret. Nitrogen balance, urinary excretion of 3-methyl histidine, and stable
isotope assessments using N-15 glycine or H-3 leucine are research tools that
assess protein status.
Rapidly changing glucose, mineral, and electrolyte status is best monitored with
serum levels. Sodium and potassium levels should be followed in infants on who
are on TPN or are receiving diuretics. Similarly, infants on TPN should have their
serum glucose concentrations monitored. In the first days after birth, sick infants
should have serum calcium, magnesium, and phosphorus levels assessed.
Chronic calcium and bone mineralization status should not be monitored solely
with serum calcium and phosphorus levels because they will usually be in the
normal to low-normal range. The serum alkaline phosphatase concentration is an
indirect measurement of bone mineralization since it is closely tied to rapid bone
turnover. An infant who is becoming osteopenic will have more rapid bone
turnover and will have a higher alkaline phosphatase level. It should be expected
that any growing premature infant will have relatively higher levels than a term
infant, but a rapidly rising weekly alkaline phosphatase level is often indicative of
active osteopenia. X-ray changes demonstrating demineralization are late findings
and indicate that the bones are at least 33% demineralized. An elevated urinary
excretion of phosphorus is also found during osteopenia of prematurity (309).
DEXA can also be used to assess bone mineralization, but is used primarily in the
research setting (310).
In general, it is unnecessary to routinely monitor trace element or vitamin status
in the healthy, growing premature infant (See Appendix J-2 for nutrients in
commercially available foods). However, higher risk infants should be monitored
periodically, depending on the micronutrient in question and the disease state of
the infant. Infants less than 1,500 g who are at high risk for BPD should have a
vitamin A level measured at birth and should be treated with supplemental
vitamin A if the level is less than 20 g/dL (161,162). Concomitant vitamin E
measurements can also be obtained. Weekly vitamin A and E levels should be

followed in infants treated for deficiency.
The recent introduction of recombinant human erythropoietin has made
monitoring of iron status an important consideration in the preterm infant. The
iron status of the premature infant can fluctuate widely; infants who are multiply
transfused have extremely high ferritin concentrations (311). Conversely, the
meager iron stores of those who receive few or no transfusions will be rapidly
consumed by erythropoiesis. Those treated with recombinant erythropoietin
experience a decrease in their ferritin levels (134). They are likely to need iron
supplementation earlier than premature infants who have been transfused.
Currently, reference values for ferritin, serum iron, and transferrin concentrations
in newborn infants have not been generated. Since iron has a narrow therapeuticto-toxic ratio, better norms for assessing iron status in preterm infants are
needed. At this time, it may be prudent to assay the serum ferritin concentration
at one month of age to assess iron stores.
REFERENCES
1. Reichman B, Chessex P, Putet G, et al. Diet, fat accretion and growth in
premature infants. N Engl J Med 1981;305:1495-1500.
2. Ziegler EE, O'Donnell AM, Nelson SE, et al. Body composition of the reference
fetus. Growth 1976;40:239-241.
3. Lemons JA, Bauer CR, Oh W, et al. Very low birth weight outcomes of the
National Institute of Child Health and Human Development neonatal research
network, January 1995 through December 1996. Pediatrics 2001;107:E1.
4. Cooke RJ, Griffin IJ, McCormick K, et al. Feeding preterm infants after
hospital discharge: effect of dietary manipulation on nutrient intake and growth.
Pediatr Res 1998;43:355-360.
5. Wahlig TM, Georgieff MK. The effects of illness on neonatal metabolism and
nutritional management. Clin Perinatol 1995;22:77-96.
6. Mathew OP. Science of bottle feeding. J Pediatr 1991;119: 511-519.
7. American Academy of Pediatrics Work Group on Breastfeeding. Breastfeeding

and the use of human milk. Pediatrics 1997;100: 1035-1039.
P.407
8. Pritchard JA. Fetal swallowing and amniotic fluid volume. Obstet Gynecol
1966;28:606-610.
9. Montgomery RK, Mulberg AW, Grand RJ. Development of the human

gastrointestinal tract: twenty years of progress. Gastro-enterology
1999;116:702-731.
10. Wood JD. Intrinsic neural control of intestinal motility. Annu Rev Physiol
1981;43:33-51.
11. Gryboski JD. The swallowing mechanisms of the neonate: I. Esophageal and
gastric motility. Pediatrics 1965;35:445-449.
12. Broussard DL. Gastrointestinal motility in the neonate. Clin Perinatol

1995;22:37-59.
13. Ittmann PI, Amarnath R, Berseth CL. Maturation of antroduodenal motor

activity in preterm and term infants. Dig Dis Sci 1992;37:14-19.
14. Siegel M, Lebenthal E, Krantz B. Effect of caloric density on gastric

emptying in premature infants. J Pediatr 1984;104:118-122.
15. Berseth CL. Gestational evolution of small intestine motility in preterm and
term infants. J Pediatr 1989;115:646-651.
16. Berseth CL. Neonatal small intestinal motility: Motor responses to feeding in
term and preterm infants. J Pediatr 1990;117: 777-782.
17. Al-Tawil Y, Klee G, Berseth CL. Extrinsic neural regulation of antroduodenal

motor activity in preterm infants. Dig Dis Sci 2002;47:26567-26563.
18. Berseth CL. Feeding methods for the preterm infant. Semin Neonatol
2001;6:417-424.
19. Schanler RJ. The use of human milk for premature infants. Pediatr Clin
North Am 2001;48:207-219.
20. Sunshine P. Digestion and absorption of proteins. In: Bloom RS, Sinclair JC,
Warshaw JB, eds. Selected aspects of perinatal gastroenterology. Evansville:
Mead Johnson, 1977:(11)17-21.
21. Euler AR, Byrne WJ, Cousins LM, et al. Increased serum gastrin
concentrations and gastric acid hyposecretion in the immediate newborn period.
Gastroenterology 1977;72:1271-1273.
22. Hadorn B, Zoppi G, Schmerling DH. Quantitative assessment of exocrine

pancreatic function in infants and children. J Pediatr 1968;73:39-50.
23. Watkins JB. Mechanisms of fat absorption and the development of

gastrointestinal function. Pediatr Clin North Am 1975;22: 721-730.
24. Balistreri WF. Anatomical and biochemical ontogeny of the gastrointestinal

tract and liver. In: Tsang RC, Nichols B, eds. Nutrition during infancy, 1st ed.
Philadelphia: Hanley & Belfus, 1988:33-55.
25. Hamosh M. Fat digestion in the newborn: role of lingual lipase and
preduodenal digestion. Pediatr Res 1979;13:615-622.
26. Hamosh M. Lingual and breast milk lipases. Adv Pediatr 1982;29:33-67.
27. Jensen RG, Jensen GL. Specialty lipids for infant nutrition: I. Milks and
formulas. J Pediatr Gastroenterol Nutr 1992;15: 232-245.
28. Hernell O, Olvecrona T. Human milk lipases. II. Bile salt stimulated lipase.
Biochim Biophys Acta 1974;369:234-244.
29. Hamosh M, Bitman J, Wood DL, et al. Lipids in milk and the first steps in
their digestion. Pediatrics 1985;75[Suppl]:146-150.
30. Watkins JB, Szczepanik P, Gould JB, et al. Bile salt metabolism in the
human premature infant. Gastroenterology 1975;69:706-713.
31. Lifschitz CH. Carbohydrate needs in preterm and term newborn infants. In:
Tsang RC, Nichols B, eds. Nutrition during infancy, 1st ed. Philadelphia: Hanley
& Belfus, 1988:122-140.
32. Koldovsky O. Development of the functions of the small intestine in

mammals and man. Basel, Switzerland: S Basel and AG Karger, 1969:168-189.
33. Gray GM. Carbohydrate absorption and malabsorption. In: Johnson LR, ed.
Physiology of the gastrointestinal tracts. New York: Raven Press 1981:10631081.
34. Auricchio S, Rubino A, Murset G. Intestinal glycosidase activities in the

human embryo, fetus and newborn. Pediatrics 1965;35: 944-949.
35. MacLean WC Jr, Fink BB. Lactose malabsorption by premature infants:

magnitude and clinical significance. J Pediatr 1980;97: 383-388.
36. Lebenthal E, Lee PC. Glycoamylase and disaccaridase activities in normal

subjects and in patients with mucosal injury of the small intestine. J Pediatr
1980;97:389-394.
37. Georgieff MK. Taking a rational approach to the use of infant formulas.
Contemp Pediatrics 2001;18:112-130.
38. Salmenpera L, Perheentupa J, Siimes MA. Exclusively breast-fed healthy

infants grow slower than reference infants. Pediatr Res 1985;19:307-312.
39. Duncan B, Schifman RB, Corrigan JJ Jr, et al. Iron and the exclusively
breast-fed infant from birth to six months. J Pediatr Gastroenterol Nutr
1985;4:421-425.
40. Wright AL, Holberg CJ, Taussig LM, et al. Relationship of infant feeding to
recurrent wheezing at age 6 years. Arch Pediatr Adolesc Med 1995;149:758763.
41. Widdowson EM, Spray CM. Chemical development in utero. Arch Dis Child
1951;26:205-214.
42. Babson SG, Benda GI. Growth graphs for the clinical assessment of infants
of varying gestational age. J Pediatr 1976;89: 814-820.
43. Lubchenco L, Hansman C, Boyd E. Intrauterine growth in length and head

circumference as estimated from live births at gestational ages from 26 to 42
weeks of gestation. Pediatrics 1966;37: 403-408.
44. Alexander GR, Himes JH, Kaufman RB, et al. A United States national
reference for fetal growth. Obstet Gynecol 1996;87: 163-168.
45. Butte NF. Meeting energy needs. In: Tsang RC, Zlotkin SH, Nichols B, et al,
eds. Nutrition during infancy, 2nd ed. Cincinnati: Digipub 1997:57-82.
46. Hamosh M. Lipid metabolism in premature infants. Biol Neonat 1987;50

[Suppl 1]:50-64.
47. Ross Pediatrics. Composition of feedings for infants and young children.
Ross Ready Reference. Columbus, Ohio: Ross Products Division, Abbott
Laboratories, 1996.
48. Motil KJ. Meeting protein needs. In: Tsang RC, Zlotkin SH, Nichols B, et al,
eds. Nutrition during infancy, 2nd ed. Cincinnati: Digipub 1997:83-104.
49. Stocker FP, Wilkoff W, Mietinen OS, et al. Oxygen consumption in infants
with heart disease. J Pediatr 1972;80:43-51.
50. Weinstein MR, Oh W. Oxygen consumption in infants with

bronchopulmonary dysplasia. J Pediatr 1981;99:958-961.
51. Wahlig TM, Gatto CW, Boros SJ. Metabolic response of preterm infants to
variable degrees of respiratory illness. J Pediatr 1994;124:283-288.
52. Mrozek JD, Georgieff MK, Blazar BR, et al. Neonatal sepsis: effect on protein
and energy metabolism. Pediatr Res 1997;41:237A.
53. Heinig MJ, Nommsen LA, Peerson JM, et al. Energy and protein intakes of
breast-fed and formula-fed infants during the first year of life and their
association with growth velocity. The DARLING study. Am J Clin Nutr

1993;58:152-161.
54. Butte NF, Wong WW, Ferlic L, et al. Energy expenditure and deposition of
breast-fed and formula-fed infants during early infancy. Pediatr Res
1990;28:631-640.
55. Whyte RK, Campbell D, Stanhope R, et al. Energy balance in low birth
weight infants fed formula of high or low medium chain triglyceride content. J
Pediatr 1986;108:964-971.
56. Sauer PJJ, Dane HF, Visser HKA. Longitudinal studies on metabolic rate,
heat loss, and energy cost of growth in low birth weight infants. Pediatr Res
1984;18:254-259.
57. Klein CJ. Nutrient requirements for preterm infant formulas. J Nutr
2002;132:1395S-1577S.
58. Hay WW Jr. Fetal and neonatal glucose homeostasis and their relation to
small for gestation age infants. Semin Perinatol 1984;8:101-116.
59. Kliegman RM, Morton S. Sequential intrahepatic metabolic effects of enteric

galactose alimentation in newborn rats. Pediatr Res 1988;24:302-307.
60. DiGiacomo JE. Carbohydrates: metabolism and disorders. In: Hay WW Jr,
ed. Neonatal nutrition and metabolism. St. Louis: Mosby 1991:93-109.
61. Lilien LD, Pildes RS, Srinivasan G, et al. Treatment of neonatal

hypoglycemia with mini-bolus and intravenous glucose infusions. J Pediatr
1980;97:295-298.
62. Collins JW Jr, Hoppe M, Brow K, et al. A controlled trial of insulin infusion
and parenteral nutrition in extremely low birth weight infants with glucose
intolerance. J Pediatr 1991;118: 921-927.
63. Binder ND, Rasschko PK, Benda GI, et al. Insulin infusion with parenteral
nutrition in extremely low birth weight infants with hyperglycemia. J Pediatr
1989;114:273-280.
64. Burr GO, Burr MM. A new deficiency disease produced by rigid exclusion of
fat from the diet. J Biol Chem 1929;82:345-367.
P.408
65. American Academy of Pediatrics, Committee on Nutrition. Nutritional needs
of low birth weight infants. Pediatrics 1985;75: 976-986.
66. Lucas A. Long-chain polyunsaturated fatty acids, infant feeding and

cognitive development. In: Dobbing J, ed. Developing brain and behaviour: the
role of lipids in infant formula. San Diego: Academic Press 1997:3-36.
67. Innis SM. Essential fatty acids in growth and development. Prog Lipid Res
1991;30:39-103.
68. Innis SM. Polyunsaturated fatty acid nutrition in infants born at term. In:
Dobbing J, ed. Developing brain and behaviour: the role of lipids in infant
formula. San Diego: Academic Press 1997: 103-140.
69. Carson SE, Werkman SH, Tolley EA. Effect of long-chain n-3 fatty acid
supplementation on visual acuity and growth of preterm infants with and
without bronchopulmonary dysplasia. Am J Clin Nutr 1996;63:687-697.
70. Carlson SE, Werkman SH, Peeples JM, et al. Arachidonic acid status
correlates with first year growth in preterm infants. Proc Natl Acad Sci U S A
1993;90:1073-1077.
71. O'Connor DL, Hall R, Adamkin D, et al. Growth and development in preterm
infants fed long-chain polyunsaturated fatty acids: a prospective, randomized
controlled trial. Pediatrics 2001;108:359-371.
72. Auestad N, Halter R, Hall R, et al. Growth and development in term infants
fed long-chain polyunsaturated fatty acids: a double-masked randomized,
parallel, prospective multi-variate study. Pediatrics 2001;108:372-381.
73. European Society of Paediatric Gastroenterology and Nutrition (ESPGAN).

Nutrition and feeding of preterm infants. Acta Paediatr Scand 1987;336
[Suppl]:1-14.
74. Uauy RD, Birch DG, Birch EE, et al. Effect of dietary omega-3 fatty acids on
retinal function of very-low-birth-weight neonates. Pediatr Res 1990;28:485492.
75. Birch EE, Birch DG, Hoffman DR, et al. Dietary essential fatty acid supply
and visual acuity development. Invest Ophthalmol Vis Sci 1992;33:3242-3253.
76. Hoffman DR, Birch EE, Castaneda YS, et al. Visual function in breast-fed
term infants weaned to formula with or without long-chain polyunsaturates at 4
to 6 months: a randomized clinical trial. J Pediatr 2003;142:669-677.
77. Jensen CL, Heird WC. Lipids with an emphasis on long-chain

polyunsaturated fatty acids. Clin Perinatol 2002;29:261-281.
78. Lucas A, Stafford M, Morley R, et al. Efficacy and safety of long-chain

polyunsaturated fatty acid supplementation of infant-formula milk: a
randomised trial. Lancet 1999;354:1948-1954.
79. Wroble M, Mash C, Williams L, et al. Should long chain polyunsaturated

fatty acids be added to infant formula to promote development? J Appl Dev
Psychol 2002;23:99-112.
80. MacLean WC Jr, Benson JD. Theory into practice: the incorporation of new
knowledge into infant formula. Semin Perinatol 1989;13:104-111.
81. Penn D, Schmidt-Sommerfeld E, Wolf H. Carnitine deficiency in premature

infants receiving total parenteral nutrition. Early Hum Dev 1980;4:23-34.
82. Helms RA, Whitington PF, Mauer EC, et al. Enhanced lipid utilization in
infants receiving oral L-carnitine during long-term parenteral nutrition. J Pediatr
1986;109:984-988.
83. Lorenz JM, Kleinman LI. Otogeny of the Kidney. In: Tsang RC, Nichols B,
eds. Nutrition during infancy. Philadelphia: Hanley & Belfus 1988:58-80.
84. Ziegler TR, Gatzen C, Wilmore DW. Strategies for attenuating protein-
catabolic responses in the critically ill. Annu Rev Med 1994;45:459-480.
85. Keshen TH, Jaksic T, Jahoor F. Measurement of the protein metabolic

response to surgical stress in extremely low birthweight neonates. Pediatric Res
1997;41:234A.
86. Thureen PJ, Melara D, Fennessey PV, et al. Effectof low versus high
intravenous amino acid intake on very low birth weight infants in the early
neonatal period. Pediatr Res 2003;53:24-32.
87. Carver JD, Pimentel B, Cox WI, et al. Dietary nucleotide effects upon
immune function in infants. Pediatrics 1991;88:359-363.
88. Goldman AS, Garza C, Nichold B, et al. Effect of prematurity on the

immunologic system in human milk. J Pediatr 1982;101: 901-905.
89. Mestecky J, Blair C, Ogry PL et al. Immunology of milk and the neonate.
Adv Exp Med Biol 1990;310:1-177.
90. Uauy R, Stringel G, Thomas R, et al. Effect of dietary nucleosides on growth

and maturation of the developing gut in the rat. J Pediatr Gastroenterol Nutr
1990;10:497-503.
91. Graham GG, Placko RP, Morales E, et al. Dietary protein quality in infants
and children. Am J Dis Child 1970;120:419-423.
92. Davidson M, Levine SZ, Bauer CH, et al. Feeding studies in low-birth-weight
infants. I. Relationships of dietary protein, fat, and electrolytes to rates of
weight gain, clinical courses, and serum chemical concentrations. J Pediatr
1967;70:695-713.
93. Kagan BM, Stanincova V, Felix NS, et al. Body composition of premature
infants: relation to nutrition. Am J Clin Nutr 1972;25: 1153-1164.
94. Kashyap S, Schulze KF, Forsyth MS, et al. Growth, nutrient retention, and
metabolic response in low birth weight infants fed varying intakes of protein and
energy. J Pediatr 1988;113: 713-721.
95. Schulze KF, Stefanski M, Masterson J, et al. Energy expenditure, energy

balance and composition of weight gain in low birth weight infants fed diets of
different protein and energy content. J Pediatr 1987;110:753-759.
96. Heird WC, Kashyap S, Gomez MR. Protein intake and energy requirements
of the infant. Semin Perinatol 1991;15:438-448.
97. Anderson TL, Muttart C, Bieber MA, et al. A controlled trial of glucose vs.
glucose and amino acids in premature infants. J Pediatr 1979;94:947-951.
98. Denne SC. Protein and energy requirements in preterm infants. Semin
Neonatol 2001;6:377-382.
99. Ziegler EE, Thureen PJ, Carlson SJ. Aggressive nutrition of the very low
birthweight infant. Clin Perinatol 2002;39:225-244.
100. Uauy R, Greene HL, Heird WC. Conditional nutrients. In: Tsang RC, Lucas
A, Uauy R, et al, eds. Nutritional needs of the preterm infant. Baltimore:
Williams & Wilkins, 1993:267-280.
101. van Goudoeveer JB, Wattimena JDL, Carnielli VP, et al. Effect of
dexamethasone on protein metabolism in infants with bronchopulmonary
dysplasia. J Pediatr 1994;124:112-118.
102. Wu PY, Edwards NB, Storm MC. Plasma amino acid pattern in normal term
breast-fed infants. J Pediatr 1986;109:347-349.
103. Clark SC, Karn CA, Ahlrichs JA, et al. Acute changes in leucine and
phenylalanine kinetics produced by parenteral nutrition in premature infants.
Pediatr Res 1997;41:568-574.
104. Gross SJ, David RJ, Bauman L, et al. Nutritional composition of milk
produced by mothers delivering preterm. J Pediatr 1980;96: 641-644.
105. Schanler RJ, Oh W. Composition of breast milk obtained from mothers of

premature infants as compared with breast milk obtained from donors. J Pediatr
1980;96:679-681.
106. Feeley RM, Eitenmiller RR, Jones JB, et al. Calcium, phosphorus and
magnesium contents of human milk during early lactation. J Pediatr
Gastroenterol Nutr 1983;2:262-267.
107. Kashyap S, Towers HM, Sahni R, et al. Effects of quality of energy on

substrate oxidation in enterally fed, low-birth-weight infants. Am J Clin Nutr
2001;74:374-380.
108. American Academy of Pediatrics. Nutritional needs of preterm infants. In:

Kleinman R, ed. Pediatric nutrition handbook, 4th ed. Elk Grove Village, IL: AAP,
1998:55-79.
109. Vitamin and mineral requirements in preterm infants. Tsang RC, ed. New
York: Marcel Dekker, 1985:1-212.
110. Holliday MA. Requirements for sodium chloride and potassium and their
interrelation with water requirement. In: Tsang RC, Nichols B, eds. Nutrition
during infancy. Philadelphia: Hanley & Belfus 1988:160-191.
111. Koo WWK, Tsang RC. Calcium, magnesium, phosphorus and vitamin D. In:
Tsang RC, Lucas A, Uauy R, et al, eds. Nutritional needs of the preterm infant.
Baltimore: Williams & Wilkins 1993:135-156.
112. Schanler RJ, Garza C. Improved mineral balance in very low birth weight
infants fed fortified human milk. J Pediatr 1987; 112:452-456.
113. Oski FA. The hematologic aspects of the maternal-fetal relationship. In:
Oski FA, Naiman JL, eds. Hematologic problems in the newborn, 3rd ed.
Philadelphia: WB Saunders, 1982:32-61.
114. Chockalingam UM, Murphy E, Ophoven JC, et al. Cord transferrin and
ferritin levels in newborn infants at risk for prenatal uteroplacental insufficiency
and chronic hypoxia. J Pediatr 1987;111:283-286.
115. Georgieff MK, Landon MB, Mills MM, et al. Abnormal iron distribution in
infants of diabetic mothers: spectrum and maternal antecedents. J Pediatr
1990;117:455-461.
P.409
116. Petry CD, Eaton MA, Wobken JD, et al. Placental transferrin receptor
localization and binding characteristics in diabetic pregnancies characterized by
increased fetal iron demand. Am J Physiol 1994;E507-E514.
117. Georgieff MK, Petry CD, Mills MM, et al. Increased N-glycosylation and
reduced transferrin binding capacity of transferrin receptor isolated from
placentas of diabetic mothers. Placenta 1997;18:563-568.
118. American Academy of Pediatrics. Iron deficiency. In: Kleinman R, ed.

Pediatric nutrition handbook, 4th ed. Elk Grove Village, IL: AAP, 1998:233-246.
119. Committee on Nutrition, American Academy of Pediatrics. Iron fortification

of infant formulas. Pediatrics 1999;104:119-123.
120. Siimes MA, Vuori E, Kuitunen P. Breast milk iron: a declining concentration
during the course of lactation. Acta Paediatr Scand 1979;68:29-31.
121. Lonnerdal B. Iron in human milk and cow's milkeffects of binding ligands
on bioavailability. In: Lonnerdal B, ed. Iron metabolism in infants. Boca Raton,
FL: CRC Press, 1990:87-103.
122. Pisacane A, De Vizia B, Valiante A, et al. Iron status in breast-fed infants. J

Pediatr 1995;127:429-431.
123. Pizarro F, Yip R, Dallman PR, et al. Iron status with different infant feeding
regimens: relevance to screening and prevention of iron deficiency. J Pediatr
1991;118:687-692.
124. Innis SM, Nelson CM, Wadsworth LD, et al. Incidence of iron-deficiency
anaemia and depleted iron stores among nine-month-old infants in Vancouver,
Canada. Can J Pub Health 1997;88:80-84.
125. Lonnerdal B, Hernell O. Iron, zinc, copper and selenium status of breastfed infants and infants fed trace element fortified milk-based infant formula.
Acta Paediatr 1994;83:367-373.
126. Foman SJ, Ziegler EE, Serfass RE, et al. Erythrocyte incorporation of iron is
similar in infants fed formulas fortified with 12 mg/L or 8 mg/L of iron. J Nutr
1997;127:83-88.
127. Oski FA. Iron-fortified formulas and gastrointestinal symptoms in infants: a

controlled study. Pediatrics 1980;66:168-170.
128. Nelson SE, Ziegler EE, Copeland AM, et al. Lack of adverse reactions to
iron-fortified formula. Pediatrics 1988;81:360-364.
129. Hall RT, Wheeler RE, Benson J, et al. Feeding iron-fortified premature
formula during initial hospitalization to infants less than 1800 grams birth
weight. Pediatrics 1993;92:409-414.
130. Winzerling JJ, Kling PJ. Iron deficient erythropoiesis in premature infants
measured by blood zinc protoporphyrin/heme. J Pediatr 2001;139:134-136.
131. Cooke RW, Drury JA, Yoxall CW, et al. Blood transfusion and chronic lung
disease in preterm infants. Eur J Pediatr 1997;156: 47-50.
132. Inder TE, Clemett RS, Austin NC, et al. High iron status in very low birth
weight infants is associated with an increased risk of retinopathy of prematurity.
J Pediatr 1997;131:541-544.
133. Widness JA, SewardVJ, Kromer IJ, et al. Changing patterns of red blood
cell transfusion in very low birth weight infants. J Pediatr 1996;129:680-687.
134. Shannon K, Keith J, Mentzer W, et al. Recombinant human erythropoietin

stimulates erythropoiesis and reduces erythrocyte transfusions in very low birth
weight preterm infants. Pediatrics 1995;95:1-8.
135. Lozoff B, Jimenez E, Hagen J, et al. Poorer behavioral and developmental

outcome more than 10 years after treatment for iron deficiency in infancy.
Pediatrics 2000;105:E51.
136. Roncagliolo M, Garrido M, Walter T, et al. Evidence of altered central

nervous system development in infants with iron deficiency anemia at 6 mo:
delayed maturation of auditory brainstem responses. Am J Clin Nutr
1998;68:683-690.
137. Algarin C, Peirano P, Garrido M, et al. Iron deficiency anemia in infancy:

long-lasting effects on auditory and visual system functioning. Pediatr Res
2003;53:217-223.
138. Berger HM, Mumby S, Gutteridge JM. Ferrous ions detected in ironoverloaded cord blood plasma from preterm and term babies: implications for
oxidative stress. Free Radic Res 1995;22: 555-559.
139. Chockalingam U, Murphy E, Ophoven JC, et al. The influence of gestational

age, size for dates and prenatal steroids on cord transferrin in newborn infants.
J Pediatr Gastroenterol Nutr 1987;6: 276-280.
140. Jansson LT. Iron, oxygen stress and the preterm infant. In: Lonnerdal B,
ed. Iron metabolism in infants. Boca Raton, FL: CRC Press, 1990:73-85.
141. Buonocore G, Perrone S, Longini M, et al. Non protein bound iron as early
predictive marker of neonatal brain damage. Brain 2003;126:1224-1230.
142. Pollak A, Hayde M, Hayn M, et al. Effect of intravenous iron

supplementation on erythropoiesis in erythropoietin treated premature infants.
Pediatrics 2001;107:78-85.
143. Peters C, Georgieff MK, DeAlarcon P, et al. The effect of chronic

erythropoietin administration on iron in newborn lambs. Biol Neonate
1996;70:218-228.
144. Zlotkin SH, Atkinson S, Lockitch G. Trace elements in nutrition for

premature infants. Clin Perinatol 1995;22:223-240.
145. Reifen RM, Zlotkin SH. Microminerals. In: Tsang RC, Lucas A, Uauy R, et
al, eds. Nutritional needs of the preterm infant. Baltimore: Williams & Wilkins,
1993:195-207.
146. Bayliss PA, Buchanan BE, Hancock RGV, et al. Tissue selenium accretion in
premature and full-term human infants and children. Biol Trace Elem Res
1985;7:755-759.
147. Sinkin RA, Phelps DL. New strategies for the prevention of
bronchopulmonary dysplasia. Clin Perinatol 1987;14:599-620.
148. Lipsky CL, Spear ML. Recent advances in parenteral nutrition. Clin
Perinatol 1995;22:141-155.
149. Greene H, Hambridge K, Schanler R, et al. Guidelines for the use of

vitamins, trace elements, calcium, magnesium, and phosphorus in infants and
children receiving total parenteral nutrition: report of the Subcommittee on
Pediatric Parenteral Nutrient Requirements from the Committee on Clinical
Practice Issues of the American Society for Clinical Nutrition. Am J Clin Nutr
1988;48:1324-1342.
150. American Academy of Pediatrics. Vitamins. In: Kleinman R, ed. Pediatric

nutrition handbook, 4th ed. Elk Grove Village, IL: AAP, 1998:267-281.
151. Pereira GR. Nutritional care of the extremely premature infant. Clin
Perinatol 1995;22:61-75.
152. Feldman KW, Marcuse EK, Springer DA. Nutritional rickets. Am Fam
Physician 1990;42:1311-1318.
153. Kreiter SR, Schwartz RP, Kirkman NH Jr, et al. Nutritional rickets in AfricanAmerican breast-fed infants. J Pediatr 2000;137: 153-157.
154. Riedel BD, Greene HL. Vitamins. In: Hay WW Jr, ed. Neonatal nutrition and
metabolism. St. Louis: Mosby, 1991:143-170.
155. Hillman L, Hoff N, Salmons SJ, et al. Mineral homeostasis in very

premature infants: serial evaluation of serum 25 hydroxy vitamin D, serum
minerals and bone mineralization. J Pediatr 1985;106:970-980.
156. Wolbach SB, Howe PR. Tissue changes following deprivation of fat-soluble
vitamin A. J Exp Med 1925;42:753-777.
157. Shenai JP, Stahlman MT, Chytil F. Vitamin A delivery from parenteral
alimentation solutions. J Pediatr 1981;99:661-663.
158. Shenai JP, Chytil F, Stahlman MT. Vitamin A status of neonates with
bronchopulmonary dysplasia. Pediatr Res 1985;19:185-188.
159. Georgieff MK, Chockalingam UM, Sasanow SR, et al. The effect of
antenatal betamethasone on cord blood concentrations of retinol-binding
protein, transthyretin, transferrin, retinol and vitamin E. J Pediatr Gastroenterol
Nutr 1988;7:713-717.
160. Georgieff MK, Mammel MC, Mills MM, et al. Effect of postnatal steroid
administration on serum vitamin A concentrations in newborn infants with
respiratory compromise. J Pediatr 1989; 114:301-304.
161. Shenai JP, Kennedy KA, Chytil F, et al. Clinical trial of vitamin A
supplementation in infants susceptible to bronchopulmonary dysplasia. J Pediatr
1987;111:269-277.
162. Pearson E, Bose C, Snidow T, et al. Trial of vitamin A supplementation in

very low birth weight infants at risk for bronchopulmonary dysplasia. J Pediatr
1992;121:420-427.
163. Tyson JE, Wright LL, Oh W, et al. Vitamin A supplementation for extremelylow-birth-weight infants. National Institute of Child Health and Human
Development Neonatal Research Network. N Engl J Med 1999;340:1962-1968.
164. Oski FA, Barnes LA. Vitamin E deficiency: a previously unrecognized cause
of hemolytic anemia in the premature. J Pediatr 1967;70:211-220.
165. Johnson L, Schaffer D, Boggs TR. The premature infant, vitamin E

deficiency and retrolental fibroplasia. Am J Clin Nutr 1974; 27:1158-1173.
166. Hittner HM, Godio LB, Rudolph AJ, et al. Retrolental fibroplasia: efficacy of
vitamin E in a double-blind clinical study of preterm infants. N Engl J Med
1981;305:1365-1371.
P.410
167. Hittner HM, Godio LB, Speer ME, et al. Retrolental fibroplasia: further
clinical evidence and ultrastructural support for efficacy of vitamin E in the

preterm infant. Pediatrics 1983;71: 423-432.
168. Phelps DL, Rosenbaum AL, Isenberg SJ, et al. Tocopherol efficacy and
safety for preventing retinopathy of prematurity: a randomized, controlled,
double-masked trial. Pediatrics 1987;79: 489-500.
169. Phelps DL. Retinopathy of prematurity. Curr Probl Pediatr 1992; 22:349371.
170. Johnson L, Bowen FW, Abbasi S, et al. Relationship of prolonged

pharmacologic serum levels of vitamin E to incidence of sepsis and necrotizing
enterocolitis in infants with birth weights of 1500 grams or less. Pediatrics
1985;75:619-638.
171. Heird WC, Gomez MR. Parenteral Nutrition. In: Tsang RC, Lucas A, Uauy R,
et al, eds. Nutritional needs of the preterm infant. Baltimore: Williams & Wilkins
1993:225-242.
172. Steinhorn DM, Green TP. Severity of illness correlates with alterations in
energy metabolism in the pediatric intensive care unit. Crit Care Med
1991;19:1503-1509.
173. Cerra FB, Siegel JH, Coleman B, et al. Septic autocannibalism: a failure of
exogenous nutritional support. Ann Surg 1980;192: 570-580.
174. Fleck A. Acute phase response: implications for nutrition and recovery.
Nutrition 1988;4:109-116.
175. Harris MC, Costarino AT, Sullivan JS, et al. Cytokine elevations in critically
ill infants with sepsis and necrotizing enterocolitis. J Pediatr 1994;124:105-111.
176. Maldonato J, Gil A, Faus MJ, et al. Differences in the serum amino acid
pattern of injured and infected children promoted by two parenteral nutrition
solutions. JPEN 1989;13:41-46.
177. Canadian Paediatric Society and Nutrition Committee. Nutrient needs and
feeding of premature infants. CMAJ 1995;152: 1765-1785.
178. Priego T, Ibanez de Caceres I, Martin AI, et al. Glucocorticoids are not
necessary for the inhibitory effect of endotoxic shock on serum IGF-I and
hepatic IGF-I mRNA. J Endocrinol 2002;172: 449-456.
179. Carver JD, Wu PY, Hall RT, et al. Growth of preterm infants fed nutrientenriched or term formula after hospital discharge. Pediatrics 2001;107:683-689.
180. deRegnier R-AO, Guilbert TW, Mills MM, et al. Growth failure and altered
body composition are established by one month of age in infants with
bronchopulmonary dysplasia. J Nutr 1996;126:168-175.
181. Frank L, Sosenko IRS. Undernutrition as a major contributing factor in the

pathogenesis of bronchopulmonary dysplasia. Am Rev Respir Dis 1988;138:725729.
182. Georgieff MK, Mills MM, Lindeke L, et al. Changes in nutritional

management and outcome of very-low-birth-weight infants. Am J Dis Child
1989;143:82-85.
183. Berseth CL. Minimal enteral feedings. Clin Perinatol 1995;22: 195-205.
184. Roberts JR. Cutaneous and subcutaneous complications of calcium

infusions. JACEP 1977;6:16-20.
185. Hruszkewycz V, Holtrop PC, Batton DG, et al. Complications associated

with central venous catheters inserted in critically ill neonates. Infect Control
Hosp Epidemiol 1991;12:544-548.
186. Polin RA, St. Geme JW 3rd. Neonatal sepsis. Adv Pediatr Infect Dis
1992;7:25-61.
187. Rowen JL, Atkins JT, Levy ML, et al. Invasive fungal dermatitis in the less
than or equal to 1000 gram neonate. Pediatrics 1995;95:682-687.
188. Duffy LF, Kerzner B, Gebus V, et al. Treatment of central venous catheter
occlusions with hydrochloric acid. J Pediatr 1989;114: 1002-1004.
189. Kashyap S. Nutritional management of the extremely-low-birth-weight

infant. In: Cowett RM, Hay W, eds. The micropremie: the next frontier. Report
of the 99th Ross Conference on Pediatric Research, 1990:115-119.
190. Zlotkin SH, Bryan MH, Anderson GH. Intravenous nitrogen and energy
intakes required to duplicate in utero nitrogen accretion in prematurely born
human infants. J Pediatr 1981;99:115-120.
191. Helms RA, Christensen ML, Mauer EC, et al. Comparison of a pediatric
versus standard amino acid formulation in preterm neonates requiring
parenteral nutrition. J Pediatr 1987;110: 466-470.
192. Helms RA, Johnson MR, Christenson ML, et al. Evaluation of two pediatric
amino acid formulations (abst). JPEN 1988;12:4.
193. Heird WC, Dell RB, Helms RA, et al. Amino acid mixture designed to
maintain normal plasma amino acid patterns in infants and children requiring
parenteral nutrition. Pediatrics 1987;80:401-408.
194. Hayes KC, Carey RE, Schmidt SY. Retinal degeneration associated with
taurine deficiency in the cat. Science 1975;188:949-951.
195. Guertin F, Roy CC, Lepage G, et al. Effect of taurine on parenteral nutritionassociated cholestasis. JPEN 1991;15:247-251.
196. Tyson JE, Lasky R, Flood D, et al. Randomized trial of taurine

supplementation for infants <1,300 gram birth weight: effect on auditory
brainstem-evoked responses. Pediatrics 1989;83:406-415.
197. White HB, Turner AC, Miller RC. Blood lipid alterations in infants receiving
intravenous fat-free alimentation. J Pediatr 1973;83:305-313.
198. Costarino AT, Baumgart S. Modern fluid and electrolyte management of

the critically ill premature infant. Pediatr Clin North Am 1986;33:153-178.
199. Dunham B, Marcuard S, Khazanie PG, et al. The solubility of calcium and
phosphorus in neonatal parenteral nutrition solutions. JPEN 1991;15:608-611.
200. Pelegano JF, Rowe JC, Carey DE, et al. Effect of calcium/phosphorus ratio
in mineral retention in parenterally fed premature infants. JPEN 1991;12:351355.
201. Hoehn GJ, Carey DE, Raye JR, et al. Alternate-day infusion of calcium and
phosphate in very low birth weight infants: wasting of the infused mineral. JPEN
1987;6:752-757.
202. Kimura S, Nose O, Seino Y, et al. Effects of alternate and simultaneous

administration of calcium and phosphorus on calcium metabolism in children
receiving total parenteral nutrition. JPEN 1986;10:513-516.
203. Fitzgerald KA, MacKay MW. Calcium and phosphate solubility in neonatal
parenteral nutrient solutions containing Trophamine. Am J Hosp Pharm
1986;43:88-93.
204. Schmidt GL, Baumgartner TG, Fischlschweiger W, et al. Cost containment

using cysteine HCl acidification to increase calcium/ phosphate solubility in
hyperalimentation solutions. JPEN 1986;10:203-207.
205. Thorp JW, Boeckx RL, Robbins S, et al. A prospective study of infant zinc
nutrition during intensive care. Am J Clin Nutr 1981;34:1056-1060.
206. Heller RM, Kirchner SG, O'Neill JA, et al. Skeletal changes of copper
deficiency in infants receiving prolonged total parenteral nutrition. J Pediatr
1978;92:947-949.
207. Lane HW, Barroso AO, Englert D, et al. Selenium status of seven chronic
intravenous hyperalimentation patients. JPEN 1982;6: 426-431.
208. Stahl GE, Spear ML, Hamosh M. Intravenous administration of lipid

emulsions to premature infants. Clin Perinatol 1986;13: 133-162.
209. Hageman JR, McCullough K, Gora P, et al. Intralipid alterations in

pulmonary prostaglandin metabolism and gas exchange. Crit Care Med
1983;11:794-798.
210. McKeen CR, Brigham KL, Bowers RE, et al. Pulmonary vascular effects of
fat emulsion infusion in unanesthetized sheep. Prevention by indomethacin. J
Clin Invest 1978;61:1291-1297.
211. Hammerman C, Aramburo MJ. Decreased lipid intake reduces morbidity in

sick premature neonates. J Pediatr 1988;113: 1083-1088.
212. Gilbertson N, Kovar IZ, Cox, DJ, et al. Introduction of intravenous lipid
administration on the first day of life in the very low birth weight neonate. J
Pediatr 1991;119:615-623.
213. Eggert LD, Rusho WJ, MacKay MW, et al. Calcium and phosphorus
compatibility in parenteral nutrition solutions for neonates. Am J Hosp Pharm
1982;39:49-53.
214. Adamkin DH, Radmacher PG, Klingbeil RL. Use of intravenous lipid and
hyperbilirubinemia in the first week. JPEN 1992;14: 135-139.
215. Aschner JL, Punsalang A, Maniscalco WM, et al. Percutaneous central

venous catheter colonization with Malassezia furfur: incidence and clinical
significance. Pediatrics 1987;80:535-539.
216. Merritt RJ. Cholestasis associated with total parenteral nutrition. JPEN
1986;5:9-22.
217. Koo WWK, Kaplan LA, Horn J, et al. Aluminum in parenteral solutions
sources and possible alternatives. JPEN 1986;10: 591-595.
P.411
218. Koo WWK, Kaplan LA, Bendon R, et al. Response to aluminum in
parenteral nutrition during infancy. J Pediatr 1986;109: 877-883.
219. Alfrey AC. Aluminum. Adv Clin Chem 1983;23:69-91.
220. Sedman AB, Klein GL, Merritt RJ, et al. Evidence of aluminum loading in
infants receiving intravenous therapy. N Engl J Med 1985;312:1337-1343.
221. ASCN/A.SPEN. Working Group on Standards for Aluminum Content of

Parenteral Nutrition Solutions. Parenteral drug products containing aluminium
as an ingredient or a contaminant: response to food and drug administration
notice of intent and request for information. JPEN 1991;15:194-198.
222. Sann L, Durand M, Picard J, et al. Arm fat and muscle areas in infancy.
Arch Dis Child 1988;63:256-260.
223. Georgieff MK, Sasanow SR, Pereira GR. Serum transthyretin levels and
protein intake as predictors of weight gain velocity in premature infants. J
Pediatr Gastroenterol Nutr 1987;6:775-779.
224. Georgieff MK, Sasanow SR, Mammel MC, et al. Cord prealbumin values in
newborn infants: effect of prenatal steroids, pulmonary maturity and size for
dates. J Pediatr 1986;108:972-976.
225. Vallee BL, Galdes A. The metallobiochemistry of zinc enzymes. Adv

Enzymol Relat Areas Mol Biol 1984;56:283-430.
226. Lau C, Smith EO, Schanler RJ. Coordination of suck-swallow and swallow
respiration in preterm infants. Acta Paediatr 2003;92: 721-727.
227. Meier P. Bottle and breastfeeding effects on transcutaneous oxygen

pressure and temperature in preterm infants. Nurs Res 1988;37:36-41.
228. Lucas A, Morley R, Cole TJ, et al. A randomised multicentre study of

human milk versus formula and later development in preterm infants. Arch Dis
Child Fetal Neonatal Ed 1994;70:F141-F146.
229. Bernbaum JC, Pereira GR, Watkins JB, et al. Nonnutritive sucking during
gavage feeding enhances growth and maturation in premature infants.
230. Field T, Ignatoff E, Stringer S, et al. Nonnutritive sucking during tube

feedings: effects on preterm neonates in an intensive care unit. Pediatrics
1982;70:381-384.
231. Widstrom AM, Marchini G, Matthieson AS, et al. Nonnutritive sucking in
tube-fed preterm infants: effects on gastric motility and gastric contents of

Somatostatin. J Pediatr Gastroenterol Nutr 1988;7:517-523.
232. Toce SS, Keenan WJ. Enteral feeding in very-low-birth-weight infants. Am

J Dis Child 1987;141:436-444.
233. Roy RN, Pillnitz RP, Hamilton JR, et al. Impaired assimilation of nasojejunal
feeds in healthy low-birth-weight newborn infants. J Pediatr 1977;90:431-434.
234. Wells DH, Zachman RD. Nasojejunal feeds in low-birth-weight infants. J

Pediatr 1975;87:276-279.
235. Challacombe D. Bacterial microflora in infants receiving nasojejunal tube

feeding [Letter]. J Pediatr 1974;85:113.
236. Kennedy KA, Tyson JE, Chamnanvankij S. Rapid versus slow rate of
advancement of feedings for promoting growth and preventing necrotizing
enterocolitis in parenterally fed low-birth-weight infants. Cochrane Database
Syst Rev 2000;2:CD001241.
237. Lucas A, Bloom SR, Aynsley-Green A. Gut hormones and minimal enteral
feeding. Acta Paediatr Scand 1986;75:719-723.
238. Slagle TA, Gross SJ. Effect of early low-volume enteral substrate on
subsequent feeding tolerance in the very low birth weight infants. J Pediatr
1988;113:526-531.
239. Dunn L, Hulman S, Weiner J, et al. Beneficial effects of early hypocaloric

enteral feeding on neonatal gastrointestinal function: preliminary report of a
randomized trial. J Pediatr 1988;112:622-629.
240. Meetze W, Valentine C, Sacks J, et al. Effects of gastrointestinal (GI)

priming prior to full enteral nutrition in very low birth weight (VLBW) infants.
Pediatr Res 1990;27:287A(abst).
241. Neu J, Valentine C, Mietze W. Scientifically-based strategies for nutrition of

the high-risk low birth weight infant. Eur J Pediatr 1990;150:2-13.
242. Hughes CA, Dowling RH. Speed of onset of adaptive mucosal hypoplasia
and hypofunction in the intestine of parenterally fed rats. Clin Sci 1980;59:317327.
243. Ryan AS, Wenjun Z, Acosta A. Breastfeeding continues to increase into the
new millennium. Pediatrics 2002;110:1103-1109.
244. Freed GL, Clark SJ, Sorenson J, et al. National assessment of physicians'
breast-feeding knowledge, attitudes, training, and experience. JAMA
1995;273:472-476.
245. World Health Organization. Protecting, promoting and supporting breastceeding: the special role of maternity services. Geneva, Switzerland: WHO,
1989.
246. Spisak S, Gross SS. Second Followup Report: the Surgeon General's
Workshop on Breastfeeding and Human Lactation. Washington, DC: National
Center for Education in Maternal and Child Health, 1991.
247. Braveman P, Egerter S, Pearl M, et al. Problems associated with early

discharge of newborn infants. Pediatrics 1995;96: 716-726.
248. Williams LR, Cooper MK. Nurse-managed postpartum home care. J Obstet
Gynecol Neonatal Nurs 1993;22:25-31.
249. Gielen AC, Faden RR, O'Campo P, et al. Maternal employment during the
early postpartum period: effects on initiation and continuation of breastfeeding. Pediatrics 1991;87:298-305.
250. Frederick IB, Auerback KG. Maternal-infant separation and breast-feeding:

the return to work or school. J Reprod Med 1985;30:523-526.
251. Spisak S, Gross SS. Second Followup Report: the Surgeon General's
Workshop on Breastfeeding and Human Lactation. Washington, DC: National
Center for Education in Maternal and Child Health, 1991.
252. Wilson MH. Feeding the healthy child. In: Oski FA, DeAngelis CD, Feigin
RD, et al, eds. Principles and practice of pediatrics. Philadelphia: JB Lippincott,
1990:553-572.
253. Rohr FJ, Levy HL, Shih VE. Inborn errors of metabolism. In: Walker WA,
Watkins JB, eds. Nutrition in pediatrics. Boston: Little, Brown, 1983:412-422.
254. American Academy of Pediatrics, Committee on Drugs. The transfer of

drugs and other chemicals into human milk. Pediatrics 1994;93:137-150.
255. American Academy of Pediatrics, Committee on Pediatric AIDS. Human

milk, breastfeeding, and transmission of human immunodeficiency virus in the
United States. Pediatrics 1995; 96:977-979.
256. Centers for Disease Control and Prevention. Recommendation for assisting
in the prevention of perinatal transmission of human T-lymphotropic virus type
III/lymphadenopathy-associated virus and acquired immunideficiency
syndrome. MMWR 1985;34:721-723.
257. Briggs GG, Freeman RK, Yaffe SJ. Drugs in Pregnancy and Lactation.
Baltimore: Williams & Wilkins, 1990.
258. Uauy-Dagach R, Mena P. Nutritional role of Omega-3 fatty acids during the
perinatal period. Clin Perinatol 1995;22:157-175.
259. Kovar MG, Serdula MK, Marks JS, et al. Review of the epidemiologic
evidence for an association between infant feeding and infant health. Pediatrics
1984;74:615-638.
260. Frank AL, Taber LH, Glezen WP, et al. Breast-feeding and respiratory virus
infection. Pediatrics 1982;70:239-245.
261. Saarinen UM. Prolonged breast feeding as prophylaxis for recurrent otitis
media. Acta Paediatr Scand 1982;71:567-571.
262. Lucas A, Cole TJ. Breast milk and neonatal necrotising enterocolitis. Lancet
1990;336:1519-1523.
263. Mortensen EL, Michaelsen KF, Sanders SA, et al. The association between
duration of breastfeeding and adult intelligence. JAMA 2002;287:2365-2371.
264. Mayer EJ, Hamman RF, Gay EC, et al. Reduced risk of IDDM among breastfed children. Diabetes 1988;37:1625-1632.
265. Koletzko S, Sherman P, Corey M, et al. Role of infant feeding practices in

development of Crohn's disease in childhood. Br Med J 1989;298:1617-1618.
266. Davis MK, Savitz DA, Graubard BI. Infant feeding and childhood cancer.
Lancet 1988;2:365-368.
267. Rosenblatt KA, Thomas DB. Lactation and the risk of epithelial ovarian
cancer. WHO Collaborative Study of Neoplasia and Steroid Contraceptives. Int J
Epidemiol 1993;22:192-197.
268. Newcomb PA, Storer BE, Longnecker MP, et al. Lactation and a reduced
risk of premenopausal breast cancer. N Engl J Med 1994;330:8-87.
269. Saarinen UM, Kajosaari M. Breastfeeding as prophylaxis against atopic

disease: prospective follow-up study until 17 years old. Lancet 1995;346:10651069.
270. Lucas A, Brooke OG, Morley R, et al. Early diet of preterm infants and
development of allergic or atopic disease: randomised prospective study. Br
Med J 1990;300:837-840.
271. American Academy of Pediatrics. Pediatric nutrition handbook, 4th ed.

Kleinman RE, ed. Elk Grove Village, IL: AAP, 1998:29-42.
P.412
272. Fomon SJ, Ziegler EE, Thomas LN, et al. Excretion of fat by normal fullterm infants fed various milks and formulas. Am J Clin Nutr 1970;23:1299-1313.
273. Henderson TR, Hamosh M, Hayman L. Serum long chain polyunsaturated

fatty acids (LC-PUFA) are adequate in full term breast fed infants in spite of low
milk LC-PUFA after >3 months lactation. Pediatr Res 1996;39:311A.
274. American Academy of Pediatrics, Committee on Nutrition. Soy protein

formulas. Recommendations for use in infant feeding. Pediatrics 1983;72:359363.
275. Saylor JD, Bahna SL. Anaphylaxis to casein hydrolysate formula. J Pediatr
1991;118:71-74.
276. Ellis MH, Short JA, Heiner DC. Anaphylaxis after ingestion of a recently
introduced hydrolyzed whey protein formula. J Pediatr 1991;118:74-77.
277. Schanler RJ. Human milk for preterm infants: Nutritional and immune
factors. Semin Perinatol 1989;13:69-77.
278. Uauy R, Hoffman DR. Essential fatty acid requirements for normal eye and
brain development. Semin Perinatol 1991;15: 449-455.
279. Carlson SE, Werkman SH, Rhodes PG, et al. Visual-acuity development in
healthy preterm infants: effect of marine-oil supplementation. Am J Clin Nutr
1993;58:35-42.
280. Kleinman RE, Walker WA. The enteromammary immune system. Dig Dis
Sci 1979;24:876-882.
281. Hutchens TW, Henry JF, Yip T-T, et al. Origin of intact lactoferrin and its
DNA-binding fragments found in the urine of human milk-fed preterm infants.
Evaluation by stable isotope enrichment. Pediatr Res 1991;29:243-250.
282. Balmer SE, Wharton BA. Diet and faecal flora in the newborn: breast milk
and infant formula. Arch Dis Child 1989;64: 1672-1677.
283. Kling PJ. Roles of erythropoietin in human milk. Acta Paediatr 2002;91
[Suppl]:31-35.
284. Dvorak B, Fituch CC, Williams CS, et al. Increased epidermal growth factor
levels in human milk of extremely premature infants. Pediatr Res 2003;54:1519.
285. Atkinson SA, Bryan MH, Anderson GH. Human milk feeding in premature
infants: protein, fat and carbohydrate balances in the first 2 weeks of life. J
Pediatr 1981;99:617-624.
286. Roy RN, Chance GW, Radde IC, et al. Late hyponatremia in very low birth
weight infants (<1.3 kilograms). Pediatr Res 1976;10:526-531.
287. Schanler RJ. Calcium and phosphorus absorption and retention in preterm
infants. J Exp Med 1991;2:24-29.
288. Schanler RJ, Oh W. Nitrogen and mineral balance in preterm infants fed
human milk or formula. JPEN 1985;4:214-219.
reference fetus. Growth 1976;40:329-341.
290. Stocks RJ, Davies DP, Allen F, et al. Loss of breast milk nutrients during
tube feeding. Arch Dis Child 1985;60:164-166.
291. Williams FH, Pittard WB 3rd. Human milk banking: Practical concerns for
feeding premature infants. J Am Diet Assoc 1981; 74(5):565-568.
292. Puczynski M, Rademaker D, Gatson RL. Burn injury related to improper

use of microwave ovens. Pediatrics 1983;72:714-715.
293. Henderson MJ, Dear PRF. Dicarboxylic aciduria and medium chain
triglyceride supplemented milk. Arch Dis Child 1986;61: 610-611.
294. Koo WW, Tsang RC. Mineral requirements of low-birth-weight infants. J Am

Col Nutr 1991;10:474-486.
295. Lake AM, Walker WA. Neonatal necrotizing enterocolitis: a disease of

altered host defense. Clin Gastroenterol 1977;6:463-480.
296. Hillman LS, Hoff N, Martin LA, et al. Osteopenia, hypocalcemia, and low 25hydroxyvitamin D (25-CHD) serum concentration with use of soy formula.
Pediatr Res 1979;13:A448(abst).
297. Georgieff MK. Taking a rational approach to the use of infant formulas.
Contemp Pediatrics 2001;18:112-130.
298. Georgieff MK, Hoffman JS, Pereira GR, et al. The effect of neonatal caloric
deprivation on head growth and one-year developmental status in preterm
299. Cooke RJ, Griffin I, Wells J, et al. Formula feeding preterm infants after
hospital discharge: 2. Effects on body composition. Pediatr Res 1996;39:306A.
300. Miller HC, Hassanien K. Diagnosis of impaired fetal growth in newborn

301. Georgieff MK, Sasanow SR, Chockalingam UM, et al. A comparison of the
mid-arm cirumference/head circumference ratio and ponderal index for the
evaluation of newborn infants after abnormal intrauterine growth. Acta Paediatr
Scand 1988;77: 214-219.
302. Bjerkedal T, Bakketeig L, Lehmann DH. Percentiles of birthweights of

single live births at different gestational periods. Acta Paediatr Scand
1973;62:449-457.
303. Sterky G. Swedish standard curves for intra-uterine growth. Pediatrics

1970;46:7-8.
304. Casey PH, Kraemer HC, Berbaum J, et al. Growth status and growth rates
of a varied sample of low birth weight, preterm infants: a longitudinal cohort
from birth to three years of age. J Pediatr 1991;119:599-604.
305. Georgieff MK, Mills MM, Zempel CE, et al. Catch-up growth, muscle and fat
accretion, and body proportionality of infants one year after newborn intensive
care. J Pediatr 1989;114: 288-292.
306. Hack M, Weissman B, Borawski-Clark E. Catch-up growth during childhood

among very low-birth-weight children. Arch Pediatr Adolesc Med
1996;150:1122-1129.
307. Gurney JM, Jelliffee D. Arm anthropometry in nutritional assessment:

normogram for rapid calculation of muscle circumference and cross-sectional
muscle and fat areas. Am J Clin Nutr 1973;26:912-915.
308. Georgieff MK, Amarnath UM, Murphy EL, et al. Serum transferrin levels in
the longitudinal assessment of protein energy status in preterm infants. J
Pediatr Gastroenterol Nutr 1989;8:234-239.
309. Shenai JP, Jhaveri BM, Reynolds JW, et al. Nutritional balance studies in
very low-birth-weight infants: role of soy formula. Pediatrics 1981;67:631-637.
310. Koo WW. Laboratory assessment of nutritional bone disease in infants. Clin
Biochem 1996;29:429-438.
311. Inder TE, Clemett RS, Austin NC, et al. High iron status in very low birth
weight infants is associated with an increased risk of retinopathy of prematurity.
J Pediatr 1997;131:541-544.
Chapter 23
Breastfeeding and the Use of Human Milk in the
Neonatal Intensive Care Unit
Kathleen A. Marinelli
Kathy Hamelin
OVERVIEW
Over the past two decades, as advances in technology have markedly improved our success in neonatal
medicine, we have concomitantly recognized that nutrition is the cornerstone of the care we provide to sick and
preterm neonates. The introduction and refinement of total parenteral nutrition, and the development of
specialty premature enteral formulas have paralleled these improved outcomes. Over this same time period,
there has also been an increasing awareness by both the general public and the medical community of the
short- and long-term advantages of both human milk and breastfeeding. Based on a rapidly increasing body of
research, there is no question that human milk is uniquely superior to other forms of nutrition for infants. The
most recent American Academy of Pediatrics (AAP) policy statement on the use of human milk reminds us that
the breastfed infant is the reference or normative model against which all alternative feeding methods must be
measured regarding growth, health, development, and other short- and long-term outcomes (1). This position is
endorsed and echoed by the American College of Obstetrics and Gynecology (2), the American Academy of
Family Physicians (3), the American Dietetic Association (4), and the Canadian Pediatric Society (5). The
evidence for the advantages of human milk to not only babies, but mothers, families and society in such diverse
areas as health, nutrition, development, and immunology, with psychological, social, economic, and
environmental impact (1,2,3,4) is so compelling that the U. S. government has made the support and promotion
of breastfeeding a national public health priority. With the release of the Surgeon General's HHS Blueprint for
Action on Breastfeeding (6) and the United States Breastfeeding Committee's Breastfeeding in the United
States: A National Agenda (7) the federal government has embraced the Healthy People 2010 breastfeeding
goals of 75% initiation, 50% breastfeeding at 6 months, and 25% breastfeeding at 1 year (8). The June 2004
launching of a 3 year, $40 million National Breastfeeding Awareness Campaign by the U.S. Department of
Health and Human Services, Office of Women's Health, with the tag line Babies were born to be breastfed
exclusively for 6 months, is clear indication of the commitment to breastfeeding promotion and support (9).
Produced in conjunction with the Ad Council, whose previous credits include Only you can prevent forest fires,
A mind is a terrible thing to waste, and Friends don't let friends drive drunk, these Public Service
Announcements will target the general market, particularly first time parents, and the African American
community, as rates of breastfeeding are lowest among this population (10).
In the ideal scenario, a discussion of breastfeeding generally conjures up a Madonna-like picture of a robust,
healthy term newborn, eagerly latched and nursing well with a mother who has had an uncomplicated delivery
and is supported and empowered by her ability to continue to nourish her baby. However, the reality of the
neonatal intensive care unit (NICU) is often in stark contrast to this. Are the uncertainty, the stress, the
technology, the constantly changing and critical nature of our patients and our
P.414
environment incompatible with the concept of breast-feeding? On the contrary, for these most vulnerable of our
patients and their families, the exponentially increasing body of research supports that both the provision of
human milk and breastfeeding are not only as important as they are in the full term population, but in fact may
be more critical to the ultimate health and developmental outcome of these sick and premature babies. The AAP
policy statement specifically recommends human milk as the preferred feeding not only for healthy term infants,
but for sick and premature infants as well (1).
It is incumbent on us as practitioners to have the knowledge and the expertise to promote and successfully
support lactation in the NICU population. Unfortunately, most of us received a cursory education on lactation, if
at all, during medical school, residency and fellowship training. The same is true of nursing, nurse practitioner
and physician assistant training. This is evidenced by the lack of physician knowledge of and confidence in the
subject (11,12,13), the variable and non-evidence-based information present in current general pediatric
textbooks (14) and the paucity of neonatal textbooks that even include it. This chapter presents our current
knowledge of the unique benefits of the use of human milk in a preterm and sick NICU population, supporting an
evidence-based rationale for its important role in our therapeutic regimen. It will also detail the challenges to
the provision of human milk, including the decision to express human milk and breastfeed, initiating and
maintaining lactation with a breast pump, breast milk supply, the use of donor human milk, the developmental
progression toward breastfeeding, the use of alternative feeding methods, and supporting breastfeeding in the
NICU and after discharge.
BACKGROUND
Prior to the advent of NICUs and the technology that has made them possible, most premature infants did not
survive. Those that were developmentally and physiologically mature enough did so if they could be kept warm,
and nourished. The source of that nourishment was human milk. It would then not be an unreasonable leap of
faith to say, that until this past century, the survival of premature infants was in large part dependent on the
provision of human milk. As early as 1907, Pierre Budin, at L'Hpital Maternit in Paris, encouraged mothers of
premature infants to breastfeed to improve survival (15). Julius Hess, who in Chicago began the first
continuously operating center for premature infants in the United States, wrote in 1922 by far the best results
are obtained in the premature infant weighing less than 1500 grams when it is fed human milk (16). He
advocated that human milk was the choice for feeding premature infants, with artificial milk a poor substitute,
resulting in increased mortality. It is astounding that at that time there was even positive discussion about the
survival of very low birth weight babies, let alone the association of improved survival with human milk feedings!
So why are we just now re-discovering the value of human milk in the neonatal unit? In 1947 Gordon showed
that premature infants fed two different formulas based on bovine milk gained weight faster than infants who
were fed human milk (17). It had also been previously shown that human milk did not support bone
mineralization in premature infants unless supplemented with calcium and phosphorus (18). Based on studies
like these, the use of human milk was abandoned in the United States for formulas that provided higher protein
and mineral intakes. Although the latter was an important observation that needs to be considered today, what
was not realized at the time, was how much was lost to obtain this gain.
INCIDENCE OF BREASTFEEDING OR PROVISION OF HUMAN MILK

Until very recently (19,20), the only on-going large-scale source of breastfeeding data in the United States has
been the Ross Laboratories Mothers Survey (21,22), which was developed in 1954 and is periodically updated.
Ross Products is a division of Abbott Laboratories, a manufacturer of infant formulas. Surveys are mailed to
large numbers of mothers who have given birth within the previous year. Included among concerns of using this
data to define breastfeeding rates and goals in the United States are low response rates (average 28% per
month since 1997) (21), inability to determine exclusivity of breastfeeding, no differentiation between
breastfeeding and breast milk feeding, and conflict of interest in a formula company monitoring breastfeeding
rates. Although the population of NICU babies or premature babies per se are not analyzed, data is reported on
the subset of babies with birth weights less than 2,500 grams, which would combine both premature babies and
small for gestational age term babies. For 1990 the in-hospital breastfeeding rates in this weight category are
given as 36.5% (compared to overall population breastfeeding rate of 51.5%) with steady increases to 62.7%
(compared to overall increase to 69.5%) in 2001 (21,22). In comparison, recently published data from the
Centers for Disease Control and Prevention (CDC), using questions added to the National Immunization Survey
for the third quarter of 2001, showed a similar overall U.S. initiation of 65.1% (95% CI: 59.5%-70.7%) (19).
The CDC has also looked at breastfeeding in 10 states using the Pregnancy Risk Assessment and Monitoring
System (PRAMS). For 1993 they report 57% overall initiation, with 40.3% for babies less than 2500 grams, and
49.4% for babies admitted to the NICU. This improved in 1998 to 67.5% overall initiation, 57.9% low birth
weight initiation, and 64.3% NICU initiation (20). These data, although less commercially biased, are still
difficult to interpret and compare to other studies as a result of inconsistency of definitions. The definition of
initiation of breastfeeding varies between studies, and often includes any attempt at putting a baby to breast or
initiating lactation with a breast pump in the first several
P.415
days of life or during the initial hospitalization. It also does not indicate whether other liquids are given at that
time.
These same reports have tried to look at duration of breastfeeding by surveying for any breastfeeding activity
later in the first year of life. The Ross Mothers Surveys report continued breastfeeding at 6 months in babies less
than 2,500 grams birth-weight increasing from 9.5% in 1990 to 22.1% in 2001 (with concomitant changes in
the entire population of 17.6% to 32.5%) (21,22). This data looked at babies with any breastfeeding;
including only breastfed, breast milk/breastfeeding mixed with formula feedings (no quantification determined),
and only occasional breast milk/breastfeeding (again no quantification). For 2001 they added the category of
exclusive breastfeeding, reporting for all infants in-hospital as 46.3%, falling to 17.2% at 6 months; for babies
less than 2,500 grams, 27.1% in-hospital falling to 8.4% at 6 months (21). Exclusive was defined as fed only
human milk; no supplemental formula and/or cow's milk. No information on solid foods or other non-formula
supplements (e.g., water, juices) fed to infants were collected (21). The PRAMS study looked at predominant
breastfeeding at 10 weeks postpartum and showed decreases between 1993 to 1998 of 58.5% to 57.9% in the
entire population, 47.9% to 45.1% in the low birth weight subgroup, and 55.1% to 47.3% in the NICU subgroup
(20). When duration data is given, it has historically been any breastfeeding or breast milk consumption, with
no differentiation made for whether other liquids or solids are ingested and in what quantities. This issue was
addressed in 1988 by the Interagency Group for Action on Breastfeeding, who developed a set of definitions to
standardize terminology. In the system they describe, full breastfeeding is distinguished from partial
breastfeeding, with full subdivided into categories of exclusive and almost-exclusive breastfeeding, and partial
differentiated into three levels (23). The hope was that consistent widespread implementation of these
definitions would assist researchers and agencies to describe, interpret and compare breastfeeding practices
accurately. Clearly, this has not occurred.
In March 2004, the Breastfeeding Committee for Canada issued a document on breastfeeding definitions. They
used work previously done in this area, including the definitions from the Interagency Group for Action on
Breastfeeding, to develop definitions in an algorithmic format that will facilitate data collection that is consistent,
and can be used to compare breastfeeding practices between Canadian provinces and territories (Table 23-1)
(24). These definitions are well thought out, and clearly delineate the amount of human milk vs. other liquids
that are being consumed. The one thing they do not do, however, is to separate out breast milk feedings from
breastfeeding.
When attempting to elucidate trends in breastfeeding and human milk consumption specifically in a NICU
population, one can examine a number of reports from individual NICUs. For example, in one author's NICU
(KM), breast-feeding rates have been tracked over a 15-year period. During this time, breast-feeding promotion
and support has been greatly increased (which will be detailed later in this chapter). In the entire NICU
population, only 20.2% of babies received any human milk (either by breastfeeding or alternative feeding
method) in 1989, which steadily increased to 70.8% in 2002. Continuing to receive any human milk at discharge
increased from 5.9% to 55.9% of all NICU discharges over this same time period. Even more striking is the data
for the population less than 1,500 grams: initiation increased from 10.6% to 82.6% with continued provision of
any human milk at discharge from 1.6% to 38.2% (25).
TABLE 23-1 BREASTFEEDING DEFINITIONS
Definition
Description
Exclusive Breast milk
No food or liquid other than breast milk, not even water, is given to the
infant from birth by the mother, health care provider, or family member/
supporter
Total breast milk
No food or liquid other than breast milk, not even water, is given to the
infant from birth by the mother, health care provider, or family member/
supporter during the past 7 days
Predominant breast milk Breast milk, given by the mother, health care provider, or family member/
supporter plus 1 or a maximum of 2 feeds of any food or liquid including
nonhuman milk, during the past 7 days
Partial breast milk
Breast milk, given by the mother, health care provider, or family member/
supporter plus 3 or more feeds of any food or liquid including nonhuman
milk, during the past 7 days
No breast milk
The infant/child receives no breast milk
Breast milk includes breastfeeding, expressed breast milk or donor milk and undiluted drops or syrups
consisting of vitamins, mineral supplements or medicines.
From The Breastfeeding Committee for Canada Breastfeeding Definitions, March 2004, with permission.
Although the numbers vary, overall trends show an increase in initiation and continuation of breastfeeding
activity over the past decade, with significantly fewer babies in low birth weight or NICU categories than
healthy, full term babies breastfeeding at any time point. Single institutional reports from the United States and
Canada have substantiated that NICU breastfeeding rates are lower than in their equivalent well-baby
populations (26,27,28,29), with rapid attrition over the course of the hospitalization and after discharge
(29,30,31,32,33). Factors often associated with continued lactation in this population are mothers who are
older, married, of Caucasian race, with more than a high school education, a good social support system and
having babies with increasing birth weight (27,30,33,34,35). An interesting recent multicenter study by Powers
and associates used an administrative database to look at breastfeeding in 42,891 neonates admitted to 124
NICUs from January 1999 to December 2000 (36). They show that 50% of neonates discharged home from
P.416
NICUs are not receiving human milk, and they confirm that greater birth weight, older gestational age, white
race, increasing maternal age, and married parents are associated with increased likelihood of receipt of at least
some human milk at discharge. Of note, they also showed that site of care is a significant independent factor
associated with human milk use. This raises questions regarding site differences that are more or less likely to
promote successful breastfeeding in an intensive care setting.
It is also important to note that these trends in North American NICUs are not necessarily replicated in other
parts of the world. As early as the 1980s, reports from European (37,38,39), Brazilian (40), and Australian (41)
units demonstrate breastfeeding initiation rates, duration, and eventual exclusivity in premature babies that are
equal to or higher than those we currently see in our term healthy population! Importantly, these countries have
breastfeeding cultures, in which breastfeeding in the term healthy population is the norm, with close to 100%
initiation.
ADVANTAGES OF HUMAN MILK FOR PREMATURE AND

COMPROMISED BABIES
Human milk is a living, changing fluid. It contains over 200 known components, including live lymphocytes,
macrophages, neutophils; immunoglobulins, complement, and other host defense factors; lactoferrin; enzymes
and hormones such as corticosteroids, erythropoietin, and insulin to name a few, in addition to its nutrients.
There are complex interactions between these components, which likely enhance and contribute to their
functions. It is ever-changing, from the beginning to the end of a feeding session, throughout the course of the
day, and over the course of lactation. It can never be replicated, no matter how much advertising to the
contrary is implied. Adding a new component to artificial formula that is present in human milk does not in any
way guarantee that its function or performance will be identical. Human milk has changed and adapted over the
course of human evolution to provide exactly what human infants need. It differs from the milk of other
mammalian species, including bovines, which have concomitantly evolved to provide what the young of each of
those species require to optimally grow and mature.
Accepting that human milk is the species-specific gold standard in nutrition for human babies, the outcomes
associated with its use are then the norm to which other forms of nutrition are compared. That being the case,
instead of a discussion of the benefits of breastfeeding as has been our habit, it is more accurate to look at the
risks of artificial feeding. There is a large evidence-based body of literature detailing improved health and
developmental outcomes of human milk-fed term babies and their mothers, over artificially fed babies, which
have been well reviewed elsewhere (1,2,3,4,42,43,44,45,46). The economic savings to families, health care
payers, employers, and society have also been studied. Table 23-2 summarizes these disadvantages of using
artificial formulas.
TABLE 23-2 SUMMARY OF THE DISADVANTAGES OF NOT BREASTFEEDING
Infant/Child:
Increased incidence/severity of:
Diarrheal diseases
Respiratory infections
Otitis media
Urinary tract infections
Infant botulism
Sudden infant death syndrome
Sepsis
Meningitis
Allergic diseases
Economic:
Families:
Cost of formula
Increased sick child visits
Increased medication use and costs
Lost wages for sick child care
Employers:
Employee absence for sick child care
Reduced productivity
Potential for higher health insurance rates
Payers:
Type 1 Diabetes mellitus
Increased physician fees
Celiac disease
Some childhood cancers (leukemias,
Increased emergency room fees

Laboratories/x-rays fees
lymphomas)
Inflammatory bowel disease
Increased prescriptions
Hospitalization costs
Mother:
Increased potential risk for:
Society:
Use of natural resources in production of
Postpartum blood loss

Anemia
artificial feedings
Environmental costs of production and wastes
Premenopausal breast cancer

Ovarian cancer
generated
Osteoporosis
Inability to take advantage of Lactational
Amenorrhea Method (LAM) of family planning
There is every reason to assume that these same advantages of human milk to term, healthy babies also apply
to preterm and sick neonates. Additionally, there is increasing research-based evidence of both short- and longterm positive effects on prematurity-related conditions, including nutrition, gastrointestinal (GI) function, host
defense, neurodevelopment, and physiological well-being.
Nutritional Advantages
Both the AAP (1) and the Canadian Paediatric Society (5) strongly recommend that breast milk is not only the
preferred nutrition for healthy term infants, but for all infants, including premature and sick newborns, with rare
exceptions (1). For an excellent in-depth examination of this topic please see Chapter 22. The reader is also
referred to a recent review article of the use of human milk for premature infants (47).
There are several points that are worth reiterating here. Preterm milk is different than term milk (Table 23-3).
Notably, preterm milk has higher concentrations of protein, fatty acids, sodium, and chloride (48,49), which
interestingly, are all components required in higher amounts by babies born early. This phenomenon was
initially attributed to lower milk volumes produced by mothers of preterm babies, thus causing a concentrating
effect on these nutrients.
P.417
However, contrary to this theory, other components of preterm milk are present in the same concentrations as
in term milk. It has subsequently been shown that preterm milk has similar volumes as term milk, so this is a
true occurrence. Some have speculated that this is a maternal adaptation to the delivery of her premature baby,
although others suggest that these differences are the end result of the interruption in maturation of the
mammary gland during pregnancy. In a recent study looking at total nitrogen, fat, lactose and carbohydrate
concentrations, gestational age at birth (GA) was inversely related to carbohydrate concentration;
postmenstrual age (PMA; an indicator of autonomous developmental processes not affected by the moment of
birth) was not related to milk composition; although postnatal age (PNA) was related to a decrease in total
nitrogen and an increase in lactose concentration. This data was interpreted to indicate that PNA strongly
influences the development of the composition of very preterm human milk, GA affects carbohydrate content
with a negligible effect on the nutritional value of the milk, although PMA has no effect (50).
TABLE 23-3 COMPARISON OF PRETERM TO TERM HUMAN MILK
Increased in Preterm Milk
Total nitrogen
Protein nitrogen
Long-chain fatty acids
Medium-chain fatty acids
Short-chain fatty acids
Sodium
Chloride
Magnesium (?)
Iron
Unchanged in Preterm Milk
Volume
Calories
Lactose (?less)
Fat
Linolenic acid
Potassium
Calcium
Phosphorous
Copper
Zinc
Osmolality
Vitamin B112
Lawrence KA, Lawrence RM. eds. Reprinted with permission from
Breastfeeding. A Guide for the Medical Profession, 1999:445.
The higher concentration of nutrients in preterm human milk all decrease to approximately term milk levels over
the course of the first postnatal month, regardless of the gestational age of the baby at delivery, although the
premature baby's increased needs continue until approximately term corrected gestational age. With the
advantage the earlier higher concentrations, particularly of protein and electrolytes, that premature milk
afforded no longer being present, we are often required to fortify human milk for the smallest babies. Babies
less than 1,500 grams have been shown to require fortification with more calories, protein, calcium, phosphorus,
sodium chloride and some vitamins to preclude poor growth rates, hyponatremia, hypochloremia, and
osteopenia (47). Larger, more mature babies thrive on mother's milk alone. Because human milk content differs
not only over the course of lactation, but during the course of a feed or milk expression session, at different
times of the day, and for those babies requiring feeding by alternative methods, by the method used, it is
critical to monitor these very low birth weight babies for growth rates, serum sodium levels, and bone
mineralization status (see Chapter 22).
Protein
Human milk protein is 80% whey, as opposed to bovine milk protein, which is 80% casein. The whey in human
milk, -lactalbumin, is much more easily digested than bovine whey, which is -lactalbumin, an important factor
to consider for premature babies with immature gut function. Additionally, human milk protein also includes
nucleotides, secretory immunoglobin A (sIgA) and other immunoglobulins, and an enzyme, lysozyme, all of
which are thought to aid in host defense; growth factors that stimulate gut growth and maturation; a variety of
hormones; and enzymes (e.g., mammary amylase, lipases) that enhance the immature intestinal tract's ability
to digest nutrients. The amino acid taurine, which serves many functions in the newborn including bile acid
conjugation, osmoregulation, neurotransmission, and as an antioxidant and a growth factor, is present in high
concentrations in human milk and is almost absent in bovine milk. Hence, it is added to artificial feedings. Unlike
bovine milk, human milk is also low in phenylalanine and tryrosine, which the premature and newborn infant are
poorly equipped to metabolize. For these reasons among others, human milk protein composition is welladapted to the needs of premature babies (see Chapter 22) (51).
Lipids
Human milk lipids have sparked the greatest interest in milk components recently, with the increasing body of
literature on the positive neuro-developmental effects of the long-chain polyunsaturated fatty acids (LC-PUFA's),
in particular docosahexaenoic acid (DHA) and arachidonic acid (AA). They are found in phospholipids in the
brain, retina and red blood cell membranes. These LC-PUFA's are not readily synthesized by preterm babies, and
are normally delivered via the placenta. They occur naturally in human milk, but are not found in bovine milk.
For premature babies, they must be delivered via an external source, in this case easily by human milk. Because
of studies that show improved neuro-developmental outcome and visual function in breastfed preterm babies
and in formula-fed preterm babies supplemented with exogenous sources of LC-PUFA's, DHA and AA have now
been added commercially to most term and preterm infant formulas. Of concern is that these additives are of
plant origin and are structurally different than human LC-PUFA's (see Neurodevelopmental Advantages).
Human milk lipids are an easily digested source of energy, in part as a result of their composition and in part as
a result of their packaging with lipases in the milk, providing approximately 50% of the calories.
P.418
Additionally they provide cholesterol, which is an essential component of membranes. Human milk-fed babies
show significantly higher plasma cholesterol levels than formula-fed babies (52). Although one might then
expect them to have higher cholesterol levels than formula-fed babies as adults, the opposite has been found to
be true (53). Additionally, coronary artery disease has been shown to be less frequent in persons up to 20 years
of age who were initially breastfed (54). It has been postulated that this early exogenous exposure to
cholesterol, a necessary nutrient, keeps endogenous cholesterol production down-regulated, thus resulting in
lower cholesterol levels in adult life. The mechanisms remain to be elucidated.
Carbohydrates
The disaccharide lactose is the predominant carbohydrate in human milk. It is a ready source of energy, and is
broken down by the enzyme lactase, located in the brush border of the intestinal mucosa, to galactose and
glucose, necessary for energy supply to the rapidly growing brain. Lactase activity is low in premature babies,
but is readily inducible by exposure to lactose, enabling them to absorb more than 90% from human milk. The
remaining unabsorbed lactose contributes to softer stool consistency, and colonization of the gut by
nonpathogenic fecal flora. It also enhances calcium absorption, critical to preventing nutritional rickets in
prematures. Oligosaccharides, present in human milk as well, act in host defense by preventing bacterial
attachment to intestinal mucosa, thus serving a protective role for the relatively immunocompromised preterm
infant.
Energy
Several previous studies have suggested that human milk-fed term (55,56) and preterm (57,58) infants have
lower sleeping energy expenditure compared with formula-fed infants. A recent randomized, cross-over study of
gavage-fed preterm babies showed significantly lower energy expenditure in the human milk fed babies at
prefeeding, during feeding, and postfeeding measurements (59). With the vast difference in composition of the
nutrients and other factors between human and artificial milk, it is not possible to say what causes this
difference, but it is an intriguing difference that remains to be investigated.
Gastrointestinal Advantages
In addition to the species specificity and superior digestibility of the nutrients in human milk for premature
babies, human milk also favorably affects the function and maturation of the GI tract. It has been shown in vivo
to decrease intestinal permeability in preterm infants when compared to preterm formula (60). Several studies
have found that human milk promotes more rapid gastric emptying than artificial formula (61,62), with one
revealing that on average, human milk emptied twice as fast as formula (62). This has implications for clinical
practice. Delayed gastric emptying, which generally presents clinically as measured gastric residuals or
vomiting, prevents advancement of enteral nutrition. Babies who cannot reach full enteral feeds require longer
periods of parenteral nutrition, with the concomitant risks inherent in both prolonged intravenous catheter usage
(e.g., infection, thrombosis, chemical infiltrates) and prolonged intravenous nutrition (e.g., mineral or
electrolyte imbalance, hepatic damage). Any of these can impact length of stay, which in turn has economic and
social/family implications, all of which are important to consider in the therapeutic plan in our NICUs today. In
fact, other studies suggesting that human milk is better tolerated by the GI tract of premature babies than
formula have looked at the marker time to full feeds. Several have shown that infants fed human milk achieve
full enteral feeds significantly faster than those fed artificial formulas (63,64). This would also be expected to
have a positive impact on length of stay.
Another related finding is the induction of lactase activity by feedings. Lactase, the enzyme responsible for the
digestion of lactose, is present in the fetal intestine early in gestation, but the greatest increase occurs during
the third trimester. Hence, premature babies are lactase-deficient at birth. In one study, lactase activity was
induced in preterm babies (26-30 weeks gestation) by the initiation of enteral feeds. Of greatest significance,
the highest levels of enzyme activity were seen with the introduction of early feeds (4 days old) as opposed to
standard feeds (15 days old) and in human milk-fed vs. formula-fed babies (65). There was also an inverse
correlation between lactase activity at 28 days, and the time to achieve full enteral feeds. It appears that the
level of lactase activity may be a marker of intestinal maturity, with human milk use being directly related to the
progression of that maturity.
There are a large number of bioactive components in human milk that are not present in formulas. They
variously provide antiinflammatory effects or protection from infectious agents; or are hormones and growth
factors that influence development; or are immune function modulators (66). This is an active area of research
with many questions remaining to be answered. At least a few of these factors have activity suggesting they
may be involved in GI maturation, growth and motility (67). Epidermal growth factor (EGF) is a major growthpromoting cytokine that stimulates proliferation of intestinal mucosa and epithelium and strengthens the
mucosal barrier to antigens (68,69). In an animal model, EGF isolated from human milk has been shown to
facilitate gut healing after induced injury (69). Other factors identified in human milk, known as human growth
factors I, II and III and insulin-like growth factor, have been shown to have growth-promoting functions,
including stimulation of deoxyribonucleic acid (DNA) synthesis and cellular proliferation. In vivo studies in animal
species have shown remarkable increases in the mass of intestinal mucosa after feedings with colostrum, which
contains these factors, but not after feedings with artificial milk (70,71). It is intriguing to postulate that
P.419
factors such as these in human milk may be important to the maturation and development of the premature GI
tract, as well as being important to intestinal repair after damage as a result of disease processes such as
necrotizing enterocolitis.
Host Defense Advantages

One of the single most extraordinary advantages of the use of human milk over formula in the preterm and
NICU patient is the effect on host defense and infections. This reason alone is enough to make the use of human
milk the gold standard in this population. The myriad of hormones, factors, cytokines, proteins, enzymes,
nucleotides, antioxidants, immunoglobulins, and live, functioning leukocytes contained within human milk (72),
their interactions and the milieu in which they effect action, is not and cannot ever be duplicated in an artificial
milk. And our knowledge is just beginning to scratch the surface.
One can think of human milk as not only the perfect form of nutrition, but also as baby's first immunization.
One type of immunization is the passive transfer of antibody. Examples include the use of intravenous immune
globulin, tetanus immune globulin, rabies immune globulin, or respiratory syncytial immune globulin. The
immunoglobulins represented in a mother's milk are a history of many of the infectious agents she has been
exposed to, and developed antibody to, during her lifetime. By transferring these antibodies to her baby through
her milk, she is then in effect immunizing her baby against those organisms. This process is a continuum from
the one begun in utero by the passage of maternal antibody across the placenta. Taking this one step further is
the concept of the enteromammary immune system, first proposed in 1979 by Kleinman and Walker (73) (Fig.
23-1). In this schema, antigen presented to the maternal mouth and gut is brought into proximity of lymphoid
follicles in the maternal GI tract. The antigen's presence commits maternal lymphoblasts to specific IgA
production against that antigen. These lymphoblasts migrate via the mesenteric nodes and the thoracic duct into
the systemic circulation, in which they find their way into the active mammary tissue. There these cells
manufacture sIgA, which is secreted into the milk. When the infant ingests the milk, the immunoglobulin present
functions in the infant's gut as protection against that specific pathogen. Most sIgA is not absorbed in the
infant's intestine, but rather plays an active role in mucosal defense. Although intact immunoglobulin has been
found in infant urine, signifying some systemic absorption, most survives intact in the GI tract, and is excreted
into the feces. The environment of the neonatal intensive care unit is full of potentially pathogenic organisms.
What makes this concept of the enteromammary immune system even more appealing is that during skin to
skin (kangaroo) care, a neonate is held by the mother at her chest, and is touched, kissed, and caressed. The
mother exposes herself to any potential pathogens to which her baby has also been in contact. Through the
enteromammary immune system, she makes antibody to those particular organisms, and then in subsequent
feedings, she passes that antibody to her baby, to help protect him/her. Imagineindividually prepared
immunizations to help protect each different baby in the NICU!
Similarly to our previous discussion of concentrations of specific nutrients, many of these immune modulators
are in higher concentration in preterm milk than term milk, helping to compensate for the premature baby's
immature immune function (Table 23-4). When major factors were quantified and compared between colostrum
of mothers delivering both prematurely (28-36 weeks) and at term (38-40 weeks), mean concentrations of IgA,
lysozyme, lactoferrin, and absolute counts of total cells, macrophages, lymphocytes, and neutrophils were found
to be significantly higher in the preterm colostrum (74). The degree of prematurity did not influence the
antiinfective levels in the colostrum. However, the total cells and macrophages were significantly higher in the
colostrum of mothers delivering at 28-32 weeks of gestation compared to 33-36 weeks gestation (p < 0.05)
(74). These differences make the use of early colostrum and human milk critical in the care of premature and
sick babies, both as prevention and also possibly in treatment of infectious disease.
Another likely factor in improved immune defense with the use of human milk is the affect on fecal flora. The
normal flora of the intestines of a breastfed infant is predominantly the gram positive bacteria Lactobacillus
bifidus. Non-human milk-fed babies are colonized with many more types of bacteria, most of which are the
potentially pathogenic gram negative bacteria. With establishment of Lactobacillus as the predominant bacteria
inhabiting the premature infant's GI tract, the likelihood of a serious or life-threatening gram negative infection
would be expected to decrease.
It has long been known that breastfed babies are at decreased risk of a number of infectious diseases
respiratory infections, otitis media, gastroenteritis, and diarrheaand are at decreased risk of mortality. These
protections accrue in a dose-dependent fashioni.e., the more breast milk a baby receives, statistically the
better protected they are. Although much of the earlier work was done in Third-World countries, and commonly
held beliefs are that breastfeeding does not make a difference in industrialized nations like the United States,
there are now many studies that show significant impact in these populations as well (75,76,77,78). There has
even been a study looking at breast-feeding and the risk of postneonatal deaths in the United States which
shows promoting breastfeeding has the potential to save or delay about 720 postneonatal deaths each year
(79)! It is probably fair to assume that babies admitted to NICUs, once attaining term corrected gestational age
and discharged to home, will accrue similar advantages from breast milk/breastfeeding as their healthier term
counterparts in these studies. But even more importantly for our population, an increasing body of research has
been done looking at the effects of a diet of
P.420
human milk compared to preterm formula in premature and low birth weight babies, with respect to clinical
infections. The data is clearpremature babies fed human milk are at significantly less risk of serious diseases,
including necrotizing enterocolitis, urinary tract infections, sepsis, and meningitis.
Figure 23-1 The Enteromammary Immune System. Adapted from Kleinman RE, Walker, WA. The
enteromammary immune system: an important new concept in breast milk host defense. Dig Dis Sci
1979;24:880.
As early as 1971, a report from Sweden showed a protective effect of breastfeeding against sepsis in the
newborn (80). Then in 1980, Narayanan and her group in India reported that even partial use of human milk
(81), and subsequently exclusive use (82) could significantly decrease the incidence of infection in a premature
and low birth weight population. In the first study, infants fed human milk supplemented with formula had a 6%
incidence of sepsis, compared to a 21% incidence in those fed formula alone (81). In the second, in 62 infants
studied, no sepsis occurred in those infants receiving human milk exclusively, although six episodes occurred in
those fed formula (82). A further study by this group looked at the dose-response effect of human milk in the
prevention of infection (83). Infections noted were sepsis, diarrhea, pneumonia, meningitis, conjunctivitis,
pyoderma, thrush, and upper respiratory infections. The strongest effect was noted in those babies fed exclusive
human milk, followed by partial feedings of human milk. In a United States study, a decrease in blood culture
proven sepsis was seen in those infants fed their own mothers' milk compared to those fed formula (27% vs.
58%, p < 0.05) (63) More recently in another U.S. neonatal unit, El-Mohandes and associates showed a
significantly decreased incidence of sepsis, in each of three time periods through the first 38 days of life, for
infants admitted to the NICU who received human milk vs. formula (odds ratio for sepsis in human milk-fed
infants was 0.4, 95% confidence limits, 0.15 to 0.95, p =0.04) (84).
TABLE 23-4 COMPARISON OF ANTIINFECTIVE PROPERTIES OF PRETERM AND TERM MOTHER'S
COLOSTRUM
Preterm Colostrum
Term Colostrum
Total protein (g/L)
0.43 1.3
0.31 0.05*
IgA (mg/g protein)

IgG (mg/g protein)
310.5 70
7.6 3.9
168.2 21*
8.4 1
IgM (mg/g protein)

Lysozyme (mg/g protein)
39.6 23
1.5 0.5
36.1 16
1.1 0.3*
Lactoferrin (mg/g protein)

Total cells/ ml3
165 37
6794 1946
102 25*
3064 424*
Macrophages
Lymphocytes
4041 1420
1850 543
1597 303*
954 143*
842 404
512 178**
Neutrophils
* p <0.001; ** p < 0.005
Modified from Mathur NB, Dwarkadas AM, Sharma VK, et al. Anti-infective factors in preterm human
colostrum. Acta Paediatr Scand 1990;79;1039-1044.
P.421
Hylander and colleagues published an elegant study in 1998 in which they looked at 212 VLBW infants born at
Georgetown University Medical Center (85). They not only looked at incidence of infection among these babies
in relationship to their type of feeding, but they also controlled for confounding factors. Infection was
documented as clinical signs of sepsis along with positive cultures for pathogenic organisms from one or more of
the following sites: blood, spinal fluid, urine, stool, pleural fluid, nasopharyngeal, intravascular catheter,
umbilicus, eye, or surgical wound. Additionally, pneumonia (by chest radiograph) and necrotizing enterocolitis
(Bell's classification) were included. Human milk feeding was defined as receiving any human milk, with
supplemental formula feedings when milk was not available. Formula-fed babies received only formula. They
found that the incidence of infection (human milk 29.3% vs. formula 47.2%) and sepsis/ meningitis (human
milk 19.5% vs. formula 32.6%) differed significantly by type of feeding (Fig. 23-2). Human milk was
independently correlated with a reduced odds ratio of infection (OR = 0.43, 95% CI, 0.23-0.81), controlling for
gestational age, 5-minute APGAR score, days on mechanical ventilation, and days without enteral feeds. Human
milk feeding was also independently correlated with a reduced odds ratio of sepsis/meningitis (OR = 0.47, 95%
CI 0.23-0.95), controlling for gestational age, mechanical ventilation days, and days without enteral feeds. And
remember, the human milk-fed babies were supplemented with formula, so if they had been exclusively
breastfed, one wonders if the difference would have been even more remarkable?
Figure 23-2 The incidence of all infections and of significant infections (sepsis/meningitis) in VLBW babies in
formula vs. breast milk-fed groups, is significantly different even when controlling for confounding factors.
Infection OR = 0.43; 95% CI 0.23-0.81; sepsis/meningitis OR = 0.47; 95% CI 0.23-9.95) n = 212. From
Hylander MA, Strobino DM, Dhanireddy R. Human milk feedings and infection among very low birth weight
infants. Pediatrics 1998;102:e38.
In 1999, Schanler and his colleagues showed that in a group of premature infants between 26 to 30 weeks
gestation, those fed predominantly fortified human milk were discharged earlier (73 19 vs. 88 47 days) and
had lower incidence of necrotizing enterocolitis (NEC) and late-onset sepsis than infants fed preterm formula
(86). This data on NEC confirms a previous study by Lucas and Cole (87), who showed that in a cohort of 926
infants with birthweights below 1850 grams, formula fed babies were six to ten times more likely to develop
NEC than those fed human milk exclusively; and three times more likely than babies who received a
combination of human milk and formula supplements. Once again, this data shows a dose-response effect of
human milk. While reviewing the data on NEC it is also important to point out that in addition to these
significant clinical decreases in incidence of NEC, there are also concomitant significant savings in economic
costs. This is an enormous issue to us as practitioners, to the health care system as a whole, and certainly to
the individual families we care for. A reduction in the cases of NEC would engender significant savings in medical
charges, and length of stay (LOS). Bisquera and associates (88) showed that infants with surgical NEC exceeded
LOS by 60 days over matched controls, and medical NEC by 22 days over
P.422
controls. Based on LOS, the estimated total hospital charges per baby for surgical NEC averaged $186,200 more
and medical NEC $73,700 more than controls. This translated into yearly additional hospital charges at their
institution for NEC of $6.5 million, or $216,666 per survivor! When parents' time and wages and prevention of
premature deaths are also considered, the total savings in the United States alone are more than $3 billion if
only 75% of preterm infants received human milk feedings and their incidence of NEC decreased from 7% to 1%
(89)! The significance of these statistics cannot be ignored.
Just as the etiology of NEC is not completely understood, and is certainly multi-factorial, so the mechanism
underlying the protection against NEC afforded by breast milk is also not completely understood, and most likely
multi-factorial. One study showed a decreased prevalence of NEC in premature infants who were given enteral
doses of a serum-derived IgA-IgG product, with IgA predominance, similarly to that found in human milk (90).
This leads one to postulate that the immunoglobulins, in particular IgA, present in human milk have a role to
play in NEC prevention. It is also known that platelet-activating factor (PAF) is one of the inflammatory
mediators that is known to be higher in babies with NEC, and that in experimental models, it causes bowel
necrosis similar to that seen in NEC, when injected intravenously. PAF is rapidly broken down by PAF
acetylhydrolase, which has been shown to be not only present in human milk, but to be approximately 5-fold
higher in preterm than in term milk (91). This may yet be another mechanism by which breast milk is
protective. Additionally we have previously discussed the predominance of nonpathogenic fecal flora in breastfed
infants. Could this not be another protective mechanism? There are certainly many other biologically active
substances in human milk that may ultimately be found to have a role in the prevention of NEC.
It is unclear how much human milk is needed to see these effects. In the previously discussed Hylander study
(85), the human milk-fed group also received supplements of formula and still showed significant benefit.
Narayanan (83) showed a dose-response in that exclusive human milk was more protective than partial, which
was still more protective than no human milk. And Lucas and Cole showed that in NEC, partial human milk
feeding was protective, but less so than exclusive human milk feedings (87). Furman and colleagues, looking at
119 VLBW babies, identified a daily threshold amount of at least 50 ml/kg/day of maternal milk through week 4
of life as needed to decrease rates of sepsis in this group (92). These data are convincing that by providing
human milk, we do make a difference in these serious morbidities in even the tiniest of our patients. And it
certainly appears that there is a dose-response relationship. That being the case, with all the potential
morbidities that our patients face, and the costs both economically and emotionally for these families and
society, it is without a doubt worth our efforts to make human milk the gold standard in neonatal intensive care
for these reasons alone. It has also been shown that upper respiratory symptoms are reduced for those low
birth weight babies through seven months corrected age who continue to receive human milk after discharge
from the NICU (93). Although not significant, a trend appears to exist for otitis media, bronchiolitis, and
gastroenteritis as well. More studies with larger cohorts are needed to confirm these data.
Neurodevelopmental Advantages
Much press has been given to recent work that shows improved cognitive development in babies who receive
human milk. In 1988 Morley and associates showed an 8-point cognitive advantage using the Bayley Scales of
Infant Development for 771 infants with birth weights less than 1850 grams (94). After controlling for
demographic and perinatal factors, a 4.3-point advantage still remained. When this cohort was followed up at
7.5 to 8 years of age, infants who had received human milk by tube (rather than breastfeeding) continued to
show an 8.3-point IQ advantage (over half a standard deviation) even after adjustments were made for
difference in mother's education and social class (95). They also showed a dose response relationship between
the proportion of human milk in the diet and subsequent IQ. In a second randomized prospective study, these
same researchers compared premature infants who received mature donor human milk with those who received
premature formula as their early enteral nutrition. These diets were compared as sole enteral feeds or as
supplemental feeding to their own mother's expressed milk. No differences in outcome were seen at 18 months
between the two diet groups, despite the low nutrient content of mature donor milk in relation to preterm
formula. Additionally, when they compared the infants from this study fed solely on mature donor milk with
infants from a previous study fed solely standard term formula, the infants fed the mature donor milk had
higher developmental scores, again supporting the positive role of human milk on cognitive development (96).
Bier and colleagues have also looked at the effects of human milk on both cognitive and motor development.
They found that in a cohort of 39 premature infants, those who received human milk had significantly higher
motor scores at 3 and 12 months corrected ages than those fed formula. Additionally, when adjusting for
oxygen requirement and maternal vocabulary scores, human milk-fed babies continued to show an advantage in
both cognitive and motor scores at 12 months. There was also an association between the amount of human
milk intake while in the special care nursery (dose response relationship!), and cognitive development at both 7
and 12 months corrected ages (97).
There has been one meta-analysis of controlled studies looking at the question of human milk and cognitive
development (98). It demonstrated a 3.16-point higher score for cognitive development in human milk-fed
babies compared with formula after adjustment for significant covariates. This difference was observed as early
as 6
P.423
months and was sustained through 15 years of age, the last time of reliable measurement. Longer duration of
breast-feeding was accompanied by greater differences in cognitive development (dose related response).
Whereas normal weight infants showed a 2.66-point difference in IQ scores between breastfed and formula fed
groups, the difference was even more remarkable, a 5.18-point difference, in low-birth-weight infants. The
results of this meta-analysis suggest that not only does human milk contribute to higher neurodevelopment, but
that the effect is even more striking in a much more high-risk population, the low-birth-weight infant.
Postulated to be responsible for at least a portion of these neurodevelopmental advantages of human milk is the
presence of long-chain polyunsaturated fatty acids (LC-PUFAs), which until recently, were not present in
formulas (Please see Chapter 22 for a more in-depth review). Docosahexaenoic acid (DHA), normally accounts
for greater than one-third of the total fatty acids of the gray matter of the brain and the retina of the eye (99).
Most of the prenatal accumulation of DHA in these tissues occurs in the third trimesterthus by definition,
premature infants are deficient compared to their term counterparts. Animal studies have shown that deficiency
of DHA in neural tissues during development leads to behavioral and retinal changes.
Other examples of the effects of human milk on neurological maturation have also been observed. Premature
infants receiving human milk have been shown to have faster brainstem maturation than those on formula
(100). Visual acuity and the development of retinopathy of prematurity have also been studied. A number of
studies were performed before the routine addition of the LC-PUFAs to preterm formulas. In one, the
researchers showed improved visual acuity of preterm infants up to four months of age in supplemented vs.
nonsupplemented formulas (99). In another, improved retinal function was present in LC-PUFA sufficient VLBW
neonates fed human milk or supplemented formula as compared to an unsupplemented formula group with
lower LC-PUFA cell composition (101). Better visual evoked potentials and acuity in both preterm and term
infants at 57 weeks postconception has been seen in those fed human milk as opposed to those fed formula
(102). Additionally, another interesting study in which healthy term infants who breastfed to 4 or 6 months, and
then were weaned, were randomly assigned to commercial formulas with or without DHA and arachidonic acid
(ARA) supplements. At 1 year of age, the level of DHA as measured in the red blood cells was reduced by 50%
from weaning level in the unsupplemented group, although there was an increase of 24% in the supplemented
group (103). The conclusions drawn from this study were that the critical period during which a dietary supply of
DHA and ARA can contribute to optimizing visual development in term infants extends through the first year of
life. This supports the AAP recommendation that breastfeeding continue through the first year of life (1), and
begs consideration then of the length of time breastfeeding should be encouraged and supported for the baby
born prematurely.
Physiological Advantages
Breastfeeding is widely assumed to be more stressful than bottle feeding for premature infants, an assumption
that has led in many intensive care units to the introduction of bottle feedings as the first oral feeds, and
postponing attempts at breastfeeding until babies can prove themselves with the bottle. There are also often
concerns for the small premature baby's ability to maintain temperature while breastfeeding, leading to rules
restricting initiating breastfeeding until a certain weight is achieved. Additionally, there is the widely held belief
that the suck-swallow-breathe mechanism is not mature until approximately 34 weeks gestation. Because of
the fear that they will choke, desaturate and aspirate, this leads to more rules concerning not initiating oral
feeds, including breastfeeding, until at least 34 weeks corrected gestational age. It is important to point out that
there is no scientific evidence to back any of these statements. On the contrary, the exact opposite is the case
as we will discuss. It is also important to understand that these policies, in addition to being unsupported
scientifically, are also harmful in that they (a) preclude premature babies and their mothers from early
breastfeeding experiences; (b) allow babies whose mothers want to breastfeed to learn to suck from a bottle,
which then in many cases makes it hard for them to transfer to the breast at a latter time, as the sucking
mechanisms are different; and (c) introduces breastfeeding so late in the hospital stay, that in addition to
struggling to overcome what they have learned with an artificial nipple, the mother and baby are often
discharged before they have had time to learn together and develop the confidence and skills necessary to allow
successful breastfeeding in this population. It is important to re-emphasize our medical dictum in this
circumstance: Primum non nocere.
Data does exist that enables us to develop breastfeeding policies that are physiologic, safe and supportive of
breast-feeding. As early as the 1980s, Meier was publishing data concerning physiologic stability at breast
compared to bottle-feeding. She was able to show that in babies less than 1,500 grams at the time of first
feeding, different sucking mechanisms were employed for breast and bottle, with better coordination of suckswallow-breathe during breastfeeding, particularly in the smaller less mature babies. Concomitantly, there are
markedly different patterns of transcutaneous oxygen pressure (tcPO2) for the two methods of feeding. TcPO2
patterns suggest less ventilatory interruption during breastfeeding than during bottle-feeding with greater
declines in tcPO2 during bottle-feeding than during breast-feeding over the course of a feeding session.
Additionally, babies became significantly warmer during breastfeeding than during bottle-feeding (104,105,106).
Similar studies done by Bier and colleagues in first VLBW (107), and then ELBW (108) babies showed they could
tolerate beginning breast- and bottle feedings at the same postnatal age; that they were
P.424
less likely to have oxygen desaturations to less than 90% during breastfeeding; and that they had lower intakes
during breastfeeding. Given that there is no data supporting the safety of initiating bottles as first feeds, and
that there is data to show that during breastfeeding premature babies are more physiologically stable, it would
then seem reasonable, safe and scientifically sound to allow mothers to put their babies to nuzzle at breast and
to begin early steps toward breastfeeding, once they are physiologically ready!
CHALLENGES TO THE PROVISION OF HUMAN MILK

The Decision to Express Human Milk and Breastfeed
Expressing human milk and subsequently breastfeeding a premature infant has been described as a contribution
to the care of her infant that only the mother can make. This is a very powerful, tangible opportunity for these
mothers. Additionally, this feeding choice has been considered the one aspect of natural caregiving that the
mother does not have to forfeit when she delivers a sick or premature infant. In the past, the decision to
provide expressed human milk for the premature infant was regarded as a matter of parental choice with
professional responsibility limited to implementing the parent's decision. With increasing and irrefutable
evidence that human milk supports the optimal health and nutrition of sick and premature infants, NICU care
providers must actively encourage mothers to initiate lactation and provide milk to their infants, even if they do
not plan to ultimately feed at the breast.
Parental decisions related to infant feeding must be based on informed choice. Emphasizing the importance of
human milk to the health of the infant assists mothers in making an informed decision about feeding method.
Health care providers should share research-based evidence about the superiority of human milk and the value
of this intervention for the sick or premature infant. It is also important to communicate the maternal benefits of
lactation to these mothers, both in terms of her immediate ability to be actively involved in the care of her baby,
no matter how sick he or she is, and also in terms of the health benefits she will accrue (2,3). Optimally, this
information should be discussed with the family as soon as the premature birth of their baby becomes a
possibility. With the understanding that this is not always feasible, it must be discussed as soon after the
infant's birth as realistic, so that milk expression can be initiated and colostrum expressed when the hormones
of lactation are optimal. Even mothers who had indicated prior to the birth of their child that they desired to
feed formula, are often more than willing to initiate milk expression to obtain colostrum for their babies when
they understand all the advantages it confers. Written information should reinforce that early human milk is
considered medicine that prevents infection in this vulnerable population. Although staff may worry that this
approach is coercive or may make mothers feel guilty about initial feeding plans, mothers who are educated to
make an informed choice to provide human milk for their infant report they are thankful for the information that
assisted them to support the health of their infant. It is important that families understand that a decision to
initiate milk expression does not commit them to breastfeeding. They can decide to stop at any point along the
continuum. This is very empowering to these women, who with the birth of a baby who requires intensive care,
acutely feel the loss of normal maternal control. Research has demonstrated that structured education and
support programs within NICUs are effective in increasing lactation initiation rates, often in mothers who had
intended to formula feed (109,110). The positive breastfeeding outcomes of such programs clarify the critical
role of health care providers in sharing the science of human milk with mothers so that they can make an
informed choice related to infant feeding.
Initiating and Maintaining Lactation with a Breast Pump

Mothers who provide breast milk and eventually breastfeed their premature infants must develop their milk
supply through manual or mechanical milk expression techniques. These techniques may be suboptimal
compared to actually breastfeeding in stimulating and maintaining a full milk supply. As a result, compromised
milk production is a well-documented constraint to extended breastfeeding in the premature mother-infant
dyad. An understanding of the physiology of lactogenesis and the effects of various pumps and pumping styles
on milk production is critical to understanding this issue and supporting mothers to establish and maintain
optimal milk volumes.
The transition from pregnancy to lactation is called lactogenesis. The first half of pregnancy is characterized by
growth and proliferation of the mammary ductal system; during the second half of pregnancy, secretory activity
increases and the alveoli become distended by accumulating colostrum (111). The capacity of the mammary
gland to secrete milk after approximately 16 weeks gestation is referred to as lactogenesis I. The onset of
copious milk secretion 2 to 8 days after birth is called lactogenesis II. Triggered by a rapid drop in serum
progesterone levels after the delivery of the placenta, lactogenesis II results in a rapid increase in milk volumes
from 36 to 96 hours postpartum. Ongoing lactation is dependent on a delicate interplay of hormones and
effective stimulation and emptying of the breast. Interference with these processes can delay and/or suppress
milk production (112).
Prolactin and oxytocin are the predominant hormones of lactation. Prolactin, the milk-making hormone, is
essential for both initiating and maintaining milk supply. As progesterone and estrogen levels abruptly drop after
birth, the anterior pituitary gland is no longer inhibited by these
P.425
hormones and releases pulsatile surges of prolactin in response to suckling at the breast. Frequent
breastfeeding in early lactation stimulates the development of receptors to prolactin in the mammary glands and
results in a faster increase in milk synthesis (113). The number of prolactin receptors increases in early
breastfeeding and remains constant thereafter (114,115). Therefore, early breast stimulation during this critical
period is a predictor of later milk production. Oxytocin is the hormone responsible for the removal of milk from
the breast. Excreted by the posterior pituitary gland in response to suckling, oxytocin causes the myoepithelial
cells surrounding the alveoli to contract and milk to be ejected into the ducts in which it is available to the
newborn. The secretion of oxytocin is negatively affected by stress and pain. Therefore, the birth of a premature
infant (often by cesarean delivery) and the subsequent stressors associated with this event can interfere with
the milk ejection reflex or let down of milk from the breast.
After the first few days postpartum, lactation shifts from endocrine control (hormone driven) to autocrine control
(driven by milk removal). Galactopoiesis (the maintenance of milk production) is driven by the quality and
quantity of milk removal. As long as milk is being removed from the breasts, the alveolar cells will continue to
make milk. This supply-demand phenomenon regulates milk production to match intake by the infant (112).
The birth of a premature infant can negatively influence milk production. If a mother has delivered very
prematurely, mammary development may be poor because the mother may not have received the full
component of pregnancy related hormones to prepare the breasts for lactation (116,117). Additionally, the close
infant contact most frequently experienced by mothers following term delivery is limited or absent after
premature delivery. As a result, the neurohormonal stimulus of the lactogenic hormones is impaired. An
additional barrier to optimal milk expression is the fact that anxiety, fatigue and emotional stress, all powerful
inhibitors of lactation, are experienced almost universally in mothers of premature infants (118). As a result, for
mothers of premature infants, concerns related to milk production and transfer of adequate milk to the infant
are primary reasons for discontinuing breastfeeding or providing supplements. Early, frequent and optimal
stimulation of maternal milk supply must replace the natural breastfeeding process to ensure adequate milk
production and duration of breastfeeding in this population.
Current research points to three factors that are independently associated with optimizing milk production in
mothers who are pumping for a preterm baby. After controlling for maternal age, race, marital status and
maternal education, factors significantly associated with ongoing milk supply and breastfeeding at term include
initiating breast pumping before 6 hours postpartum, pumping more that 6 times a day, and skin-to-skin contact
with the infant (119,120,121). An additional factor that optimizes ongoing milk production is expressing milk for
an adequate duration of time to completely empty the breasts (120). The degree of breast emptying is a strong
stimulus for milk synthesis and may be even more important than the frequency of pumping. Clinical advice
therefore should include early (<6 hours postpartum), frequent (8-10 times in 24 hours) and effective pumping
to achieve milk volumes between 800 and 1000 ml per day by 2 weeks postpartum (112) (Table 23-5). This
oversupply of milk provides a reserve against diminishing milk production later in lactation.
TABLE 23-5 OPTIMAL INITIAL MILK EXPRESSION REGIME
Lactogenesis IEstablishing initial milk supply
Begin milk expression as soon after birth as possible (optimally within 6 hours)
Pump frequently (no less than 8 times in 24 hours)
Pump at least once at night (between 1 AM and 4 AM)
Use a full-sized (hospital grade) electric breast pump with the ability to pump both breasts
simultaneously
Increase pump suction until milk is flowing and comfort maintained
Hold infant skin-to-skin prior to pumping if possible
Use breast massage prior to and during milk expression
Pump for 10 to 15 minutes and/or until all milk droplets cease flowing
Maximize rest and minimize stress as much as possible
Lactogenesis IIMaintaining milk supply
Continue with above regime

To ensure breasts are emptied, do not pump a specific amount of time, but pump until milk stops
flowing, for 2 to 3 minutes
Mothers who initiate long-term milk expression require a hospital grade electric breast pump; the clinical
challenge is ensuring that these pumps are available. Although hospital-grade electric pumps are available
through pharmacies, lactation consultants and home health agencies, many mothers are unable to incur this
expense. Support and advocacy from health care providers related to the necessity of this equipment for the
health of the infant may assist with payment from third party payers. Although some researchers advocate
simultaneous pumping as a significant predictor of eventual milk volumes (122,123), other studies have not
supported the necessity of pumping both breasts simultaneously (119). However, simultaneous (double)
pumping requires less time and effort than sequential (single) pumping; this difference may influence ultimate
maternal commitment and ongoing willingness to continue pumping over time.
Human Milk Supply

Despite their best efforts, mothers of premature infants may experience persistent low milk volumes or milk
production that decreases over time. Researchers have described a frequent phenomenon of a decreasing milk
supply during the second month of pumping (124), which is very often seen by these authors. Because a breast
pump does not mimic the physical contact and closeness of an infant during breastfeeding, suboptimal hormonal
stimulation
P.426
may play a role in decreasing milk supply. Little is known about the physiologic response of mechanical milk
expression to milk production. Therefore strategies to manage low milk volumes in mothers of premature infants
have focused on pharmacological and nonpharmacological enhancement of prolactin secretion
(112,125,126,127).
Kangaroo Care, also called skin-to-skin Care, is an effective nonpharmacological method to enhance maternal
milk production through skin-to-skin contact. This intimate contact stimulates the release of the hormones of
lactation and positively influences maternal milk supply (128,129). Mothers who engage in Kangaroo Care
frequently note feelings of milk-ejection and report pumping larger volumes immediately following holding their
infants skin-to-skin. Hurst and colleagues report that mothers who held their infants in Kangaroo Care for 30
minutes per day had significantly greater milk volumes both early (2 weeks) and later (4 weeks) in the
postpartum period (128). Breast massage has also been demonstrated to increase volumes of pumped milk. It
is done similarly to a breast self-exam. Using the palmar surface of her three middle fingers, a mother starts
close to the chest wall and slowly massages in a circular motion, moving toward the areola. Mothers should be
instructed to palpate each breast after pumping and subsequently massage the firmer areas to promote
complete breast emptying (117).
Ongoing clinical observation of pumped milk volumes will ensure timely intervention to prevent significantly
inadequate milk production. Because most premature infants require a minimum of 500 mL per day at
discharge, intervention to increase milk volumes should be put into place when milk volumes fall below this level
(112). Nonpharmacological interventions to increase milk volumes include increasing pumping frequency and
resuming and/or increasing pumping at night, pumping at the infant's bedside, or while holding a picture or an
article of baby's clothing or bedding; alternatively pumping in a quiet, low-stress environment; and covering the
milk collection bottles with a receiving blanket or towel while pumping (decreases anxiousness, stress
hormones? It really works!). Pharmacological measures however should be considered if alternative strategies to
increase milk volumes have been employed unsuccessfully for 7 days, milk supply is not fully established despite
optimal breast stimulation, and/or milk supply has decreased significantly (by 50% or <500 mL/day) after
optimal initiation.
Pharmacological prolactin enhancers are another strategy to assist with increasing milk supply. Dopamine
inhibitors such as Metoclopramide and Domperidone have been used to induce lactation in clinical situations with
varying results (125,126,130). Both increase circulating prolactin levels and, with accompanying breast
stimulation, are postulated to cause an increase in prolactin receptors and subsequent milk production (113). As
a result, these medications are most effective in the early postpartum period when the breast tissue is sensitive
to the formation of prolactin receptors. Companies that manufacture these drugs do not endorse their use as
galactagogues, so they are used off label for this purpose (Use of Galactogogues in Initiating or Augmenting
Maternal Milk Supply, Appendix C-1). When other strategies to increase milk supply have been applied,
however, pharmacological alternatives may be considered to optimize milk production and support ongoing
breastfeeding.
THE USE OF MOTHER'S MILK IN NEONATAL INTENSIVE CARE

Guidelines for the collection and storage of human milk ensure optimal quality of human milk that is collected
and fed to premature infants. Milk storage and handling are slightly different for the term healthy baby/infant
compared to the hospitalized/sick/premature infant (131, 132). Collection technique, collection container and
storage conditions play a role in maintaining the unique constituents of human milk and reducing bacterial
colonization. Even with meticulous technique, no mother's milk is sterile. Attention to hand washing and
cleaning of expression equipment, however, is extremely important in reducing colonization by pathogens other
than normal skin flora. Mothers require instruction to ensure that anything coming in contact with the milk or
the breasts be cleansed thoroughly prior to each pumping session. Each mother must have her own personal kit.
Instructions should include washing all pump parts with hot soapy water after each pumping and sterilizing all
equipment once a day. Most NICUs provide sterile hard plastic containers for the collection of human milk in the
hospital and at home. These containers provide for the stability of water-soluble constituents and
immunoglobulins. Plastic milk bags are not recommended for the collection of milk for hospitalized infants as a
result of loss of milk constituents and chance of leakage and contamination during storage and handling. The
temperature at which milk is stored determines the duration of storage. Fresh milk, considered optimal for
premature infants, must be used or refrigerated, within one hour of expression. If the infant is not fed
immediately, expressed milk can be safely kept in the refrigerator for 48 hours. If the milk will not be used
within this time frame, it should be frozen. Milk that has been fortified with additives should be used with 24
hours, and should never be frozen (Appendix C-2).
All milk that is used for infant feeding should be stored in the hospital under controlled conditions.
Environmental issues are an important aspect of quality control. This includes monitoring refrigerator/freezer
temperatures, and routine cleaning and maintenance of storage units and milk preparation areas. All milk should
be clearly identified with infant's name, medical record number and date and time of collection and kept in an
environment that eliminates the potential for tampering (Appendix C-2).
Human milk, like blood, is a living fluid and should be handled as such. Proper handwashing and/or wearing
gloves during preparation and administration prevents potential bacterial contamination. Stringent quality
control standards,
P.427
including a two-nurse check system of identification, minimize potential administration errors (Appendix C-2).
The Use of Donor Human Milk

No chapter on the use of human milk in the NICU is complete without a discussion of donor human milk and its
uses. It is beyond the scope of this chapter to present this subject in great depthfor that the reader is referred
to more comprehensive discussions (133,134,135,136).
If human milk is the perfect food nutritionally, immunologically and developmentally for the sick or premature
infant, what happens when that milk is not available? What if the mother died in the postpartum period? What if
the mother, despite her best efforts, is not providing adequate volumes to cover the baby's needs? What if the
mother is HIV positive in North America? What if the mother was delivered prematurely because she had a
diagnosis of breast cancer made during pregnancy, and she must now undergo chemotherapy? What if the
mother, despite being optimally informed, decided not to provide breast milk, and now her baby has a severe
allergic colitis and is not thriving on any of the expensive hypoallergenic formulas? What if? Do these questions
sound far-fetched? They are not. They have all occurred within recent memory in one intensive care unit. The
answer has always been provide formulaor keep changing formulas around until you find one that works.
But is this the best answer? The World Health Organization (WHO) and the United Nations Children's Fund
(UNICEF), in their 2003 Global Strategy for Infant and Young Child Feeding, recommend banked donor milk as
the next option when mother's own milk is not available (137). Although the AAP fell short of endorsing the use
of donor milk in its 1997 statement (1), the AAP's 2003 Red Book does contain a section on human milk banks
and donor milk (138). The international physician breast-feeding organization, the Academy of Breastfeeding
Medicine, supports donor human milk as the preferred human milk supplement in its published clinical protocol
for supplementing healthy, term infants in the hospital (139). Many other countries in Europe and South
America have donor milk banks and make extensive use of donor human milk.
In North America, the Human Milk Banking Association of North America (HMBANA) developed guidelines that
are now mandatory for all member milk banks (140). They require thorough donor screening and testing
(similar to that required for organ donors), pasteurization of all milk, and bacteriological quality control of all
dispensed milk. The pasteurization process kills all known viruses (including CMV and HIV) and bacteria known
to date. Most of the immunologic factors remain intact. IgA is reduced by 20% to 30%; lysozyme activity is only
minimally affected; lactoferrin iron-binding capacity is reduced by as much as 60% (141). Even with these
reductions, donor milk still has more immunoglobulin, lysozyme, and lactoferrin than formula! There is little
effect of pasteurization on the long chain polyunsaturated fatty acids (141). In 2002 the six milk banks
belonging to HMBANA dispensed 501,100 ounces of milk, approximately half of which went to premature or ill
NICU babies (when surveyed in 1994) (142).
The benefits of using donor milk are the same as have been discussed for mother's own milk. In fact, many of
the studies from Europe on the prevention of NEC and neurodevelopmental outcomes were done using donor
milk when own mother's milk was not available. In addition to prematurity and low or nonexistent maternal milk
supply, other uses in the NICU include formula intolerance/allergy; short bowel syndrome; other issues of
malabsorption such as gastroschisis; multiple births; recovery from NEC; supplementation of the hypoglycemic
baby; and supplementation of the dehydrated, hyperbilirubinemic breastfed baby.
Donor milk requires a doctor's prescription. Although the milk is donated to the milk banks and not sold, there
are processing fees charged to help defray the costs of donor screening and processing of the milk. That fee is
currently approximately $3.50 an ounce plus shipping fees. In Canada, the use of donor milk is covered by the
national health plan. In the United States, its coverage by insurance and health plans is variable. It often
requires an extra effort by the physician to speak to the medical director of the health plan and educate him or
her on the benefits of using donor milk. Although the cost may seem exorbitant, it has been estimated that for
every $1 spent on donor milk, $11 to $37 in NICU costs are saved (136). Arnold calculated the cost of supplying
donor human milk and fortifier to a hypothetical preterm infant as being $1350 (134). Using the previously
discussed (86) estimated additional costs of surgical NEC averaging $186,200 more and medical NEC $73,700,
the potential for preventing a case of NEC by the use of donor human milk seems more than economical! Most
milk donated to milk banks comes from mothers who have delivered at term. However, they do receive milk
from preterm mothersmothers who have large milk supplies, and even mothers who have ultimately lost their
baby, but donate their milk as part of working through the grieving process, to help another baby. Preterm and
full term milk is processed separately. Donor preterm milk can be specifically requested, and if available, it will
be shipped to you. It is shipped frozen overnight. It is worth having a policy in place for the timely transport of
the milk from your receiving area to your unit, where it must be transferred to the freezer on arrival. For a
sample donor human milk policy, please see Appendix C-3.
FORTIFICATION OF HUMAN MILK FORVERY LOW BIRTH WEIGHT

BABIES
There are many times when babies in the NICU do not show adequate growth and good weight gain. For babies
who are receiving breast milk, there are a number of things that can be done. One of the most obvious, and
most frequently overlooked, is increasing the volumes. As neonatologists, we have firmly entrenched in our
minds that we cannot exceed
P.428
160 cc/kg/day. For some babies this is truein particular those who are fluid restricted, on diuretics, or for
those who do not tolerate advancing beyond that volume enterally. However as we also know, term babies,
when allowed to feed ad libitum, often exceed 200 to 220 cc/kg/day. So it is worth remembering that at least in
some of our population, we do not have to adhere to such tight fluid control. When we cannot advance volume
any further, there are generally two ways to advance calories with breast milk: using hind milk and using
commercial fortifiers.
Hind Milk
The fat content of human milk increases throughout the course of a feeding or a milk expression session. We
can take advantage of that to provide higher caloric milk to a baby who is either receiving breast milk through
an alternative feeding method or who is breastfeeding. Hind milk refers to the milk at the end of the feed that is
higher in fat content; foremilk is the milk from the early part of the feed that is lower in fat. Because the fat
changes along a continuum, this is in some ways an artificial construct, but one that is useful to this concept.
For a mother to provide hind milk, she must be producing more than the baby's requirement in milk per day.
Essentially it consists of having her pump off a specified volume of milk at the beginning of a session, change
containers, and then continue to pump until she is empty. If these containers are then compared, the foremilk
will look thin and bluish (lower fat content) and the hind milk will look whiter and creamier. The decision
regarding how much is foremilk and how much is hind milk is based on how much milk mother makes, and how
much is needed for a feed. For example, if she generally pumps 4 ounces of milk total, and the baby receives 2
ounces at a feeding, we would have mom pump off approximately 1 ounce from each breast as her foremilk,
and the remaining 2 ounces total as her hind milk. The containers need to remain separate and be labeled as
foremilk and hind milk respectively. If a baby is receiving some or all of his/her nutrition at the breast, this
method can still be used. Mother can express the initial one ounce from her breast, and then let the baby drain
the breast to receive the higher fat content milk. This method has been used successfully in a developing
country to improve growth when commercial fortifiers were not available (143). A word of cautionhind milk is
only higher in fat and therefore calories. It does not provide the extra protein, vitamins or minerals that many of
our babies need. These are provided by the commercial fortifiers. However, if a baby on fortified milk is not
showing acceptable rate of weight gain, one can successfully use fortified hind milk (144).
Commercial Fortifiers
Either liquid or powdered commercially manufactured substances generally provide fortification. Powdered
fortifiers add the additional nutrients without diluting out any of the substances in the human milk. Liquid
fortifiers are added in equal volume to the human milk, thus diluting out its components by half. Traditionally,
when a mother's milk production has equaled or exceeded her baby's requirements, powdered fortifier has been
used. When her production has not met her baby's needs, liquid fortifier has been used in an attempt to
extend her milk. There have recently been a number of reports of significant Enterobacter sakazakii infections
in neonates, including sepsis, meningitis and necrotizing enterocolitis, with premature and sick babies being at
highest risk (145,146,147,148,149,150). They are associated with the use of complete milk-based powdered
infant formula products from a variety of manufacturers. Powdered milk-based infant formulas are not sterile;
they are heat treated during processing, but unlike liquid formula products, not subjected to the high
temperatures for sufficient time to make the product commercially sterile (145,151). Recommendations from
the CDC included that formula products should be selected based on nutritional needs; alternatives to
powdered forms should be chosen when possible (145). At the time of this writing, there have been no reports
of infection following the use of powdered human milk fortifiers. However, based on these reports and the CDC
recommendations for powdered formula products, many NICUs are preferentially using the liquid fortifiers,
unless there is a cogent reason for using the powdered.
The protein in these fortifiers, which is derived from cow's milk, is of concern to many neonatologists and
nutritionists. One of the risks of artificial feedings compared to human milk is the development of allergic GI
symptoms to the bovine protein they contain. The ideal fortification of milk for preterm babies could be obtained
by a process known as lactoengineering, in which specific components are removed from either excess own
mother's milk, or donor milk, and then added to the human milk feeding to increase those needed nutrients.
Studies have shown babies fed own mother's milk fortified with human milk protein (152) or with human milk
protein plus human milk fat (153) grow comparably to both intrauterine growth curves and to other premature
babies who have been exclusively fed fortified preterm formulas. This is an expensive process, but is feasible,
and has been under investigation in several European countries (154,155,156). The ideal may well become
human milk proteins packaged with other components for use as fortifiers for preterm babies. In the meantime,
for these very low birthweight babies, human milk with added fortification is the ideal nutrition available.
It is a widely held belief by staff at all levels in NICUs that the addition of commercial human milk fortifiers
increases feeding intolerance; there is, however, no evidence for this. Numerous controlled clinical trials have
shown no difference in signs of feeding intolerance, increased gastric residuals, bilious gastric residuals,
abdominal distention, or blood in the stools when compared to human milk alone or preterm formulas
(157,158,159,160,161,162).
P.429
There has also been concern that nutritional additives to human milk may in fact alter its complex system of
antiinfective/host defense properties. In one clinical study fortified human milk has been shown to increase, but
not to a significant degree, the incidence of necrotizing enterocolitis (5.8% compared with 2.2%, p = 0.12), and
infections (suspected plus proven; 43% compared with 31%, p = 0.04) over human milk alone (163).
Laboratory research in this area has shown that fortification does not affect total IgA content (164,165), or
inhibition to E. coli, although it does decrease lysozyme activity by 19%, which was not considered significant
(164). When fortified human milk was evaluated under simulated nursery conditions, bacterial colony counts did
not rise appreciably for the first 20 hours under refrigeration, but did increase approximately 10-fold in the
following 4 hours under incubator conditions (165). More recently, Chan evaluated the effect of addition of
powdered fortifiers to human milk on antibacterial activity. Human milk alone inhibited the growth of E. coli,
Staphylococcus aureus, Enterobacter sakazakii, and Group B Streptococcus. Fortifier that contains iron, and iron
alone, affected the antimicrobial activity of the milk, causing no zone of inhibition for any of these organisms
(166). The fortifier containing no iron had similar inhibitory effects to human milk alone. It was postulated that
the iron may have saturated the lactoferrin in the milk, thus decreasing its antibacterial activity. Another in vitro
study looked at the effect of human milk fortifier added to human milk on the concentration of transforming
growth factor-(TGF-. TGF- is a gut peptide found in human milk that is believed to exert a maturational
effect on the neonatal gut. They found that the addition of fortifier did not affect the whole milk or aqueous
portion of TGF-, but significantly decreased its concentration in the fat fraction, in addition to altering its
molecular mass profile characteristics (167). Thus far, the data does not suggest a change in our current
practice of fortification. Whether these laboratory findings are of clinical significance remains to be seen. But it is
worth keeping in mind that the addition of exogenous substances to human milk, although our intent is good,
may in fact be found later to alter and compromise the balance of nutrients, enzymes, hormones, immunological
and other factors and thus their effects.
One final note on fortificationmothers of NICU babies are already dealing with many stresses, not the least of
which are the circumstances that brought her and her baby to the NICU. We must always be very cautious when
talking about inadequate growth and fortification with her. Rememberproviding breast milk is the one thing
only she can do for her baby. If we are careless when we broach these subjects, the message she will take away
is that her milk is not adequate, and that her infant's slow growth relates to something she is not doing right. It
is important to point out to her that her milk is the best possible nutrition for her baby, and helping the baby to
fight infections, and that she is doing an excellent job providing it. The issue at this time is that her small baby
has enormous nutritional needs, which can be met by using hind milk and/or commercial fortifiers. It is
important as well to emphasize that this is a time-limited issue, and that as her baby grows, matures, and
becomes healthier, her milk alone will provide all her baby needs.
Colostrum Oral Care

It is well known that sIgA is present in large quantities in premature colostrum (Table 23-4). Almost 75% of
ingested IgA from human milk survives passage through the intestinal tract and is excreted in the feces (168).
This immunoglobulin provides passive protection against antigens (including viruses and bacteria) crossing the
immature, permeable mucosal barrier (169). As early as 1983, Narayanan and colleagues from India reported
that small quantities of colostrum (10 cc three times a day) administered to low birth weight infants resulted in
a significantly decreased incidence of infection compared to solely formula-fed controls (170). It is common
practice in NICUs to provide oral care to babies who are not being orally fedincluding the new, ventilated
ELBW and VLBW infants, and those with significant cardiovascular instability. This is generally accomplished
using sterile water applied to a cotton-tipped swab, or to a piece of gauze. In our unit, we have instituted a
practice of colostrum oral care in which any baby who is not being enterally fed can be provided with oral care
using mother's expressed colostrum instead of sterile water. The reasoning behind this is that knowing the sIgA
is active in protecting the mucosal surfaces of these immunocompromised babies, by swabbing the oral cavity
with colostrum, we are providing some of that protection to the oral mucosa and possibly the upper GI tract if
any is swallowed. Additionally, it accomplishes another very important purpose. In the early days before a
mother can hold a very small or sick baby, before the milk she is working so hard to express can be used for
feeds, she is playing a vital role in the care of her infant. This procedure can be initiated within hours of birth,
even in the most premature babies. NICU staff or parents dip a sterile swab in the colostrum container, and
swab over the tongue and oral mucosa. The swab should not be dipped back into the container. This may be
done as frequently as every 4 hours, generally in conjunction with hands-on care. Once minimal enteral feedings
are begun, remaining colostrum should be used for those early feedings, and oral care continued according to
standard routine.
THE DEVELOPMENTAL PROGRESSION TOWARD BREASTFEEDING

In hospital breastfeeding management, the process of establishing breastfeeding in the preterm population
begins with the facilitation of frequent skin-to-skin contact
P.430
through Kangaroo (skin to skin) care (Fig. 23-3). Originally proposed by Rey and colleagues in Bogota Columbia
(171), subsequent studies in developed nations illustrate that the practice is safe, beneficial and an essential
component of high-quality neonatal intensive care (172,173,174,175). Kangaroo care has also been consistently
associated with improved milk production, enhanced maternal breastfeeding competence and extended lactation
(175,176,177,178). Mothers who provide skin-to-skin contact report that their infant makes rooting and sucking
movements toward the nipple, note feelings of milk ejection and leaking, and often express larger amounts of
milk immediately following a Kangaroo Care session (179,180). Kangaroo care provides the mother with the
opportunity to become more confident in handling her small infant. Mothers who engage in Kangaroo Care
report a sense of efficacy, increased self-esteem, and confidence in caring for their babies (181). Kangaroo care
also provides the infant with frequent access to the breast, increases opportunity for nonnutritive sucking, and
assists in the gradual transition to breastfeeding. It is the optimal position for an infant to receive gavage
feedings prior to the establishment of breastfeeding and during supplemental feedings. Tornhage and colleagues
report that infant plasma cholecystokinin levels increase when receiving nasogastric feedings during Kangaroo
Care (182). This stimulates GI function and infant growth.
Figure 23-3 Mother and 600-gram 23-week infant engaging in skin-to-skin contact.
The transition from breast milk feedings to breast-feeding is critical. When successful breastfeeding is
established in the hospital setting, breastfeeding is more likely to continue at home. However, the process of
breastfeeding the premature infant can be challenging for the infant, the mother and the health care
professional. The transition phase from providing expressed human milk to nutritive breastfeeding has not been
extensively studied. Factors to consider in transitioning the premature infant from breast milk feedings to
feeding at the breast include assessment of feeding readiness, optimizing opportunity for early breast-feeding,
encouraging increased breastfeeding as the infant nears discharge and ensuring post discharge support based
on the individual needs of the mother and infant.
Assessment of feeding readiness is determined by the maturation of the infant, measured chronologically as
corrected gestational age and influences the development of feeding skills. The bottle-feeding skill of the
premature infant has been positively correlated with the development of sucking skills and is a function of
maturation (183). The infant must be able to coordinate bursts of sucking interspersed with pauses for
breathing to manage the bottle-feeding skills. This ability is variable but often occurs by about 34 weeks'
gestational age.
Restrictions in breastfeeding policies for preterm infants are commonly based on studies of bottle-feeding, in
which it has been established that infants with immature cardi-respiratory control show less coordinated suckswallow-breathe pattern, resulting in apnea, hypoxia, and bradycardia (184). However, at the breast the
preterm infant coordinates sucking, swallowing, and breathing with minimal fluctuations in transcutaneous
oxygen pressure (104,106), and as discussed earlier, is more physiologically stable.
There is no scientific evidence linking gestational age, growth/weight milestones, or the ability to drink from a
bottle as evidence for breastfeeding readiness in a premature infant. Although maturation plays a role in the
premature infant's ability to breastfeed, clinicians observe a wide range of variability related to breastfeeding
readiness and competence among premature infants, with infants as young as 34 weeks gestation fully
breastfeeding although some term infants take several weeks to effectively breast-feed. This suggests that the
emergence of breastfeeding competence in preterm infants is a multifactorial process that is dependent on both
infant and maternal factors. The role of experience and learning in acquiring breastfeeding skills has recently
been investigated (185). Nyquist and colleagues suggest that the development of nutritive sucking is not solely
maturational but a result of learning and extrinsic factors such as maternal-infant interaction and the frequency
and time spent breastfeeding (185). The same investigators examined early oral behavior of preterm infants
during breastfeeding via electromyographic study (186). Data provided evidence of early sucking competence
during breastfeeding, with wide individual variations. The authors concluded that preterm infants are capable of
suckling at the mother's breast at low maturational levels and that both maturation and experience play a role in
breastfeeding success (186). With experience and maturation, preterm infants will demonstrate increasing
competence in latching onto the breast and maintaining a latch. Over time, they will subsequently engage in
more efficient suckling and demonstrate gradual increase in vigor, intake velocity and volume at the breast
(187). Research also indicates that breastfeeding is less stressful for premature infants than bottle-feeding with
less hypoxia, apnea, bradycardia, and oxygen desaturation experienced during breastfeeding than during bottle
feeding (104). Therefore, when infant stability is

P.431
attained, neonatal staff should encourage early and frequent opportunities to breastfeed.
Figure 23-4 Premature infant at the breast.
Readiness criteria for early breastfeeding sessions include individual assessment of physiological stability during
Kangaroo Care or holding, wakeful periods during feeding, rooting when hungry, and periods of nonnutritive
sucking. The goal of early breastfeeding for a premature infant is to promote an enjoyable experience for both
mother and baby although simultaneously teaching proper position and latch at the breast. Some oral feeding is
possible at the breast between 28 to 30 weeks gestation; by 32 to 34 weeks gestation, some infants may be
able to take a full breastfeed once or twice a day, while others may become proficient at breast; from 35 weeks
onward, efficient breastfeeding that maintains growth is possible (Fig. 23-4).
Early breastfeeding sessions provide opportunity to introduce breastfeeding as the first oral feeding experience
for the premature infant. These are in effect, practice sessions. Mothers do not feel the pressure to make it
work because their baby's nutrition does not depend on their success. So they become comfortable handling
the baby, and working on the skills of positioning and latching, while enjoying this process and the time spent
with their baby. These sessions are also excellent times to give the baby a gavage feed, in effect teaching the
baby as well. Over time the baby becomes imprinted with this is the position I am in, this is the taste and
smell of my mother (and maybe even a little taste of the milk that can be expressed onto her nipple!), and this
is the nice feeling I get when my stomach becomes full. During this early breastfeeding stage, mothers should
be encouraged to put baby to the breast one to two times per day. Making the mother and baby as comfortable
as possible is key. Chairs used for these postpartum women should be padded and give the mother good
support with enough room to maneuver and provide support to the infant's body. The large reclining chairs used
in geriatric wards or dialysis units make perfect breastfeeding or skin to skin chairs. Mothers often maintain
better position if a small stool is used under their feet to raise their legs somewhat. Pillows may be necessary to
help support her arms in correct position. The baby needs to be supported as well. Commercial breastfeeding
pillows, made of hospital grade materials that can be wiped off with disinfectant between babies, are available
and work well. They can be covered with a towel or receiving blanket for comfort. Alternatively, some come with
removable, washable covers that can be changed between babies. In our experience they seem to work better
than using bed pillows, which are another option. Neonatal staff should provide a constant presence to reassure
the mother, point out the positive aspects of this early breastfeeding, ensure infant stability, and optimize the
latch of the premature infant at the breast.
The premature infant has unique characteristics that may interfere with latch at the breast. These include low
muscle tone, limited energy, and the propensity to fall asleep at the breast from fatigue rather than satiety
(188). Compared to a full-term infant, physical characteristics include a proportionately larger head, weaker
neck muscles, and a mouth that is smaller in relationship to the areola and breast. These characteristics require
breastfeeding positions that assist in placing the infant's mouth over the areola. Failure to support the infant in
this optimal position will result in inability to compress the lactiferous sinuses, decreased milk transfer and
nipple trauma. Additionally, failure to provide adequate support for the premature infant at the breast can result
in the baby slipping away from a good latch and tiring easily as a result of the additional effort expended during
breastfeeding to try and stay latched. Mothers may become discourage because of some of these issues. It is
important to reassure them that these are time-limited concerns. As the baby grows and develops, the
hypotonia will improve. Early on for many of these dyads, the problem of trying to fit a quarter-sized nipple
into a dime-sized mouth is a real problem. It is easy to point out to a frustrated mom, that although her nipple
is not expected to grow any more, her baby's mouth will!
Effective techniques to assist the premature infant to latch vary and depend on the configuration of the breast
and the strength and skill of the infant (188). Data from nonnutritive sucking and bottle-feeding reveal that the
amount of suction that the premature infant generates
P.432
while at the breast is maturationally dependent (189). Because of this limitation, small premature infants need
to be placed and kept on the nipple/areola, because limited suction inhibits their ability to bring and keep the
nipple /areola in the correct position for milk extraction. Mothers of small premature infants require instruction
to place and support the baby at the breast. This includes supporting the baby's head at the nape of the neck
during latch and throughout breastfeeding. Additional head support assists weak neck musculature of the
premature infant to maintain neck stability and prevents undirected head movement that can lead to airway
collapse, apnea, and bradycardia (190). For some mothers with large breasts, it is helpful to roll a towel or
blanket to place under the breast to hold it up, to make it easier for mom to see as she is placing the baby, and
easier for the baby to stay latched because the breast is being supported. Supportive positions for breastfeeding
the premature infant include the underarm/football hold or cross-cradle (across the lap) positions. These
positions support the infant's head and torso, guide the infant to the nipple/areolar area, ensure optimal latch,
and subsequently maintain the baby in a close flexed position throughout the feeding. To maintain the airway
and to help in swallowing in the cross-cradle position, it is important to maintain the infant's body facing the
mother, keeping the baby's ear, shoulder and hip in a straight line, and then flexing their legs around the
mother's body (Fig. 23-5). In the football hold, care should be taken to ensure that the baby's neck is not overly
flexed (Fig. 23-6). The typical breastfeeding Madonna hold, which most mothers have seen so will naturally try,
in which the baby is held with her arm on the same side of her body as the breast she is feeding from, is not
appropriate for premature or other low tone babies. The baby's head, which is supported in the crook of the
mother's arm, tends to fall through, and is poorly supported for optimal feeding.
Figure 23-5 Cross-cradle position at the breast.
When optimal position is attained, the infant's head can be moved toward the breast, by the hand supporting
the neck and head. The mother's nipple can then be lightly brushed against the infant's mouth. The infant will
spontaneously open his mouth when ready. The mother should then gently guide the infant onto the nipple/
areolar area. The mother with everted, elongated nipples may find it easier to elicit a sucking response at the
breast as her nipple will automatically stimulate the roof of the baby's mouth and elicit the sucking response.
Mothers with flatter nipples may require additional assistance to achieve latch. This may include manipulating
the nipple to make a teat which will then stimulate the infant suck reflex. It can be done by either having the
mother use the electric pump briefly right before the feed to draw the nipple out, or by having the mother roll
her nipple between her thumb and forefinger. This requires additional assistance from nursing staff or a
lactation consultant during the first several breastfeeding sessions or until the infant has demonstrated effective
latch.
Figure 23-6 Football hold with a premature baby at breast, demonstrating compression of the breast to
increase milk flow during the feed.
It is important here to discuss the concept of on-cue feedings. In neonatal units, we are very entrenched in
calculating, ordering, measuring and reporting the exact input and the intervals at which it must occur. It is
critical to our care when our patients are very small or very sick. However, we need to learn to relax that
somewhat as we begin to introduce feedings at breast. If the baby is due for an interval feeding at a specific
time but is asleep, it often will prove fruitless to try to put the baby to breast at that time. However, an hour
later, before the next feed is due, when the baby is awake and the mother is at the bedside, it is the right time
to allow them to practice breastfeeding. If the baby takes nothing or if the baby actually ingests a few ccs, it
really does not matter. But the dyad have had the perfect opportunity to learnthat is what matters. As the
baby
P.433
matures, bedside staff will often notice a pattern to their wakeful periods. It is not unusual for them to be at
night. So a mother who comes in every afternoon to put the sleeping baby to breast may have gotten nowhere.
But if she changes her pattern, and manages to come in at night when the baby is often active, their
breastfeeding skills may take off and amaze everyone. From there, we often see the babies becoming more
wakeful for feeds, so on-cue breastfeeding times can advance in number. If the baby is awake 2 hours after the
last feed, and is giving feeding cues, it is perfectly reasonable to have her put the baby to breast again, even
though it is too early based on our feeding orders. That is what the mother will do once she is homethat is
what we must ultimately aim for before discharge.
Mothers must be taught indicators of milk transfer/ swallowing at the breast. This includes the soft caw sound
of swallowing. This also includes the open-pause-close pattern of suckling during which the wide-open pause of
the infant's mouth during breastfeeding indicates that the infant is swallowing a mouthful of milk. When
evidence of milk transfer diminishes during a feed, switching breasts and using breast compression may
increase volume of milk ingested. During breast compression, the mother uses her free hand, which has been
supporting her breast, to grasp the breast in a C-hold (fingers on one side, thumb on the other) and when the
baby is suckling, compresses her breast between her fingers to eject more milk (Fig. 23-6). This will often keep
the baby interested and actively suckling. When bursts of these indicators are evident during breast-feeding,
and accuracy of intake is important, intake at the breast can be estimated by the use of prefeeding and
postfeeding weights (191,192). The difference between pre- and post-weights in grams equals the amount of
milk transfered in cc's. Volume of intake at the breast can be extremely variable. It is, however, virtually
impossible to estimate by observing the feed, even by experts in lactation. Most premature infants will
experience marginal intake during early breastfeeding. However some infants can consume adequate quantities
of milk during early breastfeeding because mothers have copious milk supply and it flows readily. Thus milk
volume and ejection can compensate for a marginally effective suck in some small premature infants. Test
weighing or prefeeding and postfeeding weights under identical conditions may be helpful in determining the
ability of the infant to stimulate milk transfer, amount of supplements required, and maternal milk supply. The
scale used must be an accurate electronic scale. Under identical conditions means the baby must be weighed
immediately before and after the feeding session, with the same clothing, diaper, blankets, leads, tubes, and so
on at both of these times. Using this data, instead of trying to guess how much supplementation must be given
for that feed, the difference between the ordered feeding amount and the amount taken at breast can be
calculated and given via gavage. Prefeeding and postfeeding weights may also provide an indicator of
breastfeeding progress and therefore reassurance for mother and staff. Test weighing should be introduced
when it appears that milk transfer is occurring or when discharge is imminent. This information can assist in
tailoring ongoing breastfeeding support and discharge preparation.
ONGOING BREASTFEEDING SESSIONS

The Use of Alternative Feeding Methods
The goal of ongoing breastfeeding sessions is to support the infant to consume adequate volumes of milk from
the breast in anticipation of discharge. Research indicates that premature infants are able to coordinate the
suck, swallow and breathe pattern of breastfeeding before bottle-feeding (104). This data also illustrates that
mean volumes of milk transfer/intake during early breastfeeding sessions is often minimal but that the range for
intake at the breast is variable and may be quite high. Early breastfeeding sessions are almost always
supplemented by gavage feedings, which optimize this early learning experience. Typically, other methods of
supplemental and/or complemental feedings are initiated when the maturity, energy and ability of the infant to
suckle increases. These include bottle-feeding, cup feeding, finger feeding, and feeding tube at the breast.
Currently there is limited evidence to guide hospital practice related to transitioning the premature infant to the
breast. Although protocols to eliminate or minimize bottle feeds have been published and used clinically (193),
the majority have not been tested in randomized controlled trials. Hospital policies and practices related to the
use of supplemental or complemental feedings however can potentially interfere with breastfeeding outcomes
(194,195).
The majority of neonatal units introduce bottle-feeding as a supplementation method for the preterm infant who
is not exclusively breastfed during neonatal hospitalization. The degree that bottle-feeding may interfere with
the acquisition of breastfeeding skills is controversial. Most experts would agree that the oral dynamics of
breastfeeding and bottle-feeding are significantly different (196,197). Breast-feeding is dependent on the
massaging action of the tongue and jaw while bottle-feeding depends more on suction. Additionally, the
characteristic of milk flow from the bottle is different from the flow of milk from the breast. A quicker flow that
provides immediate reward for minimal sucking may be another factor in the preference of bottle over
breastfeeding. Therefore the suckling pattern learned through bottle-feeding may result in infants resisting
latching to the breast, refusing the breast, or having difficulty with latch-on (198,199). Research indicates that
the use of artificial nipples interferes with breastfeeding success in term infants and that they may be
detrimental in preterm infants as well (194,200,201). Consequently, for preterm infants whose supplementary
feedings are provided by bottle, subsequent breastfeeding may be difficult. Alternative feeding methods that do
not adversely affect breastfeeding duration in this population are highly desirable. This need, coupled with
recommendations emerging from the 1991
P.434
WHO/UNICEF Baby Friendly Initiative (202,203) that no artificial nipples or pacifiers be used if breastfeeding is
to be established, has resulted in an increased use of alternate feeding methods in neonatal units. Methods that
may foster the transition of preterm infants to full breastfeeding include continued tube feeding, cup feeding,
finger feeding, or a nursing supplementer tube (feeding tube device). A nipple shield may also be a
consideration at this stage of breastfeeding. Currently there is no agreement on the best method of
supplementing the preterm infant or the timing of transitioning the preterm infant to the breast.
Supplementation method and timing are based on a combination of maternal and infant factors. Various levels
of evidence support each strategy.

Until the preterm infant establishes a pattern of nutritive oral feeding, periodic gastric tube feedings are
required. Although guidelines and protocols for gastric feeding differ, continuous nasogastric tube placement is
the most common protocol and tube sizes include numbers 5, 6.5, and 8 French (204). Nasogastric feedings
have been associated with negative impacts on breathing, sucking, irritation, and delayed transition to oral
feeding. A prospective randomized trial evaluated the impact of an indwelling smaller bore feeding tube (3.4
French) during the transition to breastfeeding as an alternative to supplementation by bottle. Infants who
received nasogastric supplementation were 4.5 times more likely to be partially and 9.4 times fully
breastfeeding at discharge compared to the bottle- fed group (205). Method of supplementation continued to be
predictive of long-term breastfeeding as well (205). Although there were risk and stressors associated with
continued nasogastric feeding (greater numbers of apnea and bradycardia requiring stimulation) the authors
suggest that this method can be used safely during hospitalization to support ongoing breastfeeding.
Cup feeding has been used primarily in full term infants as a short-term method of providing a supplement to a
breastfeeding baby. Early reports suggested that cup feeding is a safe alternative method of supplementation
for preterm infants 30 weeks gestational age and older (206). Breastfeeding rates were increased when cup
feeding replaced bottle-feeding and replaced or supplemented tube feedings in this early study (206). More
recently, the safety of cup feedings and bottle feedings were compared in a prospective randomized cross-over
study of 56 infants 34 weeks at birth by Marinelli and colleagues (207). Results indicate that a 10-fold increase
in desaturations occurs during bottle-feeding compared to cup feeding. Although volumes taken were lower and
duration of feeds longer during cup feedings, the study supported the use of cup feeding as a safe alternative
feeding method for premature infants learning to breast-feed (207). The mechanics and safety of cup feeding
were also evaluated by Dowling and colleagues through observation of 18 cup feeding sessions by 8 preterm
infants with a mean gestational age of 30.7 weeks (208). Data indicate that the infants remained physiologically
stable during cup feeding but that low volume of intake may account for this stability and is also cause for
concern (208). Spillage was a common occurrence during cup feeding, indicating that cup feeding may not be as
easy for the infant or caregiver as previously described. Therefore, the efficacy and efficiency of cup feeding as
an alternative method of supplementation is questioned and requires further study (208) (Fig. 23-7).
Figure 23-7 Cup feeding.
Finger feeding is another alternative feeding method that eliminates the need for bottle-feedings while the
premature infant is learning to breastfeed and/or nearing discharge. Finger feeding is proposed as a method of
feeding that assists with the development of appropriate tongue position and movements of sucking. Finger
feeding is considered more similar to breastfeeding that bottle-feeding and can be used while the infant is
learning to breastfeed, is too tired to breastfeed or when mother is not available (194). Little data exists to
evaluate finger feeding as an alternative feeding method. Oddy and Glenn assessed the effectiveness of finger
feedings in encouraging a breastfeeding-type suck and breastfeeding outcomes in preterm infants who required
supplementation (209). Data was collected on rates of breastfeeding at discharge before and after this
alternative feeding method was introduced. Results indicated higher rates of breastfeeding on discharge in
preterm infants who were supplemented with finger feeding vs. bottle-feeding (209). One of the concerns with
finger feeding is that it is introducing a hard surface into the baby's mouth, and may not ultimately fair any
better than the use of artificial bottles with nipples. It is a tool that is used widely and successfully in sucktraining babies with dysfunctional sucking mechanisms (210). Although further research is required to evaluate
this strategy, finger feeding may provide an option to support breastfeeding learning in this population.
Alternative methods of transitioning the premature infant to the breast can be used when the mother is present;
these include a nipple shield and a feeding tube device at the breast (supplemental nurser). Recent literature
supports the use of a thin silicone nipple shield to assist the premature infant to latch when the infant cannot
draw in enough nipple/areola, the areola is too puffy or the nipple
P.435
is very large (211,212). Because the nipple shield is less pliable than the maternal nipple, it can stimulate a
stronger sucking reflex. It has been hypothesized that the nipple shield functions to increase the effectiveness of
infant suck by remaining in the correct position within the infant's mouth in the absence of strong suck
pressures. A nipple shield maintains its shape when the infant pauses, keeping him/her on the breast with little
effort, minimizing the tendency to slip off during feedings. Additionally, once the baby begins to suckle, negative
pressure appears to be generated in the empty space between the nipple and the tip of the shield. These
pressures have been postulated to compensate for the weaker suck of the infant, allowing milk to accumulate in
the chamber, and thus making it more readily available (212). In these circumstances, the nipple shield may
make it easier for the premature infant to maintain attachment to the breast and extract milk. As a result,
temporary use of a nipple shield may increase both the duration of suckling and volume of milk consumed
during breastfeeding (211,212) (Fig. 23-8). In the past, nipple shields were considered taboo in the lactation
field. They were made of thick rubber, and mothers often developed decreasing milk supplies as a result of less
sensory stimulation to the breast, ultimately leading to cessation of breastfeeding. The newer thin silicone nipple
shields are less likely to precipitate these negative outcomes. Premature infants who may benefit from the use
of a thin silicone nipple shield include those with short inefficient bursts of sucking, limited energy and low
suction at the breast, who fall asleep quickly at the breast, or whose mothers have flat or ill-defined nipples that
make grasp and sustained latch difficult. A nipple shield is often well accepted by mothers because it allows the
infant to feed at the breast with increased vigor, alert periods and intake at the breast (112,212).
Figure 23-8 Silicone nipple shield with feeding tube device.
Figure 23-9 Feeding tube device at the breast.
Attempts should be made to wean the baby from the nipple shield prior to discharge. However, if the baby can
only maintain total enteral intake with the shield, discharge home can be considered, but only if there is
experienced lactation support available. The dyad must have close follow-up with intent to wean from the shield
when the baby is able to suckle well without it. Additional research is required to continue to explore the use of
a silicone nipple shield as a temporary milk transfer device for premature infants.
A feeding tube device at the breast (supplemental nurser) can be beneficial for a mother with a limited milk
supply or for an infant who achieves a good latch but is unable to transfer adequate milk volumes (214). This
device reinforces the position and latch of breastfeeding, provides for additional infant intake at the breast
without supplementary time and energy expenditure, and increases maternal milk supply through optimal breast
stimulation (Fig. 23-9). Mothers often feel initially awkward with placing the tube and then latching the baby,
but with assistance from staff and a little practice, it will become much easier for her. The tube should be placed
so that the tip is close to, but not at or past the end of the nipple. If that occurs, babies often figure out quickly
how to use it like a straw instead of latching correctly! A feeding tube device can be placed inside a nipple shield
to aid a baby with milk transfer, especially if the mother has low milk supply, or the baby has limited energy and
needs help with efficiently transferring milk (Fig. 23-8).
It is important to recognize that the safety, efficacy and breastfeeding outcomes of all methods of
supplementing
P.436
the premature infant during the transition to full breast-feeding have not been studied systematically in
controlled trials. Therefore, the method, timing and duration of supplementation should be based on individual
assessment of the infant and mother and follow informed consent.
DISCHARGE PLANNING
Transition to cue-based full breastfeeding occurs over time with the majority of premature infants able to
achieve this milestone by the time they are equivalent to term. This process is facilitated by increasing the
frequency of breast-feeding as the infant's ability to transfer milk at the breast improves. Several protocols to
facilitate this outcome are documented in the literature. All require ongoing assessment of breastfeeding
adequacy and intervention as required.
Breastfeeding assessment should include regular visual assessment of the baby at the breast by staff with
knowledge and experience in the care of breastfeeding families. Several breastfeeding tools are available to
promote an objective assessment of latch, milk transfer and other attributes associated with optimal
breastfeeding (215,216,217,218,219,220). The Preterm Infant Breastfeeding Behavior Scale specifically
measures the maturational steps in breastfeeding progress and is useful in the clinical setting (220). Utilization
of an objective breastfeeding assessment tool has the potential to identify early breastfeeding issues and assist
staff to intervene in a timely and consistent manner. These tools are also excellent teaching tools for families
which assist mothers in assessing progress toward successful breastfeeding.
As the infant matures, mothers should be given opportunity to breastfeed not only on cue, but ad libitum.
Frequent on-demand or cue-based breastfeeding ad libitum provides information about the infant's energy level
to engage in full breastfeeding, rate of growth with this plan, and assists with setting realistic goals for discharge
preparation and feeding at home. Feeding options should be discussed with mothers and fathers to ensure they
participate in and are fully informed about individualized breastfeeding plans. An individualized transition to
home plan should be put into place for each mother-infant dyad based on the skills of the infant, the mother's
milk production and the infant's caloric needs. Because the behavioral feeding cues of the preterm infant are
less distinct than the full-term infant, mothers require time and learning to interpret correctly (185). The
availability of a parent suite for overnight stays facilitates this learning, allowing maximal access to the infant
and a trial of breastfeeding on cue. This opportunity can also provide valuable information related to maternal
milk supply and the adequacy of infant intake at the breast. Prior to discharge, if the facilities are available,
parents should be offered and encouraged to room-in with their baby for at least 24 hours. This is helpful in
both trouble shooting any concerns or issues that arise, and for building parental confidence that they will be
able to manage once home.
Many premature infants are otherwise ready for discharge before reaching term. Lack of maturity may continue
to interfere with their ability to fully breastfeed. Other premature infants are discharged at term but have size
and energy limitations that preclude full breastfeeding. Mothers therefore may need to modify their expectations
about their infant's breastfeeding potential at discharge. All mothers require a detailed feeding plan at discharge
that addresses the ongoing nutritional needs of the infant while supporting ongoing breastfeeding. This plan
must support the mother's knowledge of expected feeding patterns, assessment of adequate intake,
supplementation (method, frequency and amount) to ensure adequate intake, and the need for continued
pumping to ensure full milk production until the baby is fully breastfed. It is optimal that this plan be not only
developed collaboratively with the parents, but also given to them in written format to refer to once home. It
must also be communicated to the primary care provider and the lactation expert who will be following and
managing this dyad closely at home (Appendix C-4).
SUPPORTING BREASTFEEDING IN THE NEONATAL INTENSIVE

CARE UNIT AND AFTER DISCHARGE
It is critical to have a breastfeeding friendly NICU to optimize a dyad's chances of successfully breastfeeding.
One of the largest roadblocks to this is the inaccurate and inconsistent advice and assistance that is given to
these mothers and families by the health care professionals (relatives and friends as well) they come into
contact with. Much of that is based on knowledge deficits in lactation, attitudes and beliefs that influence what
we say and how we present it, and personal factors such as age, education and own breastfeeding experiences
(221,222). Additionally, nurses, who spend more time at the bedside than any other caregivers, are viewed by
families as breastfeeding authorities, regardless of their knowledge or expertise. Although a number of studies
have looked at the effect of education on caregiver attitudes on breastfeeding, one has examined this approach
in the NICU. Siddell, Marinelli, and colleagues examined the change in breastfeeding knowledge and attitudes in
their NICU staff after a structured mandatory 8-hour educational intervention (223). Nurses' breastfeeding
knowledge and some attitudes improved significantly, with indication that other attitudes were changed as well.
Although more study is necessary, it appears that the time and effort placed on educating a NICU staff so that
advice and assistance can become more consistent, has the potential to make a difference.
The environment also must be breastfeeding friendly. Many mothers are very shy about breastfeeding openly,
and their privacy must be respected. Because milk volume is often increased when a mother pumps at her
baby's bedside, pumping is another activity that requires some
P.437
privacy. Portable screens in sufficient numbers are an easy way to cordon off a small area. It is interesting to
observe that once pumping and breastfeeding become the cultural norm in a given unit, it is not unusual to see
mothers request the screens less. Having enough portable hospital-grade electric breast pumps, comfortable
chairs for skin to skin and breastfeeding, breastfeeding pillows, and stools are equally as important. If space
permits, small rooms in which mothers can rest between feeding sessions, express milk, or take their babies to
as discharge nears are ideal, as are rooms for overnight/rooming-in.
Another major issue is the influence of the formula/ drug companies on our NICU culture, and that of medicine
as a whole. The giving out of free formula samples and discharge gift packs have been associated with
decreased duration of breastfeeding in the less vulnerable well-baby population (224,225). The Baby-Friendly
Hospital Initiative was designed by the United Nations Children's Fund (UNICEF) and the WHO as an intervention
to increase breastfeeding rates (202,203). One of the issues it addresses is the negative influence of these
products, and other informational materials provided by these companies for our patientsfor example booklets
on breastfeeding. Two recent U.S. studies, one in a NICU, have shown that these interventions can improve
breastfeeding initiation rates (226,227). Instead of discharging a vulnerable breastfeeding NICU family with a
discharge gift pack from a formula company, some units have put together their own packs with noncommercial
information and supplies to support them.
The period immediately following discharge from the neonatal unit is a time of extreme vulnerability for mothers
and may precipitate a breastfeeding crisis (188). Follow up care in the community is critical and must be
arranged and provided prior to discharge home. Access to a telephone support number and referral to a
lactation consultant or a breastfeeding clinic has been shown to increase the transition to home and support
ongoing breastfeeding in this vulnerable group (Appendix C-4).
SPECIAL CIRCUMSTANCES
The Near-Term Infant
Infants born at 35 to 37 weeks gestation are considered near term. In the past, these small but physically stable
infants were cared for in a special care nursery until close to 40 weeks gestation. Currently, most stable nearterm infants are placed on the postpartum unit with their mothers and treated as though they are full-term.
Many of these infants appear deceptively vigorous at first glance but have subtle immaturities that may
compromise their outcomes. Clinical challenges in temperature control and glucose/ metabolic stability are
examples of this immaturity. Potential difficulties in establishing and maintaining breast feeding present an
additional challenge in this population.
Common breastfeeding problems experienced by term infants and their mothers are magnified in infants who
are born before term. These include the ability to latch to the breast, the incidence of jaundice, weight loss in
the first few days to weeks of life, and the establishment of maternal milk supply. As a result, the near term
infant is at risk of hypoglycemia, dehydration and slow growth/failure to thrive. Recent literature has reported
an increased incidence of kernicterus among breastfed infants; this literature identifies breastfed infants born at
less than 37 weeks gestation at increased risk for this devastating and preventable morbidity (228,229). These
issues are more likely to occur in near term infants who are breastfeeding.
The predominant characteristic of the near term infant that interferes with breastfeeding is decreased stamina.
This characteristic results in less effective suckling at the breast, decreased milk transfer and suboptimal breast
stimulation. Additionally, the suck, swallow, breathe cycle of the near term infant may not be fully developed,
further compromising intake at the breast. Poor muscle tone also contributes to fatigue and suboptimal
breastfeeding. A common pattern among these infants is to latch and suckle for a short time and then pause to
rest with subsequent difficulty resuming a nutritive sucking pattern. The inability of these infants to sustain a
suck/swallow/breathe pattern limits milk transfer and contributes to insufficient intake at the breast. This
scenario predisposes the near term infant to morbidity in the early newborn period (230).
The two basic principles in breastfeeding support of the near term infant are to ensure adequate infant nutrition
and to assist with establishing and maintaining maternal milk supply. The near term infant requires adequate
nutrition with minimal calorie/energy expenditure. Early frequent breastfeeding should be accompanied by
assessment of latch and intake at the breast. Once lactogenesis II is established, measurement of intake at the
breast with an electronic scale can provide information about the infant's ability to transfer milk effectively and
maternal milk supply. The infant may need to be supplemented after breast-feeding with small quantities of
expressed breast milk or formula. A full/generous milk supply assists the infant to receive adequate intake at
the breast with minimal effort. Because the near term infant may not have the ability to optimally stimulate milk
production, mothers should be encouraged to use a full-sized electric breast pump after breastfeeding to ensure
early and ongoing milk supply. Discharge from the hospital must be accompanied by a discharge plan that is
communicated to the family and care providers. Early and ongoing follow up is required to ensure infant health
and nutrition and maternal health and milk supply. With continuing support, the majority of near term infants
will transition to full breastfeeding at term.
Other infants who have decreased energy and potential for poor breastfeeding may also benefit from this
approach. The NICU graduate may be discharged before term or at term but not displaying term characteristics.
Infants born to diabetic mothers, term infants who have been ill at
P.438
birth, and term infants with physical or neurological issues that interfere with energy or feeding may have
compromised breastfeeding ability similar to the near term infant. Similar to the near term infant, these
vulnerable infants should be carefully assessed for breastfeeding competency and supported until the transition
to full breastfeeding is made. For more information see the Academy of Breastfeeding Medicine Protocol #10.
Breastfeeding the Near-Term Infant at http://www.bfmed.org.
Hypoglycemia and the Breastfeeding Infant

It is important to briefly touch on the issue of hypoglycemia and the breastfeeding infant, from the standpoint of
prevention and maintaining the breastfeeding relationship. The topics of metabolic adaptation to extra-uterine
life, glucose homeostasis, and causes of hypoglycemia are well covered in other venues (231,233). As described
by Williams (231) there are two principles to guide us in preventing hypoglycemia: (a) early and exclusive
breastfeeding meets the nutritional requirements of healthy, full-term newborn infants; and (b) healthy fullterm infants do not develop symptomatic hypoglycemia simply as a result of underfeeding. An excellent guide to
the prevention and management of hypoglycemia is the Academy of Breastfeeding Medicine Clinical Protocol
Number 1: Guidelines for Glucose Monitoring and Treatment of Hypoglycemia in the Breastfed Neonate (233). In
brief, this is put into practice by putting babies skin to skin, which facilitates normal temperature and metabolic
adaptation, to initiate breastfeeding within 30 to 60 minutes of delivery, and allowing them to room-in, and feed
on cue as frequently as every 10 to 12 times in 24 hours or more. Given these tenants, there is no need to
routinely monitor blood glucose levels in healthy, full-term breastfeeding infants (1,233). If hypoglycemia does
develop, asymptomatic babies can continue to breastfeed frequently with careful follow-up and monitoring. If a
need for intravenous glucose develops, once the baby is stable and able to feed, breastfeeding should resume as
weaning from the i.v. glucose occurs.
Jaundice and the Breastfeeding Infant

It is beyond the scope of this chapter to cover this topic in detail. The reader is referred to in-depth articles on
jaundice (234,235) that cover the pathophysiology and prevention of lack of breastfeeding jaundice. With
increasing cases of kernicterus being reported, the importance of good management and frequent follow-up of
breastfeeding babies, especially those who are feeding poorly, whose mothers have not developed a good milk
supply, or who develop jaundice, cannot be over-stated. It is critical that when a breastfed baby needs
intervention or is admitted to the hospital for treatment of dehydration or jaundice, that it be emphasized to the
mother and the family that breastfeeding is not harmful or dangerous to the baby. This can not be emphasized
enough. Mothers immediately get the message that they have done something wrong to cause their child to
have this problem. As often health care providers, in their explanations of the pathophysiology, lay blame on the
failure of breastfeeding, mothers immediately assume enormous guilt. With that comes pressure from within
themselves or their families to stop breastfeeding and begin formula feeding. It is critical that these mothers be
supported to continue breastfeeding through this family crisis. That support in the hospital setting includes
continuing breast-feeding; having a qualified person (lactation consultant, trained staff member) immediately
evaluate and begin working on fixing the mother's milk supply and the baby's latch and transfer of milk, and
allowing the mother to room-in with her baby. If supplements are needed, and the baby can latch, consider use
of a feeding tube device at breast, so the mother will continue to stimulate her breasts to produce more milk,
and her feedings can continue at breast. If milk supply is at issue, evaluate possible causes. Most commonly one
finds a baby with a poor or inadequate latch; a mother with sore nipples; and a vicious cycle of less stimulation
to the breasts resulting in inadequate milk supply. If the mother's nipples are significantly painful and
traumatized because of poor latch, she may need a rest from breastfeeding. In either of these scenarios, it is
critical that she be given the use of a hospital-grade electric pump in the hospital to use, and that she initiates
frequent expression (see Table 23-5) to empty her breasts and increase her milk supply. The gentle support of
this mother and her family are as important as the diagnosis and treatment of the dehydration and the jaundice.
CONTRAINDICATIONS TO BREAST FEEDING

There are surprisingly few contraindications to the provision of breast milk and breastfeeding. We will cover a
few of the most important here (for a more in-depth review the reader is referred to 236,237,238,239).
Infections
Human Immunodeficiency Virus
The one clear infectious contraindication to breastfeeding, in developed countries, is maternal HIV (138,240).
HIV can be transmitted through human milk, with rates varying from 5% to 20% depending on many
contributing factors, and up to 29% when the infection is acquired just before or during the breastfeeding period
(236). In developed countries in which sanitary affordable replacement feedings are readily available, women
should be counseled strongly not to breastfeed. In developing countries in which safe and affordable substitutes
are not readily available, and in which malnutrition and infectious diseases cause high infant and child mortality,
breastfeeding remains the nutrition of choice. These recommendations will continue to be re-evaluated as more
study is done on
P.439
the specifics of the transmission of HIV and the possible affects of maternal and infant antiretroviral therapy.
Human T-Cell Lymphotrophic Virus

HTLV-1 and HTLV-2 are retroviruses similar to HIV. The virus is transmitted in human milk, so women with this
infection are advised not to breastfeed. Recommendations in developed countries are the same as for HIV (138).
Hepatitis
The risk of transmission of Hepatitis A is rare, so breast-feeding is encouraged. Breastfeeding does not increase
the risk of transmission of Hepatitis B. Babies born to mothers who are positive for hepatitis B should be treated
according to recommended guidelines. There is no need to delay initiation of breastfeeding. Hepatitis C
transmission via breast milk has not been documented. Mothers should be advised it is theoretically possible,
but maternal hepatitis C is not a contraindication to breastfeeding (138).
Cytomegalovirus
Cytomegalovirus (CMV) maternal infection is a bit more complicated. It is a ubiquitous virus, and is transmitted
via breast milk. Acquisition in healthy full-term infants does not result in clinical disease, leading to the term
natural vaccination (241). Recent reports of transmission, thought to be through breast milk to premature
infants, some of whom developed significant disease, have caused some concern (242). Some units use only
frozen milk, which has been reported to kill the virus, in the smallest, most immunosuppressed babies. No
authoritative guidelines have been issued.
MEDICATIONS AND DRUGS

The vast majority of medications are compatible with lactation. Many health care providers will tell a mother to
stop breastfeeding when a medication is prescribed. This is rarely indicated, and it is always worth looking up
the drug in a reliable source. If a drug is found to be contraindicated, it is almost always possible to find another
that will be efficacious and is compatible with breastfeeding. Drugs that require cautious use when used in
breastfeeding mothers of newborns or preterm infants include: sulfonamides (jaundice possible; do not use in
G6PD deficiency); ergotamine and bromocriptine (reduce milk supply); pseudoephedrine and estrogens (may
inhibit milk supply); meperidine (neonatal sedation reported); fluoxetine (tremulousness, colic, crying and
hypotonia reported); angiotensin-converting enzyme inhibitors (may predispose to neonatal hypotension) and
anticancer agents (239). Drugs of abusemarijuana, cocaine, heroin, phencyclidineare contraindicated in
lactation. When in doubtdo not tell a mother to pump and dump or discontinue breastfeedinggo to a good
source and find out! Medications that are found in breast milk are shown in Appendix H-1a, H-1b, H-1c, H-1d
and H-1e.
Resources for Medication Use During Lactation

Publications
American Academy of Pediatrics Committee on Drugs. The transfer of drugs and other chemicals into
human milk. Pediatrics 2001; 108:776-789.
Briggs GG, Freeman RK, Yaffe SJ. Drugs in Pregnancy and Lactation: A Reference Guide to Fetal and
Neonatal Risk. 6th edition. Philadelphia: Lippincott Williams and Wilkins, 2002.
Hale, Thomas W. Medications and Mother's Milk. 11th edition. Amarillo, TX: Pharmasoft Medical Publishing,
2004.
Hale, Thomas W. Clinical Therapy in Breastfeeding Patients. 1st edition. Amarillo, TX: Pharmasoft Medical
Publishing, 1999.
Lawrence, Ruth A and Lawrence Robert M. Breastfeeding: A Guide for the Medical Profession. 6th edition.
St. Louis: Mosby, 2005.
Telephone Consultation
Lactation Study Center, University of Rochester: 1-585-275-0088
Poison and Drug Control Centers: (Check your local poison control center for availability of drug
information)
Internet
Breastfeeding PharmacologyDr. Thomas Hale http://neonatal.ttuhsc.edu/lact/index.html
CONCLUSION
In today's high-tech world of neonatal intensive care, there is one relatively low tech intervention that we as
neonatologists can give with just a little effort, understanding, and educationthe gift of mother's milk. The
research supports it. The cost/benefit ratio is very appealing. The advantages to our patients and their mothers
and families are high. The drawbacks are very low. Promoting, encouraging, and supporting mothers to succeed
is a very small price to pay. We just need to remember, and to embrace, that all babies were born to be
breastfed.
Acknowledgment
With deep gratitude to Ms. Linda Kaczmarczyk, Pediatric Clinical Librarian at Connecticut Children's Medical
Center, for her invaluable aid in preparing this chapter.
REFERENCES
1. American Academy of Pediatrics, Work Group on Breastfeeding. Breastfeeding and the use of human milk.
Pediatrics 1997;100: 1035-1039.
P.440
2. Breastfeeding: maternal and infant aspects. ACOG Educational Bulletin 2000;258:1-16.
3. American Academy of Family Physicians. Breastfeeding (Position Paper). AAFP Policies on Health Issues.
Available at http://www.aafp.org/x6633.xml. Last accessed 1/26/05.
4. Position of the American Dietetic Association: promotion of breastfeeding. J Am Diet Assoc 1997;97(6):662666.
5. Nutrition Committee, Canadian Paediatric Society (CPS). Nutrient needs and feeding of premature infants.
CMAJ 1991;144(11):1451-1454. Available at http://www.cps.ca/english/statements/N/n95-01.htm. Last
accessed 1/26/05.
6. U.S. Department of Health and Human Services. HHS Blueprint for Action on Breastfeeding. Washington: U.
S. Department of Health and Human Services; Office of Woman's Health; 2000.
7. United States Breastfeeding Committee. Breastfeeding in the United States: a national agenda. Rockville,
MD: US Department of Health and Human Services, Heath Resources and Services Administration, Maternal
and Child Health Bureau; 2001.
8. U.S. Department of Health and Human Services. Healthy People 2010: conference Editionvolumes I and
II. Washington: Department of Health and Human Services, Public Health Service, Office for the Assistant
Secretary for Health; 2000(Jan): 47-48. Available at http://www.healthypeople.gov/document/ Last accessed
1/26/05.
9. The National Women's Health Information Center. US Department of Health and Human Services. Available
at http://www.4woman.gov/ Last accessed 1/26/05.
10. The Ad Council. Breastfeeding Awareness Campaign. Available at http://www.adcouncil.org/campaigns/

breastfeeding/ Last accessed 1/26/05.
11. Freed GL, Clark SJ, Sorenson J, et al. National assessment of physician's breast-feeding knowledge,
attitudes, training and experience. JAMA 1995;273(6):472-476.
12. Freed GL, Clark SJ, Curtis P, et al. Breast-feeding education and practice in family medicine. J Fam Pract
1995;40(3):263-269.
13. Williams EL, Hammer LD. Breastfeeding attitudes and knowledge of physicians-in-training. Am J Prev Med
1995;11:26-33.
14. Philipp BL, Merewood A, Gerendas E. Breastfeeding information in pediatric textbooks needs
improvement. Academy of Breastfeeding Medicine News and Views 2002;8(3):27.
15. Budin P. The nursling: the feeding and hygiene of premature and full-term infants. London, UK: Caxton
Publishing, 1907. Maloney WJ, translator.
16. Hess JH. Premature and congenitally diseased infants. Philadelphia: Lea and Febiger, 1922.
17. Gordon HH, Levine SZ, McNamara H. Feeding of premature infants. A comparison of human and cow's
milk. Am J Dis Child 1947;73:442-452.
18. Benjamin MH, Gordon HH, Marples E. Calcium and phosphorus requirements of premature infants. Am J
Dis Child 1943; 65:412-425.
19. Ruowei L, Zhao Z, Mokdad A, et al. Prevalence of breastfeeding in the United States: the 2001 National
Immunization Survey. Pediatrics 2003;111:1198-1201.
20. Ahluwalia I, Morrow B, Hsia J, et al. Who is breast-feeding? Recent trends from the pregnancy risk
assessment and monitoring system. J Pediatr 2003;142:486-491.
21. Ryan AS, Wenjun Z, Acosta A. Breastfeeding continues to increase into the new millennium. Pediatrics
2002;110: 1103-1109.
22. Ross Products Division. Breastfeeding trends through 2000. Available at http://ross.com/aboutross/
Survey.pdf. Last accessed 1/26/05.
23. Labbok M, Krasovec K. Toward consistency in breastfeeding definitions. Stud Fam Plan 1990; 21:226.
24. Breastfeeding Committee for Canada. Breastfeeding Definitions and Data Collection Periods, March 2004.
http://www.breastfeedingcanada.ca Date last accessed: 1/26/05
25. Marinelli K. Personal communication from Connecticut Children's Medical Center NICU database: 2004.
26. Erenkrantz RA, Ackerman BA, Mezger J, et al. Breast-feeding premature infants: incidence and success.
Pediatr Res 1985;19: 199A(abst530).
27. Furman L, Minich NM, Hack M. Breastfeeding of very low birth weight infants. J Hum Lact 1998;14(1):2934.
28. Hill PD, Ledbetter RJ, Kavanaugh KL. Breastfeeding patterns of low-birth-weight infants after hospital
discharge. JOGN Nurs 1997;26(2):189-197.
29. Lefebvre FL, Ducharme M. Incidence and duration of lactation and lactational performance among mothers
of low-birth-weight and term infants. CMAJ 1989;140:1159-1164.
30. Marinelli K, Page K, Burke G. Influence of NICU admission on choice and duration of breastfeeding in
mothers of preterm infants. Academy of Breastfeeding Medicine News and Views 1998;4:23.
31. Hill PD, Ledbetter RJ, Kavanaugh KL. Breastfeeding patterns of low-birth-weight infants after hospital
discharge. JOGN Nurs 1997;26:189-197.
32. Richards MT, Lang MD, McIntosh C, et al. Breastfeeding the VLBW infant: successful outcome and
maternal expectations. Pediatr Res 1986;20:383A(abst1385).
33. Kaufman KJ, Hall LA. Influence of the social network on choice and duration of breast-feeding in mothers
of preterm infants. Res Nurs Health 1989;12:149-159.
34. Killersreiter B, Grimmer I, Buhrer C, et al. Early cessation of breast milk feeding in very low birthweight
infants. Early Hum Dev 2001;60:193-205.
35. Furman L, Minich N, Hack M. Correlates of lactation in mothers of very low birth weight infants. Pediatrics
2002;109(4):e57 Available at http://www.pediatrics.org/cgi/content/full/109/4/e57. Last accessed _____.
36. Powers NG, Bloom B, Peabody J, et al. Site of care influences breast milk feedings at NICU discharge. J
Perinatol 2003;23: 10-13.
37. Meberg A, Willgraff S, Sande HA. High potential for breastfeeding among mothers giving birth to pre-term
infants. Acta Paediatr Scand 1982;71:661-662.
38. Verronen P. Breast feeding of low birthweight infants. Acta Paediatr Scand 1985;74:495-499.
39. Hunkeler B, Aebi C, Minder CE, et al. Incidence and duration of breast-feeding of ill newborns. J Pediatr
40. Barros FC, Victoria CG, Vaughn JP, et al. Birth weight and duration of breast-feeding: are the beneficial
effects of human milk being over-estimated? Pediatrics 1986;78:656-661.
41. Yip E, Lee J, Sheehy Y. Breast-feeding in neonatal intensive care. J Paediatr Child Health 1996;32:296298.
42. Heinig MJ. Host defense benefits of breastfeeding for the infant: effect of breastfeeding duration and
exclusivity. Pediatr Clin North Am 2001;48(1):105-123.
43. Davis MK. Breastfeeding and chronic disease in childhood and adolescence. Pediatr Clin North Am 2001;48
(1):125-141.
44. Reynolds A. Breastfeeding and brain development. Pediatr Clin North Am 2001;48(1):159-171.
45. Labbok MH. Effects of breastfeeding on the mother. Pediatr Clin North Am 2001;48(1):143-158.
46. Butte N. The role of breastfeeding in obesity. Pediatr Clin North Am 2001;48(1):189-198.
47. Schanler RJ. The use of human milk for premature infants. Pediatr Clin North Am 2001;48:207-219.
48. Gross SJ, David RJ, Bauman L, et al. Nutritional composition of milk produced by mothers delivering
preterm. J Pediatr 1980;96: 641-644.
49. Hibberd CM, Brooke OG, Carter ND, et al. Variations in the composition of breast milk during the first five
weeks of lactation: implications for the feeding of preterm infants. Arch Dis Child 1982;57:658-662.
50. Maas YGH, Gerritsen J, Hart AAM, et al. Development of macronutrient composition of very preterm
human milk. Br J Nutr 1998;80:35-40.
51. Lawrence RA, Lawrence RM. Breastfeeding: a guide for the medical profession. Philadelphia: Mosby, 1999.
52. Wagner V, Stockhausen JG. The effects of feeding human milk and adapted milk formulae on serum lipid
and lipoprotein levels in young infants. Eur J Pediatr 1988;147:292-295.
53. Owen CG, Whincup PH, Odoki K, et al. Infant feeding and blood cholesterol: a study in adolescents and a
systematic review. Pediatrics 2002;110:597-608.
54. Bergstrom E, Hernell O, Persson LA, et al. Serum lipid values in adolescents are related to family history,
infant feeding and physical growth. Atherosclerosis 1995;117:1-13.
55. Butte NF, Wong WW, Ferlic L, et al. Energy expenditure and deposition of breast-fed and formula-fed
infants during early infancy. Pediatr Res 1990;28:631-640.
56. Butte NF, Smith EO, Garza C. Energy utilization of breast-fed and formula-fed infants. Am J Clin Nutr
1990;51:350-358.
P.441
57. Putet G, Senterre J, Rigo J, et al. Nutrient balance, energy utilization and composition of weight gain in
very-low-birth-weight infants fed pooled human milk or preterm formula. J Pediatr 1984;105:79-85.
58. Whyte RK, Haslam R, Vlainic C, et al. Energy balance and nitrogen balance in growing low birthweight
infants fed human milk or formula. Pediatr Res 1983;17:891-898.
59. Lubetzky RL, Vaisman N, Mimouni FB, et al. Energy expenditure in human milk- versus formula-fed
preterm infants. J Pediatr 2003;143:750-753.
60. Shulman RJ, Schanler RJ, Lau C, et al. Early feeding, antenatal glucocorticoids, and human milk decrease
intestinal permeability in preterm infants. Pediatr Res 1998;44:519-523.
61. Cavell B. Gastric emptying in infants fed human milk or infant formula. Acta Paediatr Scand 1981;70:639641.
62. Ewer AK, Durbin GM, Morgan MEI, et al. Gastric emptying in preterm infants. Arch Dis Child 1994;71:E24E27.
63. Uraizee F, Gross S. Improved feeding tolerance and reduced incidence of sepsis in sick very low
birthweight infants fed maternal milk. Pediatr Res 1989;25:298A.
64. Lucas A. AIDS and milk bank closures. Lancet 1987;1:1092-1093.
65. Shulman RJ, Schanler RJ, Lau C, et al. Early feeding, feeding tolerance, and lactase activity in preterm
66. Hamosh M. Bioactive factors in human milk. Pediatr Clin North Am 2001;48:69-86.
67. Sheard NF, Walker WA. The role of breast milk in the development of the gastrointestinal tract. Nutr Rev
1988;46:1-8.
68. Carpenter G. Epidermal growth factor is a major growth-promoting agent in human milk. Science
1980;210:198-199.
69. Petschow BW, Carter DL, Hutton GD. Influence of orally administered epidermal growth factor on normal
and damaged intestinal mucosa of rats. J Pediatr Gastroenterol Nutr 1993; 17:49-57.
70. Heird WC, Schward SM, Hansen IH. Colostrum-induced enteric mucosal growth in beagle puppies. Pediatr
Res 1984;18:512-515.
71. Widdowson EM, Colombo VE, Artavanis CA. Changes in the organs of pigs in response to feeding for the
first 24 hours after birth. II. The digestive tract. Biol Neonate 1976;28:272.
72. Xanthou M, Bines J, Walker WA. Human milk and intestinal host defense in newborns: an update. Adv
Pediatr 1995;42:171-208.
73. Kleinman RE, Walker WA. The enteromammary immune system. An important new concept in breast milk
host defense. Dig Dis Sci 1979;24:876-882.
74. Mathur NB, Dwarkadas AM, Sharma VK, et al. Anti-infective factors in preterm human colostrum. Acta
Paediatr Scand 1990;79: 1039-1044.
75. Scariati PD, Grummer-Strawn LM, Fein SB. A longitudinal analysis of infant morbidity and the extent of
breastfeeding in the United States. Pediatrics 1997;99(6)URL: Available at http://www.pedatrics.org/cgi/
content/full/99/6/e5. Last accessed 1/26/05.
76. Bachrach VR, Schwarz E, Bachrach LR. Breastfeeding and the risk of hospitalization for respiratory disease
in infancy. Arch Dis Adolesc Med 2003;157:237-243.
77. Dewey KG, Heinig MJ, Nommsen-Rivers LA. Differences in morbidity between breast-fed and formula-fed
78. Duncan B, Ey J, Holberg CJ, et al. Exclusive breast-feeding for at least 4 months protects against otitis
media. Pediatrics 1993; 91:867-872.
79. Chen A, Rogan WJ. Breastfeeding and the risk of postneonatal death in the United States. Pediatrics
2004;113:e435-e439. Available at URL: http://www.pediatrics.org/cgi/content/full/113/5/e435. Last accessed
1/26/05.
80. Winberg J, Wessner G. Does breast milk protect against septicaemia in the newborn? Lancet 1971;1:10911094.
81. Narayanan I, Prakash K, Bala S, et al. Partial supplementation with breast-milk for prevention of infection
in low-birth-weight infants. Lancet 1980;2:561-563.
82. Narayanan I, Prakash K, Gujral VV. The value of human milk in the prevention of infection in the high-risk
low-birth-weight infant. J Pediatr 1981;99:496-498.
83. Narayanan I, Prakash K, Prabhakar AK, et al. A planned prospective evaluation of the anti-infective
property of varying quantities of expressed human milk. Acta Paediatr Scand 1982;71: 441-445.
84. El-Mohandes AE, Picard MB, Simmens SJ, et al. Use of human milk in the intensive care nursery decreases
the incidence of nosocomial sepsis. J Perinatol 1997;17:130-134.
85. Hylander MA, Strobino DM, Dhanireddy R. Human milk feedings and infection among very low birth weight
infants. Pediatrics 1998;102:e38. Available at http://www.pediatrics.org/cgi/cotent/full/102/3/e38; Accessed

1/26/05.
86. Schanler RJ, Shulman RJ, Lau C. Feeding strategies for premature infants: beneficial outcomes of feeding
fortified human milk versus preterm formula. Pediatrics 1999;103:1150-1157.
87. Lucas A, Cole TJ. Breast milk and neonatal necrotising enterocolitis. Lancet 1990;336:1519-1523.
88. Bisquera JA, Cooper TR, Berseth CL. Impact of necrotizing enterocolits on length of stay and hospital
charges in very low birth weight infants. Pediatrics 2002;109:423-428.
89. Weimer J. The economic benefits of breastfeedinga review and analysis. Washington: Food and Rural
Economic Research Service; U.S. Department of Agriculture; Food and Assistance Research report no 13;
2001.
90. Eibl MM, Wolf HM, Furnkranz, et al. Prevention of necrotizing enterocolitis in low-birth-weight infants by
IgA-IgG feeding. N Engl J Med 1988;319:1-7.
91. Moya FR, Eguchi H, Zhao B, et al. Platelet-activating factor acetylhydrolase in term and preterm human
milk: a preliminary report. J Ped Gastroenterol Nutr 1994;19:236-239.
92. Furman L, Taylor G, Minich N, Hack M. The effect of maternal milk on neonatal morbidity of very low-birthweight infants. Arch Pediatr Adolesc Med 2003;157:66-71.
93. Blaymore Bier JA, Oliver T, Ferguson A, et al. Human milk reduces outpatient upper respiratory symptoms
in premature infants during their first year of life. J Perinatol 2002;22: 354-359.
94. Morley R, Cole TJ, Powell R, et al. Mother's choice to provide breast milk and developmental outcome.
Arch Dis Child 1988;63:1382-1385.
95. Lucas A, Morley R, Cole TJ, et al. Breast milk and subsequent intelligence quotient in children born
preterm. Lancet 1992;339: 261-264.
96. Lucas A, Morley R, Cole TJ. A randomized multicentre study of human milk versus formula and later
development in preterm infants. Arch Dis Child 1994;70:F141-F146.
97. Blaymore Bier JA, Oliver T, Ferguson AE, et al. Human milk improves cognitive and motor development of
premature infants during infancy. J Hum Lact 2002;18:361-367.
98. Anderson JW, Johnstone BM, Remley DT. Breast-feeding and cognitive development: a meta-analysis. Am
J Clin Nutr 1999;70: 525-535.
99. Carlson SE, Werkman SH, Rhodes PG, et al. Visual-acuity development in healthy preterm infants: effect
of marine-oil supplementation. Am J Clin Nutr 1993;58:35-42.
100. Amin SB, Merle KS, Orlando MS, et al. Brainstem maturation in premature infants as a function of
enteral feeding type. Pediatrics 2000;106:318-322.
101. Uauy RD, Birch DG, Birch EE, et al. Effect of dietary omega-3 fatty acids on retinal function of very-low-
birth-weight neonates. Pediatr Res 1990;28:485-492.
102. Birch E, Birch D, Hoffmann D, et al. Breast-feeding and optimal visual development. J Pediatr Ophthalmol
Strabismus 1993;30: 33-38.
103. Hoffman DR, Birch EE, Castaneda YS, et al. Visual function in breast-fed term infants weaned to formula
with or without long-chain polyunsaturates at 4 to 6 months: a randomized clinical trial. J Pediatr
2003;142:669-677.
104. Meier P. Bottle- and breast-feeding; effects on transcutaneous oxygen pressure and temperature in
preterm infants. Nurs Res 1988;37:36-41.
105. Meier P. Suck-breathe patterning during bottle and breastfeeding for preterm infants. In: David TJ, ed.
Major controversies in infant nutrition, International Congress and Symposium, series 215. London: Royal
Society of Medicine Press, 1996:9-20.
106. Meier P, Anderson GC. Responses of small preterm infants to bottle- and breast-feeding. MCN Am J
Matern Child Nurs 1987; 12:97-105.
107. Blaymore Bier J, Ferguson A, Anderson L, et al. Breast-feeding of very low birth weight infants. J Pediatr
1993;123:773-778.
108. Blaymore Bier JA, Ferguson AE, Morales Y, et al. Breastfeeding infants who were extremely low birth
weight. Pediatrics 1997; 100:e3.
P.442
109. Meier PP. Supporting lactation in mothers with very low birth weight infants. Pediatr Ann 2003;32:317325.
110. Meier PP, Engstrom JL, Spanier-Mingolelli SR et al. Dose of own mothers' milk provided by low-income
and non-low income mothers of very low birthweight infants (abstract). Pediatr Res 2000; 47:292A.
111. Russo J, Russo IH. Development of the human mammary gland. In: Neville MD, Daniel CE eds. The
mammary gland: development, regulation and function. New York: Plenum Press, 1987:67-97
112. Riordan J. Anatomy and physiology of lactation. In: Riordan J, ed. Breastfeeding and human lactation,
3rd ed. Massachusetts: Jones and Bartlett, 2005.
113. deCarvalho M, Anderson DM, Giangreco A, et al. Frequency of milk expression and milk production by
mothers of nonnursing premature neonates. Am J Dis Child 1985;139(5):483-485.
114. Hinds LA, Tyndale-Biscoe CH. Prolactin in the marsupial macropus engenii during the estrous cycle,
pregnancy and lactation. Biol Reprod 1982;26:391-398.
115. Zuppa AA, Tornesello A, Papacci P, et al. Relationship between maternal parity, basal prolactin levels and
neonatal breast milk intake. Biol Neonate 1988;53(3):144-147.
116. Ellis L, Picciano MF. Prolactin variants in term and preterm milk: altered structural characteristic,
biological activity and immunoreactivity. Endocr Regul 1993;27:193-200.
117. Morton JA. Strategies to support extended breastfeeding of the premature infant. Adv Neonatal Care
2002;2(5):267-282.
118. Chatterton RT, Hill PD, Aldag JC, et al. Relation of plasma oxytocin and prolactin concentrations to milk
production in mothers of preterm infants: influence of stress. J Clin Endocrinol Metab 2000;85(10):3661-3668.
119. Hill PD, Aldag JC, Chatterton RT. Effects of pumping style on milk production in mothers of non-nursing
preterm infants. J Hum Lact 1999:15(3):209-216.
120. Hill PD, Aldag JC, Chatterton RT. Initiation and frequency of pumping and milk production in mothers of
non-nursing preterm infants. J Hum Lact 2001;17(1):9-13.
121. Furman L, Minich N, Hack M. Correlates of lactation in mothers of very low birth weight infants.
Pediatrics 2002;109(4):1-7.
122. Hill PD, Aldag JC, Chatterton RT. The effect of sequential and simultaneous breast pumping on milk
volumes and prolactin levels: a pilot study. J Hum Lact 1996;12(3): 193-199.
123. Jones E, Dimmock PW, Spencer SA. A randomised controlled trial to compare methods of milk expression
after preterm delivery. Arch Dis Child Fetal Neonatal Ed 2001;85(2): 91-95.
124. Hill PD, Brown LP, Harker TL. Initiation and frequency of breast expression in breastfeeding mothers of
LBW and VLBW infants. Nurs Res 1995;44(6):352-355.
125. Budd SC, Erdman SH, Long DM, et al. Improved lactation with metoclopramide: a case report. Clin
Pediatr (Phila) 1993;32(1): 53-57.
126. da Silva OP, Knoppert DC, Angelini MM, et al. Effect of domperidone on milk production in mothers of
premature newborns: a randomized, double-blind, placebo-controlled trial. CMAJ 2001;164(1):17-21.
127. Emery MM. Galactogogues: drugs to induce lactation. J Hum Lact 1996;12(1):55-57.
128. Hurst NM, Valentine CJ, Renfro L, et al. Skin-to-skin holding in the neonatal intensive care unit influences
maternal milk volume. J Perinatol 1997;17(3):213-217.
129. Bier JA, Ferguson AE, Morales Y, et al. Comparison of skin-to-skin contact with standard contact in lowbirth-weight infants who are breast-fed. Arch Pediatr Adolesc Med 1996;150(12):1265-1269.
130. Ehrenkranz RA, Ackerman BA. Metoclopramide effect on faltering milk production by mothers of
premature infants. Pediatrics 1986;78(4):614-620.
131. Williams-Arnold LD. Human Milk Storage for Healthy Infants and Children. Sandwich, MA, Health
Education Associates Inc, 2000.
132. Arnold, LDW. Recommendation for Collection, Storage and Handling of a Mother's Milk for Her Own
Infant in the Hospital Setting. 3rd edition. Denver, The Human Milk Banking Association of North America,
1999.
133. Arnold LDW. Using banked donor milk in clinical settings. In: Cadwell K, ed. Reclaiming breastfeeding for
the United States; protection, promotion and support. Boston: Jones and Bartlett, 2002: 137-159.
134. Arnold LDW. The cost-effectiveness of using banked donor milk in the neonatal intensive care unit:
prevention of necrotizing enterocolitis. J Hum Lact 2002;18:172-177.
135. Arnold LDW. Donor human milk banking. In: Riordan J, ed. Breastfeeding and human lactation, 3rd ed.
Boston: Jones and Bartlett, 2005:409-431.
136. Wight NE. Donor milk for preterm infants. J Perinatol 2001;21: 249-254.
137. World Health Organization. Global strategy for infant and young child feeding. Geneva: WHO, 2003.
138. American Academy of Pediatrics. Human Milk. In: Red Book: 2003 Report of the Committee on Infectious
Diseases, Pickering LK, Baker CJ, Overtorf GD, Prober CG, ed. 26th ed. Elk Grove Village, IL, American
Academy of Pediatrics 2003:117-123.
139. Protocol Committee Academy of Breastfeeding Medicine. In: Cordes R, Howard CR, Powers N, et al, eds.
Clinical Protocol Number 3: Hospital guidelines for the use of supplementary feedings in the healthy term
breastfed neonate. Academy of Breastfeeding Medicine. 2002. Available at http://www.bfmed.org; accessed
1/26/05.
140. Human Milk Banking Association of North America (HMBANA). In: Tully M, ed. Guidelines for the
establishment and operation of a donor human milk bank. Raleigh, NC: HMBANA, 2003.
141. Tully DB, Jones F, Tully MR. Donor milk: what's in it and what's not. J Hum Lact 2001;17:152-155.
142. Arnold LDW. How North American donor milk banks operate: results of a survey, part 2. J Hum Lact
1997;13:243-246.
143. Slusher T, Hampton R, Bode-Thomas F, et al. Promoting the exclusive feeding of own mother's milk
through the use of hindmilk and increased maternal milk volume for hospitalized, low birth weight infants
(<1800 grams) in Nigeria: a feasibility study. J Hum Lact 2003;19:191-198.
144. Valentine CJ, Hurst NM, Schanler RJ. Hindmilk improves weight gain in low-birth-weight infants fed
human milk. J Pediatr Gastroenterol Nutr 1994;18:474-477.
145. Centers for Disease Control and Prevention. Enterobacter sakazakii infections associated with the use of
powdered infant formulaTennessee, 2001. MMWR Morb Mort Wkly Rep 2002;51:297-300.
146. Van Acker J, de Smet F, Muyldermans G, et al. Outbreak of necrotizing enterocolitis associated with
enterobacter sakazakii in powdered milk formula. J Clin Microbiol 2001;39:293-297.
147. Weir E. Powdered infant formula and fatal infection with Enterobacter sakazakii. CMAJ 2002;166
(12):1570.
148. Clark NC, Hill BC, O'Hara CM, et al. Epidemiologic typing of Enterobacter sakazakii in two neonatal
nosocomial outbreaks. Diagn Microbiol Infect Dis 1990;13(6):467-472.
149. Simmons BP, Gelfand MS, Haas M, et al. Enterobacter sakazakii infections in neonates associated with
intrinsic contamination of a powdered infant formula. Infect Control Hosp Epidemiol 1989;10(9):398-401.
150. Biering G, Karlsson S, Clark NC, et al. Three cases of neonatal meningitis caused by Enterobacter
sakazakii in powdered milk. J Clin Microbiol 1989;27(9):2054-2056.
151. US Food and Drug Administration. Health professionals letter on Enterobacter sakazakii infections
associated with use of powdered (dry) infant formulas in neonatal intensive care units. April 11, 2002; rev.
October 10, 2002. Available at http://www.cfsan.fda.gov/~dms/inf-ltr3.html. Last accessed 1/26/05.
152. Ronnholm KAR, Perheentupa J, Siimes MA. Supplementation with human milk protein improves growth
of small premature infants fed human milk. Pediatrics 1986;77:649-653.
153. Chappell JE, Clandinin MT, Kerney-Volpe C, et al. Fatty acid balance studies in premature infants fed
human milk or formula: effect of calcium supplementation. J Pediatr 1986;102:439.
154. Michaelsen KF, Skafte L, Badsberg JH, et al. Variation in micronutrients in human bank milk: influencing
factors and implications for human milk banking. J Pediatr Gasrtoenter Nutr 1990;11:229-239.
155. Voyer M, Senterre J, Rigo J, et al. Human milk lacto-engineering. Acta Paediatr Scand 1984;73:302-306.
156. Polberger S, Raiha NCR, Juvonen P, et al. Individualized protein fortification of human milk for preterm
infants: composition of ultrafiltrated human milk protein and a bovine whey fortifier. J Pediatr Gasrtoenter
Nutr 1999;29:332-338.
157. Lucas A, Fewtrell MS, Morley R, et al. Randomized outcome trial of human milk fortification and
developmental outcome in preterm infants. Am J Clin Nutr 1996;64:142-151.
P.443
158. Schanler RJ, Schulman RJ, Lau C, et al. Feeding strategies for premature infants: randomized trial of
gastrointestinal priming and tube-feeding method. Pediatrics 1999;103:434-439.
159. Reis BB, Hall RT, Schanler RJ, et al. Enhanced growth of preterm infants fed a new powdered human
milk fortifier: a randomized control trial. Pediatrics 2000;106:581-588.
160. Ewer AK, Yu VYH. Gastric emptying in preterm infants: the effect of breast milk fortifier. Acta Paediatr
1996;85:1112-1115.
161. Sankaran K, Papageorgiou A, Ninan A, et al. A randomized, controlled evaluation of two commercially
available human breast milk fortifiers in health preterm neonates. J Am Diet Assoc 1996;96:1145-1149.
162. Moody GJ, Schanler RJ, Lau C, et al. Feeding tolerance in premature infants fed fortified human milk. J
Pediatr Gastoenterol Nutr 2000;30:408-412.
163. Lucas A, Fewtrell MS, Morley R, et al. Randomized outcome trial of human milk fortification and
developmental outcome in preterm infants. Am J Clin Nutr 1996;64:142-151.
164. Quan R, Yang C, Rubinstein S, et al. The effect of nutritional additives on anti-infective factors in human
milk. Clin Pediatr 1994;33:325-328.
165. Jocson MAL, Mason EO, Schanler RJ. The effects of nutrient fortification on varying storage conditions on
host defense properties of human milk. Pediatrics 1997;100:240-243.
166. Chan GM. Effects of powdered human milk fortifiers on the antibacterial actions of human milk. J
Perinatol 2002;23:620-623.
167. Lessaris KJ, Forsythe DW, Wagner CL. Effect of human milk fortifier on the immunodetection and
molecular mass profile of transforming growth factor-alpha. Biol Neonate 2000;77:156-161.
168. Lawrence RM. Host resistance factors and immunologic significance of human milk. In: Lawrence RA,
Lawrence RM, eds. Breastfeeding: a guide for the medical profession, 5th ed. New York: Mosby, 1999;172.
169. Walker WA. Antigen penetration across the immature gut: effect of immunologic and maturational
factors in colostrum. In: Ogra PL, Dayton D, eds. Immunology of breast milk. New York: Raven Press, 1979.
170. Narayanan I, Prakash K, Verma RK, et al. Administration of colostrum for the prevention of infection in
the low birth weight infant in a developing country. J Trop Pediatr 1983;29:197-200.
171. Gomez HM, Sanabria ER, Marquette CM. The mother kangaroo programme. International Child Health
1992;3:55-56.
172. Cattaneo A, Davanzo R, the International Network for Kangaroo Mother Care, et al. Recommendations
for the implementation of kangaroo mother care for low birth weight infants. Acta Paediatrica 1998;87:440445.
173. Ludington SM, Ferreira CN, Goldstein MR. Kangaroo care with a ventilated preterm infant. Acta
Paediatrica 1998;87:711-716.
174. Tornhage CJ, Stuge E, Lindberg T, et al. First week kangaroo care in sick very preterm infants. Acta
Paediatrica 1999;88:1402-1404.
175. Whitelaw A, Heisterkamp G, Sleath K, et al. Skin to skin contact for very low birthweight infants and
their mothers. Arch Dis Child 1998;63:1377-1381.
176. Charpak N, Ruiz-Pelaez JG, Figueroa de CZ, et al. A randomized, controlled trial of kangaroo mother
care: results of follow-up at 1 year of corrected age. Pediatrics 2001;108:1072-1079.
milk volume. J Perinatol 1997;17:213-217.
178. Kirsten GF, Bergamn NJ, Hann FM. Kangaroo mother care in the nursery. Ped Clin North Am
2001;48:443-452.
179. Anderson GC. Current knowledge about skin-to-skin (kangaroo care) for preterm infants. J Perinatol
1991;11:216-226.
maternal milk volume. J Perinatol 1997;17(3):213-217.
181. Furman L, Kennell J. Breast milk and skin-to-skin kangaroo care for premature infants. Avoiding bonding
failure. Acta Paediatrica 2000:89:1280-1283.
182. Tornhage CJ, Serenius F, Uvnas-Moberg K, et al. Plasma somatostatin and cholecystokinin levels in
preterm infants during kangaroo care with and without tube-feeding. J Pediatr Endocrinol Metab 1998;11:645651.
183. Lau C, Alaguagurusamy R, Schanler RJ, et al. Characterization of the developmental stage of sucking in
preterm infants during bottle feeding. Acta Pediatrica 2000;89:846-852.
184. Daniels H, Devlieger H, Minami T, et al. Infant feeding and cardiorespiratory maturation. Neuropediatrics
1990;21:9-10.
185. Nyquist KH, Sjoden P, Ewald U. The development of preterm infants' breastfeeding behavior. Early Hum
Dev 1999;55; 247-264.
186. Nyquist KH, Farnstrand C, Eeg-Olofsson KE, et al. Early oral behavior in preterm infants during
breastfeeding: an electromoygraphic study. Acta Pediatrica 2001;90:658-663.
187. Martell M, Martinez G, Gonzalez M, et al. Suction pattern in preterm infants. J Perinat Med 1993;21:363369.
188. Morton JA. The long road home: strategies to support extended breastfeeding of the premature infant.
Adv Neonatal Care 2002;2(5):267-282.
189. Lau C, Sheena HR, Shulman RJ, et al. Oral feeding in low birth weight infants. J Pediatr 1997;130
(4):561-569.
190. Neifert M, Seacat J. Practical aspects of breastfeeding the premature infant. Perinatology Neonatology
1988;12:24-30.
191. Meier PP, Lysakowski TY, Engstrom JL, et al. The accuracy of test weighing for preterm infants. J Pediatr
Gastroent Nutr 1990; 10:62-65.
192. Meier PP, Engstrom JL, Crichton CL. A new scale for in-home test weighing for mothers of preterm and
high risk infants. J Hum Lact 1994;10:163-168.
193. Stine MJ. Breastfeeding the premature newborn: A protocol without bottles. J Hum Lact 1990;6(4):167170.
194. Newman J. Breastfeeding problems associated with early introduction of bottles and pacifiers. J Hum Lact
1990;6(2):59-63.
195. Walker M. Breastfeeding the premature infant. NAACOG's Clinical Issues in Perinatal and Women's
health. Nursing 1992;3:620-633.
196. Woolridge MW. The anatomy of infant sucking. Midwifery 1986;2:164-171.
197. Righard L, Alade MO. Sucking technique and its effect on success of breastfeeding. Birth 1992;19:185189.
198. Hill PD, Ledbetter RJ, Kavanaugh KL. Breastfeeding pattern of low-birth-weight infants after hospital
discharge. J Obstet Gynecol Neonatal Nurs 1997;26:190-197.
199. Neifert M, Lawrence R, Seacat J. Nipple confusion: towards a formal definition. J Pediatr 1995;126:125129.
200. Barros FC, Victora CG, Semer TC, et al. Use of pacifiers is associated with decreased breastfeeding
duration. Pediatrics 1995;95: 497-499.
201. Righard L. Are breastfeeding problems related to incorrect breastfeeding technique and the use of bottles
and pacifiers? Birth 1998;25:40-44.
202. World Health Organization. Protecting, Promoting and Supporting Breastfeeding: The Special Role of
Maternity Services. Geneva, Switzerland: World Health Organization; 1989.
203. World Health Organization, Division of Child Health and Development. Evidence for the Ten Steps to
Successful Breastfeeding. Geneva, Switzerland: World Health Organization; 1998.
204. Shiao PK, DiFiore TE. A survey of gastric tube practices in level 11 and level 111 nurseries. Compr
Pediatr Nurs 19:209-220.
205. Kliethermes PA, Cross ML, Lanese MG, Johnson KM, Simon SD. Transitioning preterm infants with
nasogastric tube supplementation: increased likelihood of breastfeeding. JOGNN 1999;28:264-273.
206. Lang SL, Lawrence CJ, L'E Orme R. Cupfeeding: an alternative method of infant feeding. Arch Dis Child
1994;71:365-369.
207. Marinelli KA, Burke GS, Dodd VL. A comparison of the safety of cupfeedings and bottlefeedings in
premature infants whose mothers intend to breastfeeding. J Perinatol 2001;21(6):350-355.
208. Dowling DA, Meier PP, DiFiore JM, et al. Cup-feeding for preterm infaints: mechanics and safety. J Hum
Lact 2002;18(1): 13-20.
209. Oddy WH, Glenn K. Implementing the baby Friendly Hospital initiative: the role of finger feeding.
Breastfeeding Rev 2003; 11(1):5-9.
210. Wolf LS, Glass RP. Feeding and Swallowing Disorders in Infancy. Therapy Skill Builders, San Antonio, TX;
1992.
211. Jones L, Spencer A. Establishing successful preterm breastfeeding. part 3. Pract Midwife 2002;5(6):1819.
212. Meier PP, Brown LP, Hurst NM, et al. Nipple shields for preterm infants: effect on milk transfer and
duration of breastfeeding. J Hum Lact 2000;16(2):106-114.
213. Hurst NM, Meier PP. Breastfeeding the preterm infant. In: Riordan J, ed. Breastfeeding and Human
Lactation. 3rd ed. Massachusetts: Jones and Bartlett, 2005;395-376
P.444
214. Riordan J, Hoover K. Perinatal and intrapartum care. In: Riordan J, ed. Breastfeeding and Human
Lactation. 3rd ed. Massachusetts. Jones and Bartlett, 2005:199.
215. Jensen D, Wallace S, Kelsay P. LATCH: a breastfeeding charting system and documentation tool. J
Obstet Gynecol Neonatal Nurs 1994;23(1):27-32.
216. Mulford, C. The mother-baby assessment (MBA): an Apgar Score for breastfeeding. J Hum Lact 1992;8
(2):79-82.
217. Shrago L, Bocar D. The infant's contribution to breastfeeding. J Obstet Gynecol Neonatal Nurs 1990;19
(3):209-215.
218. Matthews MK. Developing an instrument to assess infant breastfeeding behavior in the early neonatal
period. Midwifery 1988; 4:154-163.
219. Janke JR. Development of the breastfeeding attrition prediction tool. Nurs Res 1994;43(2):100-104.
220. Hedberg Nyquist K, Rubertsson C, Ewald U, et al. Development of the Preterm Infant Breastfeeding
Behavior Scale (PIBBS): a study of nurse-mother agreement. J Hum Lact 1996;1:207-219.
221. Winikoff B, Laukaran V, Meyers D, Stone R. Dynamics of mother-infant feeding: mothers, professionals,
and the institutional context in a large urban hospital. Pediatrics 1986; 77:357-365.
222. Bernaix L. Nurses' attitudes, subjective norms, and behavioral intentions toward support of breastfeeding
mothers. J Hum Lact 2000;16:201-209.
223. Siddell E, Marinelli K, Froman R, Burke G. Evaluation of an educational intervention on Breastfeeding for
NICU Nurses. J Hum Lact 2003;19(3):293-302.
224. Bergevin Y, Dougherty C, Kramer MS. Do infant formula samples shorten the duration of breastfeeding?
Lancet 1983;1:1148-1153.
225. Frank DA, Wirtz SJ, Sorenson JR, Heeren T. Commercial discharge packs and breastfeeding counseling:
effects on infant feeding practices in a randomized trial. Pediatrics 1987;80: 845-854.
226. Philipp BL, Merewood A, Miller LW, et al. Baby-Friendly Hospital Initiative improves breastfeeding
initiation rates in a US hospital setting. Pediatrics 2001;108:677-681.
227. Merewood A, Phillip BL, Vchawala N, Vcimo S, The baby-friendly hospital initiative increases
breastfeeding rates in a US neonatal intensive care unit. J Hum Lact 2003;19:166-171.
228. Maisels MJ, King E. Length of stay, jaundice and hospital readmission. Pediatrics 1998; 101:995-998.
229. Brown AK, Damus K, Kim MT, et. al. Factors relating to readmission of term and near-term neonates in
the first two weeks of life. J Perinatal Med 1999; 27(4):263-275.
230. Wight NE, Breastfeeding the borderline (near-term) preterm infant. Pediatr Ann 2003;32(5):329-336.
231. Williams A. Hypoglycemia of the newborn: review of the literature. Geneva: World Health Organization,
1997;75:261-290. Download from: http://www.who.int/reproductive-health/docs/hypoglycaemia_newborn.
htm
232. Eidelman AI. Hypoglycemia and the breastfed neonate. Pediatr Clin North Am 2001;48:377-387.
233. Protocol Committee Academy of Breastfeeding Medicine, Eidelman AI, Howard CR, et al. Clinical protocol
number 1: guidelines for glucose monitoring and treatment of hypoglycemia in breastfed neonate. Academy of
Breastfeeding Medicine. Available at http://www.bfmed.org/protos.html; accessed 1/26/05.
234. Gartner LM, Herschel M. Jaundice and breastfeeding. Pediatr Clin North Am 2001; 48:389-399.
235. Gartner LM, Herschel M. Jaundice and the breastfed baby. In: Riordan J, ed. Breastfeeding and human
lactation, 3rd ed. Boston: Jones and Bartlett, 2005;311-321.
236. Lawrence RM, Lawrence RA. Given the benefits of breastfeeding, what contraindications exist? Pediatr
Clin North Am 2001;48: 235-251.
237. Howard CR, Lawrence RA. Xenobiotics and breastfeeding. Pediatr Clin North Am 2001;48:485-504.
238. American Academy of Pediatrics. Committee on Drugs. The transfer of drugs and other chemicals into
human milk. Pediatrics 2001;108:776-789.
239. Hale TW. Medications in breastfeeding mothers of preterm infants. Pediatr Annu 2003;32:337-347.
240. World Health Organization. HIV and infant feeding: a review of HIV transmission through breastfeeding.
Publication WHO/FRH/ NUT 98.3, UNAIDS/98.5, UNICEF/PD/NUT/(J) 98.2 Geneva: WHO, 1998.
241. Stagno S, Reynolds DW, Pass RF, et al. Breast milk and the risk of cytomegalovirus infection. N Engl J
Med 1980;302: 1073-1076.
242. Vochem M, Hamprecht K, Jahn G, et al. Transmission of cytomegalovirus to preterm infants through
breast milk. Pediatr Infect Dis J 1998;17:53-58.
Chapter 24
Thermal Regulation
Michael Friedman
Stephen Baumgart
A HISTORICAL PERSPECTIVE
Tarnier was an obstetrician in Paris who first applied modern concepts of incubation to human infants starting
around 1830 (1,2). Tarnier's incubator, the couveuse, has been widely recognized as the first one designed
specifically to care for premature babies. Tarnier and his student, Budin, studied premature human incubation
into the next century, reporting almost doubled survival in infants born at less than 2 kg. In the United States,
commercialization of Tarnier's and Budin's designs occurred, and the Rotch Incubator appeared at the
Colombian Exposition in Chicago in 1893 (3,4,5). Thereafter, in 1933 Blackfan and Yaglou (6) provided humidity
along with air warming within incubators, which improved the stability of infant temperature control. In the
1940s Chappel in Philadelphia added air isolation techniques to incubator care to prevent neonatal septic
infections recognized to occur more frequently in humid environments (7,8). In 1958, Silverman and associates
(9) challenged the need for humidity in incubators and used higher air temperatures than previously reported to
care for an ever smaller premature population surviving with modern techniques.
Toward Defining the Optimal Thermal-Neutral Environment

In 1959 Cross and Hill from the United Kingdom described a metabolically neutral temperature for achieving the
optimal environmental care of newborn animals and humans, suggesting that poor temperature maintenance
resulted in increased metabolic rates of oxygen and substrate consumption (10,11,12) (Fig. 24-1). Physiologic
responses to environmental temperature during incubation of premature newborn infants were investigated
systematically in 1962 by Brck and coworkers (13,14) from Germany. In 1969 Hey and associates (15,16,17)
from the United Kingdom defined the best operational incubator temperatures for preterm babies by describing
an algorithm of air and wall temperatures, taking humidification and swaddling (insulation) into account as well.
The Hey and Katz nomograms for regulating incubators following preterm birth, growth, and maturation remain
the benchmark standard for modern human incubation to date (Fig. 24-2).
THERMAL BALANCE AT THE BEGINNING OF LIFE

Fetal Thermal Regulation
The fetus generates heat during metabolism with cellular proliferation and differentiation, maintenance of intraand extracellular ion gradients, and transport of nutrients and wastes across cell membranes. Cardiac and
skeletal muscle work also generates heat in utero (18). Fetal ovine and human studies suggest that the rate of
fetal heat production is about 33 to 47 cal/kg/minute (18,19). Fetal-maternal temperature gradients in
mammals and humans have demonstrated that a difference in temperature of only 0.45C to 0.50C between
the umbilic arterial and venous blood is sufficient to eliminate the majority of metabolic heat via the placental
circulation (i.e., by forced convective transfer into the mother's uterine circulation) (20,21,22). Probably less
than 10% to 20% of heat is dissipated from the fetal skin into the amniotic fluid (natural convection and
conduction from the uterine wall). The mother additionally serves as a heat reservoir for the fetus, favoring the
dissipation of heat as a byproduct of fetal metabolism. Fatal hyperthermia may occur with an elevation of
maternal temperature or if the mother is unable to dissipate the excess heat produced during pregnancy.
Therefore, pregnant women are advised to avoid prolonged hot baths and exertion on hot and humid days.
Maternal fever should be treated aggressively with environmental cooling and with antipyretics and antibiotics
when indicated.
Transition in the Delivery Suite

At birth, a newly born infant is immediately exposed to a wet and cold environment. Without intervention, rapid
cooling by convection from the neonate's skin into the cold delivery room air (at least a 10.0C drop) and by
P.446
evaporation at a tremendous rate (0.58 kcal/mL of water loss) may result in a drop of the infant's body
temperature at a rate of 0.2C to 1.0C/minute. Although fetal response to cold stress is relatively insensitive
prenatally (22,23), increased infant activity (crying with agitated movement characteristic of cold exposure upon
birth), vasoconstriction, and nonshivering thermogenesis (shivering is not active in the human newborn) occurs
the instant the baby hits the cold air, mediated by the sympathetic nervous system (24). Triggered by
temperature sensation of the skin, infant metabolic rate may increase by two- to threefold and thus maintain
body temperature for a period of several hours in the term subject before thermogenic reserves of glycogen and
brown fat become depleted.
Figure 24-1 Basic concepts for defining neonatal thermal-neutral environmental temperature (horizontal axis)
as the minimal observed metabolic rate (vertical axis, measured indirectly as oxygen consumption). The
shaded region of this graph (the relatively narrow thermal-neutral zone) is bounded by upper and lower critical
temperatures for nonmetabolic or physical regulation of normal body temperature (e.g., by vasoconstriction,
vasodilatation, or changes in posture). Variation of environmental temperature outside this limited range
results in a metabolic rate increase, in which infant core temperature may remain normal, but at the expense
of increased metabolic expenditure (e.g., cold stress). Outside the thermal regulatory range of metabolic heat
production, inevitable body cooling or heating results, with eventual death at environmental extremes. (From
Baumgart S. Incubation of the human newborn infant. In: Pomerance JJ, Richardson CJ, eds. Neonatology for
the clinician. Norwalk, Conn: Appleton & Lange, 1993:139-150, with permission.)
Figure 24-2 Range of temperatures needed to provide neu-tral environmental conditions for babies lying
naked on an insulated mattress in draft-free surroundings (at about 50% relative humidity) with equal mean
radiant wall temperature and air temperature. The top graph represents a 1-kg infant at birth and the bottom
graph a 2-kg infant. Optimum temperature probably approximates the lower limit of each neutral range as
defined here. Approximately 1C should be added to these operative temperatures to derive the appropriate
neutral air temperature for a single-walled incubator when room temperature is less than 27C (80F), and
more should be added if room temperature is very much less than this. (From Hey EN, Katz G. The optimum
thermal environment for naked babies. Arch Dis Child 1970;45:328-334, with permission.)
Physiology of Neonatal Thermal Response

Brown fat is especially thermogenic in the term newborn, with large reserves located between the scapula, in
the axillae and perithymic region, and in the paraspinal and peri-nephric areas. Penetrated extensively by blood
vessels, which give it the brown appearance and which conduct heat generated into the central circulation, these
adipo-cytes are laden with excess triglyceride stores and numerous mitochondria. With cold stress, the
sympathetic surge acts directly upon cell surface receptors, which stimulate
P.447
cyclic-AMP-mediated lipoprotein lipase. Thyroid hormone also surges at birth and augments this effect
(5,25,26). -Oxidation is uncoupled in brown adipocytes, however, resulting in triglyceride breakdown and
resynthesis producing heat (27).
Figure 24-3 Immediate and potentially detrimental cardiovascular response to a sympathetic surge in
norepinephrine released after birth upon exposure of the neonate's temperature-sensitive skin to a cold
extrauterine environment (convection and evaporation). (From Baumgart S. Incubation of the human newborn
infant. In: Pomerance JJ, Richardson CJ, eds. Neonatology for the clinician. Norwalk, Conn: Appleton & Lange,
1993:139-150, with permission.)
Preterm infants with immature thermogenic response and without metabolic substrate reserves deposited over
the last trimester of pregnancy fare worse. As shown in Fig. 24-3, a sympathetic surge occurs at birth, with
massive neurohumoral secretion of noradrenaline from paraaortic nodes and the fetal adrenal (28). Systemic
and pulmonary vasoconstriction result, which may end in poor oxygen uptake and relative peripheral tissue
hypoxia. Lactic acid production ensues, with demise ultimately occurring secondary to cold stress. Tarnier in the
1800s recognized this in Paris, where there was no central heating and poor home and hospital insulation.
Early Intervention
Drying infants in the delivery suite interrupts the process of evaporation, and bundling infants in cotton blankets
to prevent exposure to cold air interrupts convective heat loss and provides insulation to retain the infant's
metabolic heat. Placing infants at the mother's breast and cradled into her axillary fold engenders conductive
heat transfer from the mother to the infant. Alternatively, and especially if early intervention is required to aid
transition (e.g., suctioning or oxygen administration), the infant is dried first and then placed onto dry bedding
under a radiant warmer while these procedures are performed. A convectively warmed incubator enclosure with
air temperatures ranging from 35.0C to 37.0C and a variety of plastic swaddling heat shields have been
advocated to prevent excessive cold exposure, especially during transition and hospital transport of premature
infants (29,30,31).
Once in the nursery or mother's room, bundled infants may be placed into either an open bassinet or incubator
(naked or bundled) and provided close monitoring of either axillary (preferred, 36.0C to 36.5C) or rectal (37
C to 37.5C) temperatures through the first few hours of life. Slightly premature infants (32 to 35 weeks) or
small-for-gestational-age babies may appear to have normal body temperature at the expense of metabolically
generated heat (9). Glass and associates (32) demonstrated that premature infants nurtured in dry incubator
environments of either 35.0C (slightly cool) or 36.5C in the first few days of life maintained a normal body
temperature, but experienced more weight loss in the cooler environment.
Kangaroo Care
Skin-to-skin care, now termed kangaroo care, has been promoted for nurturing premature infants who are held
naked between the mother's breasts as if in a kangaroo's pouch. The infant is in contact with the mother's warm
skin and is close to the breast for unlimited feeding. Fathers also can provide thermal support in this way.
Kangaroo care was first reported from Bogot, Columbia, where use of conventional incubators was limited and
mortality in nonincubated preterm births high. A large randomized trial from this country recently showed that
infants less than or equal to 2.00 kg placed under kangaroo care shortly after birth for prolonged periods
achieved transition safely and grew normally, had fewer nosocomial infections, and were discharged earlier,
particularly at less than or equal to 1.80 kg (33). Significant reduction in early mortality also has been observed.
During the 1980s the kangaroo technique was promoted for nurturance of nonmechanically ventilated, growing
premature infants in Scandinavian and some other European countries. Randomized clinical trials also have
demonstrated enhanced mother-infant attachment, greater maternal self-esteem, prolonged and enhanced
lactation, increased infant alertness, and better weight gain (34). Physiologic studies have focused on
demonstrating thermal-neutral metabolic response (minimal observed oxygen consumption) and temperature
stability in stable growing premature babies during kangaroo care. Moreover, vital signs and oxygenation
parameters were demonstrated to be more stable in preterm infants recovering from bronchopulmonary
dysplasia, with absence of periodic breathing and reduced apnea and bradycardia. Behavioral studies
demonstrate more homogenous sleep patterns, less irritability later in infancy, and more direct social eye
contact with caregivers (35).
In the intensive care nursery, kangaroo care may be initiated even during mechanical ventilation with
uncomplicated patients. Mothers are instructed to wear front-opening shirts, maintain careful hygiene without
open sores or rashes, and avoid use of lotions, oils, or perfumes. Maximum skin surface area contact is desirable
with a covering blanket to avoid outward convective and evaporative heat losses. Privacy and quiet must be
provided by the nursery staff for periods of 0.5 to 1 hour initially, and careful temperature monitoring by surface
thermistor or axillary thermometer should be performed at least every 15 minutes,

P.448
along with cardiorespiratory and noninvasive oxygen monitoring when indicated. Temperature deviation of more
than 0.5C of normal skin temperature (36.5C to 37.0C) should result in termination of the session and return
to incubator care. Periods up to 4 hours may be achieved. Mothers are encouraged to pump their breasts before
and after sessions because milk production is enhanced. Parents are empowered with the care of their infants,
and kangaroo care integrates the family into the neonatal intensive care team. Presently, no adverse reports
have been published, and the use of kangaroo care in modern intensive care settings is on the rise.
CONVECTION-WARMED INCUBATORS
A modern incubator consists of an optically transparent, plastic hood (3 mm thick) covering the infant, with
sidewall and hand-access ports. The infant lays on a bed platform, underneath which a tungsten element
electronically heats the air. Air is forced over this element by a fan, circulating heated air within the hood.
Temperature may be controlled thermostatically to regulate either the air or infant skin temperature (15,36).
Thermodynamics of Incubation
The physiology of mammalian (homeothermic) thermal regulation (37) may be summarized by the equation:
in which [Q with dot above] is the rate of either metabolic heat production (left side of the equation) or of heat
loss and heat stored (right side of the equation), generally expressed in kcal/kg/hour or in W/m2 (J/sec/m2). By
convention, heat production and heat losses (or storage) are expressed as positive values. When a mammal is
successfully maintaining normal body temperature, heat storage is zero, otherwise body temperature either
increases or decreases until a new thermal equilibrium is established at another temperature. Also, when an
environmental heat loss becomes a heat gain (e.g., under a radiant warmer), the gain is ex-pressed as a
negative loss.
Convection
The rate of heat transferred from an infant's skin into the incubator environment depends in part on the
insulation provided by the dermis and subcutaneous fascia (comprised primarily of white fat deposited late in
gestation). Preterm babies have almost no fatty fascia and, therefore, are more vulnerable to heat loss through
air (and skin blood flow) convection (37). Air convection is heat loss that takes place from the skin's surface into
the surrounding environment and is summarized by the equation:
in which two forms of skin-to-air convective heat loss occur, depending on the gradient between skin and air
temperatures (T), the complex geometry of surface area exposed and air thermal density k, and air movement
velocity Vn (38). The first is natural convection, which results from the gradient of temperature between the skin
surface and surrounding air (38,39). Natural convection cells form as warm air rises from the skin, conveying
heat and body moisture away from the surface of the baby. Air thus warmed subsequently cools and falls back
toward the baby, forming the convection cell. Such cells form over the curvature of the baby's exposed body
surface area. An infant in flexion leaves less surface area exposed (38). An infant extended and flaccid is able to
dissipate more heat. Posture may be a valuable observation in deciding the thermal comfort or discomfort of
even a preterm infant.
The second form of convection is forced convective air movement, usually occurring at air velocities 0.27 m/
sec. Forced convective heat loss is roughly proportional to an exponential power (n) of the velocity (V) of air
movement. Within forced convection-warmed incubators, manufacturers strive to render still the air near the
baby. Recent estimates of natural and forced-air convective heat loss from premature neonates nursed within
incubators indicate success in this strategy because natural convection was the only major loss observed to
occur (40). Heat also is lost to a lesser degree by respiratory convection and evaporation.
An example of different incubator convection designs affecting skin-to-air heat transfer is shown in Fig. 24-4,
P.449
adapted from a study of partitional calorimetry performed by Okken and associates (41). The left incubator
partition in this figure represents a natural convection-warmed incubator with no circulating fan. Air rose
passively from heating elements underneath the baby's mattress to warm the interior of the incubator hood. The
second partition on the right represents a more standard, fan-forced convection incubator as described
previously. Nonevaporative heat loss in the passive device (non-E, the sum of convection, radiation, and
conduction) was 60% compared to 47% in the forced convection-warmed incubator, whereas evaporative heat
loss (E) was 13% higher in the forced convective environment. These authors attributed this increased
evaporative loss to disturbance by the incubator's fan of a microenvironment of humid air layered near the
baby's skin.
Figure 24-4 Partitional calorimetry performed for premature infants in two different types of convectionwarmed incubators, natural and forced air. The rate of evaporative heat loss (E) is greater proportionally for
babies nurtured in a conventional forced-air warmed environment than nonevaporative heat losses (non-E)
from convection, conduction, and radiation. Heat balance is regulated by metabolic rate (M) and nonmetabolic
mechanisms for thermal conservation (e.g., vasoconstriction). (Adapted from Okken A, Blijham C, Franz W, et
al. Effects of forced convection of heated air on insensible water loss and heat loss in preterm infants in
incubators. J Pediatr 1982;101:108-112, with permission.)
TABLE 24-1 CALCULATED BODY SURFACE AREA: BODY MASS RATIO FOR ADULTS, LOW-BIRTHWEIGHT NEONATES, AND VERY-LOW-BIRTH-WEIGHT NEONATES
Body Mass (kg)
BSA (m2)
BSA/Mass (cm2/kg)
Adult
70
1.73
250
LBW premature
1.5
0.13
870
LBW premature
1.0
0.10
1,000
VLBW
0.5
0.065
1,300
LBW, low birth weight; VLBW, very low birth weight. (Adapted from Costarino AT, Baumgart S.
Neonatal water metabolism. In: Cowett RM, ed. Principles of perinatal neonatal matabolism. New York:
Springer-Verlag, 1991:623, with permission.)
Figure 24-5 Transepidermal water evaporation from the skin of premature neonates of gestations ranging
from 25 to 40 weeks, followed longitudinally from birth over the first month of life. Dehydration is most
dangerous in the most immature babies less than 28 weeks of gestation in the first week of life before skin
keratinization occurs. (From Sedin G, Hammarlund K, Nilsson GE, et al. Measurements of transepidermal water
loss in newborn infants. Clin Perinatol 1985;12: 79-99, with permission.)
Evaporation
The premature neonate loses large amounts of water (and, therefore, latent heat at 0.58 kcal/mL) through
evaporation from the skin for several reasons (42,43,44,45,46,47,48,49). First and most important is the
premature neonate's thin epidermis lacking keratin, which normally serves as a vapor barrier for older infants,
children, and adults. Very immature skin is associated with reduced survival, and evaporative rates from infants
less than 700 g may be likened to those of severe burn victims. Second, premature neonates lose excessive
amounts of water and heat due to an increased body surface area:body mass ratio as demonstrated in Table 241 (50). The very-low-birth-weight premature infant of less than or equal to 500 grams exposes six times the
area of the adult subject per kilogram of the largely water body mass. Third, the proportion of the extracellular
water mass in the very-low-birth-weight infant, which is exposed to the external environment through the
nonkeratinized epidermal layer, is significantly larger (51).
Hammerlund and Sedin (45) (Fig. 24-5) summarized the rate of transepidermal water loss in premature
newborn infants nurtured in incubators throughout the first month of life at different gestational ages. This
figure demonstrates that the newborn at 26 weeks of gestation may lose as much as 60 grams of water per m2/
hour (more than 180 mL/kg/day or 100 kcal/kg/day). Additional amounts of water may also be lost from upper
airways during respiration of nonhumidified air in incubators.
P.450
Incubator Humidification
Humidification inside modern incubators is accomplished by the evaporation of water from a reservoir located in
the air path over the heating element beneath the incubator mattress (52). Earlier descriptions of humidification
in incubators cited the use of nebulized mist to saturate the infant's environment (80% to 90% relative
humidity) (53,54). These latter techniques resulted in the proliferation of Pseudomonas infections and were
rejected resoundingly by a number of reports (53,54,55,56). Unfortunately, abandoning all use of vapor
(invisible humidification) based on infection risks encountered with particulate mists (visible humidification) has
resulted in running incubator hoods completely dry. Relative humidity levels inside dry incubators at 36C to
36.5C may drop to less than 10% to 15%, promoting large insensible water and latent heat losses (52). In
response to these concerns, the American Academy of Pediatrics recommended only moderate use of air
humidification 50% relative humidity in the vapor phase (i.e., gaseous), and not a particulate mist (liquid phase)
(57). Put simply, if water circulating near an infant is visible, infection may be more likely.
A more recent reevaluation of incubator humidification was conducted by Harpin and Rutter (52) in 1985. Thirtythree infants less than 30 weeks of gestation and less than 2 weeks of age were nurtured in vapor-humidified
incubators saturated at between 80% and 90% relative humidity. Two infants acquired Pseudomonas sepsis and
one died. Twenty-nine babies of similar gestation and maturity were nurtured in dry incubators: one suffered an
episode of Pseudomonas sepsis and died. Those who died with Pseudomonas infections did so after the study
had ended after the first 2 weeks of life. The authors recommended that humidification be used routinely in the
first 2 weeks of life to prevent evaporative losses and prevent skin desiccation. Humidification early in life may
be prudent for incubation of the very-low-birth-weight infant, in whom heat and water losses are excessive and
probably pathologic. Modern incubators now may be equipped with sophisticated solid-state temperature and
humidity control devices to prevent condensation at higher relative humidity levels between 60% and 80%.
Although verification is yet lacking, hopefully a reduced risk of infection will occur by maintaining humidity
without visible condensation.
Conductive Heat Loss and Warming Mattresses

Conductive heat loss results from contact of a baby's body with the solid surface of a bed platform:
in which k is contact surface area and the bed's heat conductivity constant, and (T) represents the temperature
gradient between the infant's body core (containing heat) and mean skin contact surface. D is the thickness of
the bed's conducting material. In general, insulating foam rubber (about 2.5-cm thickness) and double cotton
blanket batting results in negligible conductive heat loss by providing insulation from the metal or plastic bed
table (i.e., reducing the value of k). Recently, however, incubator manufacturers and water mattress companies
have provided evidence that exogenous heat application through carefully conducted bed surfaces maintained at
less than or equal to 39C to prevent burn injury to areas of skin contact may reduce environmental heat
requirements from incubators or radiant warmers for very-low-birth-weight subjects (58,59). Moreover,
application of warmed mattresses during extreme environmental conditions encountered in infant transports
also may prove beneficial (59,60).
Radiant Heat Loss

Radiant heat loss is the least intuitive aspect of newborn incubation. Radiant heat loss occurs as a transfer of
infrared electromagnetic energy virtually instantly from one warm body to another of lesser warmth:
in which a radiant heat transfer constant (Stefan-Boltzmann), the physical emissivity of the infant's skin and
incubator's plastic walls (skin, ewalls), the absolute temperature gradient (T), and the infant's exposed surface
area and posture determine the rate of radiant heat loss within incubators (37). In a single-walled incubator,
heat transfers from the infant's skin (36.5C to 37.0C) to the mean temperature of the cooler walls of the
incubator's plastic interior (approximately 28.0C to 36.5C). The incubator walls then reradiate heat to the
nursery walls and windows, which are even cooler (18.0C to 27.0C). A massive heat sink is represented by
the nursery walls, which may determine 60% of the operant environmental temperature perceived by the skin in
room air conditions:
Double-Walled Incubators
The importance of radiant heat transfer within incubators is demonstrated in Fig. 24-6 (40). Two incubators of
different designs are compared. Partitional calorimetry for a single-walled incubator is shown on the left,
whereas a double-walled incubator design is depicted on the right. The double-walled incubator constitutes a
plastic chamber similar to the single-walled incubator with an additional inner wall suspended several
centimeters interior to the outer wall of the incubator. Warmed air is circulated between these two incubator
walls, warming both the outer and inner surfaces of the inner wall, as well as the inner surface of the outer wall
of the incubator. The result is an elevated inner wall plastic temperature exposed to an infant's skin. Radiant
heat loss to the inner wall of the incubator exposed to the infant's skin is significantly reduced. Convective heat
loss is higher in the double-walled incubator because a cooler air temperature is
P.451
required to maintain the infant when radiant heat loss is thus conserved. Because vapor pressure is constant,
evaporative heat loss is the same in both incubator devices.
Figure 24-6 Partitional calorimetry performed for preterm infants at steady-state, thermal-neutral
temperature in two forced-air, convection-warmed incubators, either with only a single wall between the infant
and surrounding nursery's environmental air and wall temperatures, or with an inner wall (double-walled
incubator) interposed between the infant and incubator's outer wall. Forced-air warming of both sides of the
inner wall reduces radiant heat loss, resulting in lower servocontrolled incubator air temperature (higher infant
convective heat loss). Evaporative loss was not altered by insertion of the double wall. (Adapted from Bell EF,
Rios GR. A double-walled incubator alters the partition of body heat loss of premature infants. Pediatr Res
Skin Surface Servocontrol Temperature

As a result of the difficulties encountered in measuring and controlling the whole environment defined by all of
the parameters described previously, Silverman and associates (36) proposed that any set of environmental
conditions that rendered a normothermic skin surface temperature would provide a thermal-neutral
environment. They defined a skin temperature servocontrol set point between 36.2C to 36.5C over the
anterior abdominal wall for thermostatic control of incubator air temperature warming (with or without
humidification, mattress warming, or radiant walls protection) to guarantee a minimal observed metabolic rate
of oxygen consumption and, therefore, a thermal-neutral environment. Proportional response, servocontrolled
skin temperature now has become the standard of care for regulating convection-warmed incubator heating
over much of the world.
Servocontrolled Incubator Homeothermy

All incubator studies demonstrating a thermal-neutral environment under a variety of conditions evaluating air
temperature, convection, and humidity; incubator wall temperature; and nursery temperature were performed
at steady-state conditions. Infant metabolic rates often were determined in noncritical subjects at rest. Such
steady-state conditions are theoretical, however, in critically ill premature neonates in thermostatically
servocontrolled incubators, subjected to the many variations encountered in the modern neonatal intensive care
nursery. Hand ports and doors may be opened and closed several times a day for access to nursing and minor
surgical procedures, as well as radiographic, ultrasound, and echocardiographic examinations. Incubators may
be invaded as often as once an hour for delivery of intensive care. This can result in overshooting and
undershooting of air temperature for as long as 1 hour after the door has been closed.
THE RADIANT WARMER BED

The radiant warmer bed is a variable-level, waist-high platform with a mattress surface (which may be either
heated or may incorporate an electronic scale), upon which the critically ill newborn lies without an encumbering
plastic enclosure. Suspended about 80 to 90 cm above this platform is a radiant heat source with an electrically
heated metal alloy wire coiled within a quartz tube. A thermostatic skin servocontrol device regulates the radiant
heat, which is distributed across near and far wavelengths of the infrared spectrum (Fig. 24-7). At full power,
the sum of infrared
P.452
irradiance is less than 100 mW/cm2 of surface area on the bed platform. The heat delivered in the near portion
of the infrared spectrum (less than 1,000 nm) is less than 10 mW/cm2 of surface area irradiated. These levels of
radiant exposure recently have been judged as biologically safe for the developing skin and eyes in the
prematurely born newborn infant when exposed briefly. The servocontrolled warmer generally operates at about
one-half this peak power, and radiant power density delivered at bed level usually is well below the 50 to 60
mW/cm2 felt to be potentially deleterious at longer periods of exposure. The head, foot, and sides of the radiant
warmer bed platform receive less radiant power and, therefore, less warming than is delivered to the center of
the bed platform. Optimally, infants are positioned in the center of this device.
Figure 24-7 Spectral irradiance measured at bed level with increasing power from a radiant warmer's heating
element. Note two peaks of energy indicating the wire element's emission (near infrared wavelength spectrum
at the left) and the quartz containment tube's reemission of absorbed heat from the wire (far spectrum at the
right). The notch in the near peak probably indicates infrared adsorption by water vapor. Integrated over the
entire peak emissions, levels of near and far infrared exposure are felt to be safe for the developing skin and
eyes. (From Baumgart S, Knauth A, Casey FX, et al. Infrared eye injury not due to radiant warmer use in
premature neonates. Am J Dis Child 1993;147: 565-569, with permission.)
Figure 24-8 Partitional calorimetry for infants nurtured naked under radiant warmers servocontrolled to
maintain anterior ab-dominal wall skin temperature at 36.5C. Shown on the left are convective, evaporative,
conductive, and radiant heat losses (from the infant's sides), and on the right are radiant heat gain (facing the
warmer) and heat generated by the infant's metabolism. (Adapted from Baumgart S. Radiant heat loss versus
radiant heat gain in premature neonates under radiant warmers. Biol Neonate 1990;57: 10-20, with
permission.)
Figure 24-8 demonstrates experience with complete partitional calorimetry for critically ill premature infants
nurtured under radiant warmers (61). Both radiant heat losses under a radiant warmer as well as heat gains are
demonstrated. The effects of evaporation, convection, and conduction and of infant metabolic heat production
are demonstrated for infants nurtured naked and supine under these devices. Heat loss comprised 64% toward
convection to the surrounding cool air of the nursery's environment. The majority of convective heat loss
occurred naturally, whereas a minor component was composed of forced convective air movement from doors
opening and closing within the nursery, nursery personnel bustling near the bedside, as well as the cycling of
heating and cooling vents supplying the nursery's ambient air control. These turbulent convective air
movements contributed to evaporative heat loss, which comprised 30% of total heat loss in this figure. Thermal
equilibrium is maintained under a radiant warmer by replacement of heat losses through convection,
evaporation, conduction, and radiation, by radiant heat gain directly from the warming element. Almost 58% of
heat replacement is derived from the servocontrolled radiant warmer. Metabolic heat production M comprises
42% of the thermal balance and is overpowered by the radiant heating element.
Thermal-Neutral Radiant Warming

Le blanc (62) summarized a metanalysis comparing radiant warmers to convectively heated incubators. The
thermal-neutral environment was defined for each warming device as the minimal observed metabolic rate of
oxygen consumption (MOMR measured in oxygen consumption, [V with dot above]O2, in mL/kg/minute) when
skin temperature was servocontrolled between 36.0C and 36.5C and at steady-state conditions. Eleven of 16
infants demonstrated slightly higher rates of oxygen consumption when nursed at similar temperatures under
radiant warmers ([V with dot above]O2 6.84 0.37 [standard error of the mean] vs. 7.45 0.44 mL/kg/
minute), an increase of 8.8% in metabolic rate under radiant warmers.
A more recent metanalysis compiled by Flenady and Woodgate (63) again compared radiant warmers to
incubators using MOMR to define the thermal-neutral environment. This review found that nursing under radiant
warmers caused a statistically significant increase in insensible water loss and a nonsignificant trend toward
higher oxygen consumption when compared to incubators. When a heat shield was added to the radiant warmer
and compared to the incubator, there was a trend toward higher insensible water loss that was not statistically
significant and there was no difference in oxygen consumption.
The assumptions involved in the calculation of MOMR are that the infant is (a) at rest and asleep, (b)
postprandial at least 2 to 3 hours, and (c) at thermal-neutral temperature conditions. Typically premature
infants demonstrate slightly higher oxygen consumptions than term infants, and small-for-gestational-age
infants manifest oxygen consumptions slightly higher than premature and term infants. The minimum
requirement for oxygen consumption in these studies demonstrates a metabolic rate of caloric expenditure of
roughly 60 to 75 kcal/kg/day. Above this amount, 9 to 10 kcal/kg/day may be required for growth. However,
intensive care conditions provided to critically ill neonates rarely approximate those of a growing premature
baby at rest and at steady-state.
Figure 24-9 shows results of oxygen consumption measurements from studies conducted in an intensive care
nursery in a 1.2-kg infant nurtured naked and supine under a radiant warmer while intubated endotracheally
and receiving ventilatory support (64). During a 90-minute period of relatively mild cold stress (servocontrol
skin temperature 35.5C), oxygen consumption fluctuated
P.453
between 5.5 and 8.5 mL/kg/minute, paralleled closely by the infant's heart rate and behavior. These
observations of preterm infant behavior linked to metabolism are typical, even in critically ill infants. Criteria for
determining MOMR suggest that oxygen consumption for this subject is 5.5 mL/kg/minute. However, the
integrated sum of behavior over the entire study period reflects a higher rate of global metabolism. Such
observations led Schulze and associates (65) to speculate that a thermal-neutral environment should be
evaluated over periods considerably longer than 10 to 30 minutes of MOMR at steady state. Figure 24-10 shows
oxygen consumption over three 90-minute periods for the same infant as shown in Fig. 24-9 at three different
radiant warmer servocontrol skin temperatures35.5C, 36.5C, and 37.5C (66). From these studies over
longer study periods, it seems clear that infant behavior may constitute a significant part of metabolic rate
determination, affecting the thermal-neutral zone. Figure 24-11 shows 18 premature infants nurtured under
radiant warmers demonstrating a thermal-neutral environmental temperature (between 36.2C and 36.5C
servocontrolled anterior abdominal wall skin temperature) in which all infant behavior was incorporated (66).
Approximately 7.2 mL/kg/minute VO2 at the warmer's 36.5C skin temperature set point represented optimal
control under intensive care circumstances. Increasing anterior abdominal wall skin temperature above this
point by 1C resulted in no significant additional reduction in metabolic rate, and when servocontrolled to 37.5
C, the gradient for heat loss (from the infant's core to the skin) narrowed sufficiently for a number of them to
become hyperthermic (38.2C to 38.5C). It is therefore recommended to avoid skin control temperatures
above
P.454
36.7C and 37.0C for infants in the weight range studied (between 0.87 and 1.60 kg).
Figure 24-9 Metabolic rate of oxygen consumption (A) parallels heart rate (B) and infant activity (C) with the
performance of indirect calorimetry in a premature baby weighing approximately 1.2 kg. Variation with infant
behavior is demonstrated in this relatively cool condition (servocontrol 35.5C). (From Baumgart S. Partitioning
of heat losses and gains in premature newborn infants under radiant warmers. Pediatrics 1985;75:89-99, with
permission.)
Figure 24-10 The metabolic rates of oxygen consumption from the same infant shown in Fig. 24-9 are
compared at 35.5C, 36.5C, and 37.5C servocontrol skin temperature. Behavioral activity is attenuated, and
basal metabolism is significantly reduced at warmer temperatures. (From Malin SW, Baumgart S. Optimal
thermal management for low birth weight infants nursed under high-power radiant warmers. Pediatrics
Figure 24-11 Significantly lower oxygen consumption is demonstrated at 36.5C compared to 35.5C
servocontrol temperature for a series of 18 infants under radiant warmers. No additional significant reduction in
oxygen consumption is achieved by increasing servocontrol temperature to 37.5C, and several infants
demonstrated deep rectal hyperthermia at 37.5C. (From Malin SW, Baumgart S. Optimal thermal management
for low birth weight infants nursed under high-power radiant warmers. Pediatrics 1987;79:47-54, with
permission.)
HEAT SHIELDING
Rigid, Plastic Heat Shields
A 1- to 2-mm thickness of plastic used as a miniature incubator hood and placed over infants on open radiant
beds has been proposed by several authors. Yeh and associates (67) reported that insensible water loss was
reduced by more than 25% for infants nurtured under a plastic hood. Bell and associates (68) failed to replicate
any difference in water loss using a rigid plastic body hood under radiant warmers. The configuration of the
plastic hoods used in each of these studies was different, in some cases permitting free air exchange at the open
ends of the hood. Moreover, the interposition of radiantly opaque plastic between infant and the radiant element
may have interfered with the delivery of radiant heat to the baby's skin (69). Disruption of the servocontrol
mechanism by interposition of a radiant opaque plastic heat sink between the infant and the warmer seems to
be a futile strategy.
Occlusive Plastic Blankets or Bags

A flexible plastic blanket made of saran seems more efficient than the body hood, particularly under a radiant
warmer (Fig. 24-12) (39,70). Saran is thin, permitting free heat exchange without excessive evaporation (70).
The flexible plastic blanket reduces the effect of forced convection, which disrupts the microenvironment of
warm, humid air near the infant's skin (39). The reduction in both convective and evaporative heat losses is
about 30% (70). Insensible water loss is diminished from about 2.0 to 1.2 mL/kg/hour. Oxygen consumption is
reduced from 9.0 to 8.0 mL/kg/minute in this study. The net effect of saran blanket heat shielding under radiant
warmers renders the environment more homeothermic, with lower power required to achieve thermal
equilibrium, both from the radiant warmer and the baby's rate of metabolic heat production.
Figure 24-12 Comparison of radiant power density delivered (right), insensible water loss (middle), and
oxygen consumption (left) in preterm neonates nurtured either naked or covered by a saran plastic blanket
under radiant warmers. The blanket significantly reduced all three measurements in all babies tested,
suggesting better environmental maintenance under saran. SEM, standard error of the mean. (From Baumgart
S. Reduction of oxygen consumption, insensible water loss, and radiant heat demand with use of a plastic
blanket for low-birth-weight infants under radiant warmers. Pediatrics 1984;74:1022-1028, with permission.)
A polyethylene plastic body bag was proposed by Vohra and associates (71) for use during delivery room
resuscitations in premature neonates. Significantly higher core temperatures and survival was reported for
babies less than 28 weeks of gestation. A multicenter randomized trial of this device has been proposed by the
authors.
Adverse effects with plastic blankets or bags have been cited. In extremely low-birth-weight infants, the
immature skin may stick to the plastic, causing maceration. Moreover, if the servocontrol thermistor fails or
becomes detached, serious hyperthermia may result and deaths have been reported. Nursing management of
thermal care should prevent plastic from directly contacting the skin wherever possible. Infant core
temperatures (axillary or rectal) should also be monitored frequently without relying on the skin servoprobe.
Finally, infection control risk with plastic blankets/bags has not been adequately documented. With careful
regard to these complications, however, these techniques are presently the most effective under radiant
warmers for more consistent thermal-neutral temperature regulation.
EPIDERMAL BARRIER PROTECTION

Semiocclusive Artificial Skin
A novel strategy for thermal protection is covering the exposed surfaces of immature skin with semiocclusive
polyurethane dressings (Tegaderm and Opsite) (72). In early studies using this technique, insensible water loss
from days 1 to 4 of life was reduced by 30% to 50%. Upon careful removal of the artificial polyurethane
dressing, skin moisture barrier development (keratinization) was consistent with development over the adjacent
naked skin sites tested. A report by Porat and Brodsky (73) indicated that applying an adherent polyurethane
layer over the entire torso and extremities early in very-low-birth-weight infants under 800 g improved fluid and
electrolyte balance, reduced patent ductus arteriosus and intraventricular hemorrhage, as well as improved
survival. Mancini and associates (74) found a similar result and reported a decrease in bacterial growth on skin
guarded by a semipermeable dressing.
P.455
Emollients and Aquaphor

Rutter and Hull (76) originally reviewed topical agents for reduction of transepidermal water loss. They found
high water content creams ineffective in this regard, whereas paraffin produced a modest effect. Calculated
evaporative rate was initially halved in preterm infants less than 30 weeks of gestation; however, this effect was
transient, with return to a higher rate of water loss within 4 to 8 hours (75,76). Nopper and associates (77)
conducted a randomized trial investigating barrier effects of a topical, petroleum-based, preservative-free

ointment (Aquaphor) applied twice daily for 2 weeks to the skin of preterm infants. Transepidermal water loss
was decreased by 67% initially compared to control infants and remained at 34% below controls after 4 to 6
hours. Dermatologic scores of skin integrity were improved in treated infants, and there was a decline in
bacterial colonization of axillary skin, with a significantly lower incidence of positive blood and spinal fluid
cultures. The authors also suggested better temperature control and fluid balance in treated subjects.
Commentary by Rutter (76) on this work suggests, however, that temperature control may not benefit babies
weighing less than 1 kg early after birth, and only high humidity incubators eliminate this problem in the
smallest infants.
Aquaphor results early on were encouraging for improving skin care and reducing insensible water loss (77);
however, the findings of a multicenter randomized trial are still awaiting publication (Vermont Oxford Network
2001) (78). One preliminary study, however, implies that bacterial infection may be more likely (79). Until
complications of any shielding technique are better defined, nurses and doctors should remain cautious adopting
them for routine care of extremely low-birth-weight patients.
HYBRID INCUBATOR/RADIANT WARMER DESIGN

One new incubator design recently marketed is the hybrid design. The accessibility of a radiant warmer is
combined with the protection of an incubator enclosure. In the incubator mode, plastic walls surround the baby,
and temperature is regulated by convection air warming. In the radiant warmer mode, these plastic walls are
rotated down below the bed platform to facilitate handling with procedures. An overhead radiant heater switches
on automatically to regulate servocontrolled skin temperature. No plastic is positioned between the infant and
radiant warming element. Although expensive, these devices are an alternative to moving fragile infants
requiring procedural interventions.
FEVER
Generally, fever may be defined as a core body temperature greater than 37.5C. Depending on the limits
chosen, approximately 1% to 2.5% of all newborns admitted to the normal nursery are febrile by rectal or
axillary temperature (80). Although relatively more uncommon, the most life-threatening cause of fever in the
neonate is bacterial sepsis. Fever, however, is an unreliable marker for infection in the newborn as less than
10% of all febrile neonates have culture-proven sepsis, and many more infected infants demonstrate
normothermia or most commonly, hypothermia.
PHYSIOLOGIC CONTROL OF BODY TEMPERATURE

Central control of body temperature occurs at the level of the hypothalamus, where a set point temperature
defines normothermia. When body temperature falls below this set point, the hypothalamus initiates heat
conservation mechanisms that include peripheral vasoconstriction and nonshivering thermogenesis. If body
temperature rises above the set point, heat dissipation mechanisms are activated; in the neonate, these consist
mainly of peripheral vasodilation and to limited degree, sweating. It is thought that fever occurs when this set
point is displaced to a higher level.
MECHANISMS PRODUCING NEONATAL FEVER

Bacterial and viral toxins, like endotoxins of bacterial membranes, are examples of exogenous pyrogens that can
initiate an immunologic cascade resulting in fever (80). Exogenous pyrogens stimulate granulocytes and macrophages to produce endogenous pyrogen (interleukin-1). Endogenous pyrogen then activates phospholipase A2,
which ultimately results in the production of prosta-glandins. One of these prostaglandins, PGE2, is thought to
be one of the mediators that raises the hypothalamic set point producing the fever seen with infection.
In addition, exposure to excess heat or insulation (excessive swaddling) can quickly increase core temperature
of newborn infants due to their poor heat dissipation mechanisms, particularly the inability to effectively sweat.
Such overheating commonly occurs when term babies are nursed in uncontrolled incubators or under radiant
warmers. Temperature elevation may also occur with increased infant metabolic rate, as seen with skeletal
muscle rigidity and status epilepticus. Another cause of temperature elevation is occasionally observed in well,
breast-feeding newborn infants on the third to fourth day of life and is felt to result from dehydration due to
insufficient maternal milk production. Finally, there are more recent reports of an increased incidence of
maternal and neonatal fever when mothers receive epidural anesthesia (81).
P.456
REFERENCES
1. Cone TE. History of the care and feeding of the premature infant. Boston: Little, Brown & Co, 1985.
2. Berthod P. La couveuse et le gavage a la maternite de Paris [doctoral thesis]. Paris: G. Rougier, 1887.
3. Marx S. Incubation and incubators. Am Med Surg Bull 1896; 9:311-.
4. Rotch TM. Description of a new incubator. Arch Pediatr 1893;10: 661-665.
5. Swanson HE. Interrelations between thyroxin and adrenalin in the regulation of oxygen consumption in the
albino rat. Endocrinology 1956;59:217-225.
6. Blackfan KD, Yaglou CP. The premature infant. a study of effects of atmospheric conditions on growth and
on development. Am J Dis Child 1933;46:1175-1236.
7. Bolt RA. The mortalities of infancy. In: Abt I-A, ed. Pediatrics, 1st ed, Philadelphia: WB Saunders, 1923.
8. Mauriceau F. Traite des maladies des femmes grosses et accouchees. Paris: Chez l'Auteur, 1669:100-.
9. Silverman WA, Fertig JW, Berger AP. The influence of the thermal environment upon the survival of newly
born premature infants. Pediatrics 1958;22:876-886.
10. Cross KW, Dawes GS, Mott JC. Anoxia, oxygen consumption and cardial output in new-born lambs and
adult sheep. J Physiol 1959;146:316-343.
11. Hill JR, Rahimtulla KA. Heat balance and the metabolic rate of new-born babies in relation to
environmental temperature, and the effect of age and of weight on basal metabolic rate. J Physiol
1965;180:239-265.
12. Hill JR. The oxygen consumption of new-born and adult mammals. Its dependence on the oxygen tension
in the inspired air and on the environmental temperature. J Physiol 1959;149:346-373.
13. Brck K, Parmelee AH Jr, Brck M. Neutral temperature range and range of thermal comfort in
premature infants. Biol Neonate 1962;4:32-51.
14. Brck K. Temperature regulation in the newborn infant. Biol Neonate 1961;3:65-119.
15. Hey EN, Katz G. The optimum thermal environment for naked babies. Arch Dis Child 1970;45:328-334.
16. Hey EN. The relation between environmental temperature and oxygen consumption in the new-born baby.
J Physiol 1969;200: 589-603.
17. Hey E. Thermal neutrality. Br Med Bull 1975;31:69-74.
18. Power GG, Schroder H, Gilbert RD. Measurement of fetal heat production using differential calorimetry. J
Appl Physiol 1984;57: 917-922.
19. Ryser G, Jequier E. Study by direct calorimetry of thermal balance on the first day of life. Eur J Clin Invest
1972;2:176-187.
20. Morishima HO, Yeh MN, Niemann WH, et al. Temperature gradient between fetus and mother as an index
for assessing intrauterine fetal condition. Am J Obstet Gynecol 1977;129:443-448.
21. Power GG, Kawamura T, Dale PS, et al. Temperature responses following ventilation of the fetal sheep in
utero. J Dev Physiol 1986;8:477-484.
22. Schroder H, Gilbert RD, Power GG. Computer model of fetal-maternal heat exchange in sheep. J Appl
Physiol 1988;65:460- 468.
23. Hodgkin DD, Gilbert RD, Power GG. In vivo brown fat response to hypothermia and norepinephrine in the
ovine fetus. J Dev Physiol 1988;10:383-391.
24. Alexander G, Williams D. Shivering and non-shivering thermogenesis during summit metabolism in young
lambs. J Physiol 1968;198:251-276.
25. Bray GA, Goodman HM. Studies on the early effects of thyroid hormones. Endocrinology 1965;76:323-328.
26. Klein AH, Reviczky A, Padbury JF. Thyroid hormones augment catecholamine-stimulated brown adipose
tissue thermogenesis in the ovine fetus. Endocrinology 1984;114:1065-1069.
27. Silva JE, Larsen PR. Adrenergic activation of triiodothyronine production in brown adipose tissue. Nature
1983;305:712-713.
28. Baumgart S. Incubation of the human newborn infant. In: Pomerance JJ, Richardson CJ, eds. Neonatology
for the clinician. Norwalk, Conn: Appleton & Lange, 1993:139-150.
29. Baum JD, Scopes JW. The silver swaddler. Device for preventing hypothermia in the newborn. Lancet
1968;1:672-673.
30. Besch NJ, Perlstein PH, Edwards NK, et al. The transparent baby bag. A shield against heat loss. N Engl J
Med 1971;284:121-124.
31. Dahm LS, James LS. Newborn temperature and calculated heat loss in the delivery room. Pediatrics
1972;49:504-513.
32. Glass L, Silverman WA, Sinclair JC. Effect of the thermal environment on cold resistance and growth of
small infants after the first week of life. Pediatrics 1968;41:1033-1046.
33. Charpak N, Ruiz-Pelaez JG, de Figueroa CZ, et al. Kangaroo mother versus traditional care for newborn
infants 2000 grams: a randomized, controlled trial. Pediatrics 1997;100:682-688.
34. Bell RP, McGrath JM. Implementing a research-based kangaroo care program in the NICU. Nurs Clin North
Am 1996;31:387-403.
35. Anderson GC. Current knowledge about skin-to-skin (kangaroo) care for preterm infants. J Perinatol
1991;11:216-226.
36. Silverman WA, Sinclair JC, Agate FJ Jr. The oxygen cost of minor changes in heat balance of small
newborn infants. Acta Paediatr Scand 1966;55:294-300.
37. Sinclair JC. Metabolic rate and temperature control. In: Smith CA, Nelson NM, eds. The physiology of the
newborn infant. Springfield, Ill: Thomas, 1976:354-415.
38. Wheldon AE, Rutter N. The heat balance of small babies nursed in incubators and under radiant warmers.
Early Hum Dev 1982;6: 131-143.
39. Baumgart S, Engle WD, Fox WW, et al. Effect of heat shielding on convective and evaporative heat losses
and on radiant heat transfer in the premature infant. J Pediatr 1981;99:948-956.
40. Bell EF, Rios GR. A double-walled incubator alters the partition of body heat loss of premature infants.
Pediatr Res 1983;17:135-140.
41. Okken A, Blijham C, Franz W, et al. Effects of forced convection of heated air on insensible water loss and
heat loss in preterm infants in incubators. J Pediatr 1982;101:108-112.
42. Baumgart S, Engle WD, Fox WW, et al. Radiant warmer power and body size as determinants of insensible
water loss in the critically ill neonate. Pediatr Res 1981;15:1495-1499.
43. Baumgart S, Langman CB, Sosulski R, et al. Fluid, electrolyte, and glucose maintenance in the very low
birthweight infant. Clin Pediatr (Phila) 1982;21:199-206.
44. Bell EF, Neidich GA, Cashore WJ, et al. Combined effect of radiant warmer and phototherapy on insensible
water loss in low-birth-weight infants. J Pediatr 1979;94:810-813.
45. Hammarlund K, Sedin G, Stromberg B. Transepidermal water loss in newborn infants. VIII. Relation to
gestational age and post-natal age in appropriate and small for gestational age infants. Acta Paediatr Scand
1983;72:721-728.
46. Hey EN, Katz G. Evaporative water loss in the new-born baby. J Physiol 1969;200:605-619.
47. Sedin G, Hammarlund K, Nilsson GE, et al. Measurements of transepidermal water loss in newborn
infants. Clin Perinatol 1985;12:79-99.
48. Williams PR, Oh W. Effects of radiant warmer on insensible water loss in newborn infants. Am J Dis Child
1974;128:511-514.
49. Wu PY, Hodgman JE. Insensible water loss in preterm infants: changes with postnatal development and
non-ionizing radiant energy. Pediatrics 1974;54:704-712.
50. Costarino AT, Baumgart S. Neonatal water metabolism. In: Cowett RM, ed. Principles of perinatal-neonatal
metabolism. New York: Springer-Verlag, 2nd ed., 1998;1045-1075.
51. Costarino AT, Baumgart S. Modern fluid and electrolyte management of the critically ill premature infant.
Pediatr Clin North Am 1986;33:1-53.
52. Harpin VA, Rutter N. Humidification of incubators. Arch Dis Child 1985;60:219-224.
53. Moffet HL, Allan D, Williams T. Survival and dissemination of bacteria in nebulizers and incubators. Am J
Dis Child 1967;114: 13-20.
54. Moffet HL, Allan D. Colonization of infants exposed to bacterially contaminated mists. Am J Dis Child
1967;114:21-25.
55. Brown DG, Baublis J. Reservoirs of pseudomonas in an intensive care unit for newborn infants:
mechanism of control. J Pediatr 1977;90:453-457.
56. Hoffman MA, Finberg L. Pseudomonas infections in infants associated with high-humidity environments. J
Pediatr 1955;46: 626- 630.
57. American Academy of Pediatrics and the American College of Obstetricians and Gynecologists. Guidelines
for perinatal care, 2nd ed. Elk Grove Village, Ill and Washington, DC: AAP and ACOG, 2002:.
P.457
58. Topper WH, Stewart TP. Thermal support for the very-low-birth-weight infant: role of supplemental
conductive heat. J Pediatr 1984;105:810-814.
59. Koch J. Physical properties of the thermal environment. In: Okken A, Koch J, eds. Thermal regulation of
sick and low birth weight neonates. Berlin: Springer-Verlag, 1995:103.
60. Sedin G. Physics of neonatal heat transfer, routes of heat loss and heat gain. In: Okken A, Koch J, eds.
Thermal regulation of sick and low birth weight neonates. Berlin: Springer-Verlag, 1995:21-.
61. Baumgart S. Radiant heat loss versus radiant heat gain in premature neonates under radiant warmers.
Biol Neonate 1990;57: 10-20.
62. Leblanc MH. Relative efficacy of radiant and convective heat in incubators in producing thermoneutrality
for the premature. Pediatr Res 1984;18:425-428.
63. Flenady VJ, Woodgate PG. Radiant warmers versus incubators for regulating body temperature in newborn
infants. Cochrane Database Syst Rev 2003;4:CD000435.
64. Baumgart S. Partitioning of heat losses and gains in premature newborn infants under radiant warmers.
65. Schulze K, Kairam R, Stefanski M, et al. Spontaneous variability in minute ventilation oxygen consumption
and heart rate of low birth weight infants. Pediatr Res 1981;15:1111-1116.
66. Malin SW, Baumgart S. Optimal thermal management for low birth weight infants nursed under highpower radiant warmers. Pediatrics 1987;79:47-54.
67. Yeh TF, Amma P, Lilien LD, et al. Reduction of insensible water loss in premature infants under the radiant
warmer. J Pediatr 1979;94:651-653.
68. Bell EF, Weinstein MR, Oh W. Heat balance in premature infants: comparative effects of convectively
heated incubator and radiant warmer, with and without plastic heat shield. J Pediatr 1980;96: 460-465.
69. Baumgart S, Fox WW, Polin RA. Physiologic implications of two different heat shields for infants under
radiant warmers. J Pediatr 1982;100:787-790.
70. Baumgart S. Reduction of oxygen consumption, insensible water loss, and radiant heat demand with use
of a plastic blanket for low-birth-weight infants under radiant warmers. Pediatrics 1984;74:1022-1028.
71. Vohra S, Frent G, Campbell V, et al. Effect of polyethylene occlusive skin wrapping on heat loss in very
low birth weight infants at delivery: a randomized trial. J Pediatr 1999;134:547-551.
72. Knauth A, Gordin M, McNelis W, et al. Semipermeable polyure-thane membrane as an artificial skin for the
premature neonate. Pediatrics 1989;83:945-950.
73. Porat R, Brodsky N. Effect of Tegaderm use on outcome of extremely low birth weight (ELBW) infants.
Pediatr Res 1993;33: 231(A).
74. Mancini AJ, Sookdeo-Drost S, Madison KC, et al. Semipermeable dressings improve epidermal barrier
function in premature infants. Pediatr Res 1994;36:306-314.
75. Rutter N, Hull D. Reduction of skin water loss in the newborn. I. Effect of applying topical agents. Arch Dis
Child 1981;56: 669- 672.
76. Rutter N. Waterproofing our infants. J Pediatr 1997;130:333-334.
77. Nopper AJ, Horii KA, Sookdeo-Drost S, et al. Topical ointment therapy benefits premature infants. J
Pediatr 1996;128:660-669.
78. Edwards WH, Conner JM, Soll RF for the Vermont Oxford Network. The effect of Aquaphor original
emollient ointment on nosocomial sepsis rates and skin integrity in infants of birth weight 501 to 1000 grams.
Pediatr Res 2001;49:388A (abstract).
79. Edwards WH, Conner JM, Soll RF. Vermont Oxford Network Neonatal Skin Care Study Group. The effect of
prophylactic ointment therapy on nosocomial sepsis rates and skin integrity in infants with birth weights of
501-1000 g. Pediatr 2004; 113: 1195-1203.
80. Craig WS. The early detection of pyrexia in the newborn. Arch Dis Child 1963;38:29-39.
81. Lieberman E, Lang JM, Frigoletto F Jr, et al. Epidural analgesia, intrapartum fever, and neonatal sepsis
evaluation. Pediatrics 1997;99:415-419.
Chapter 25
The Extremely Low-Birth-Weight Infant
Apostolos Papageorgiou
Ermelinda Pelausa
Lajos Kovacs
Tremendous progress in the survival of newborn infants has resulted in the need for population descriptors other
than premature or low birth weight. Such descriptors require subdivision by both gestational ages and birth
weights to have meaningful diagnostic and prognostic value.
In the 1960s, the term low birth weight (LBW) defined all infants with a birth weight less than 2,500 g. With
improved survival of infants born weighing less than 1,500 g in the 1970s and 1980s, the term very low birth
weight (VLBW) was introduced to better express the problems and outcomes unique to this group of infants.
In the 1990s, it became clear that a new category was necessary to reflect the prevailing reality, namely, the
large number of surviving infants born weighing less than 1,000 g. Thus, the term extremely low birth
weight (ELBW) was added to identify these infants.
In the start of the millennium, we have witnessed the survival of a new cohort of infants who weigh less than
500 grams at birth, i.e., below the weight limit which the World Health Organization had designated for
reporting live births. These infants, referred to by some authors as fetal infants or micropremies, are rarely
treated in our neonatal units but, nevertheless, reflect a new reality of modern Neonatology. Although their care
showcases the tremendous clinical and technological progress achieved in recent years, their survival has also
brought a substantial additional demand on human and financial resources, and major ethical dilemmas. Most of
the fetal infants who do survive are small for gestational age (SGA) and, at this point in time, their long-term
prognosis is not reassuring, which makes their neonatal intensive care unit (NICU) management a hotly
contentious issue (1,2,3).
In recent years, few medical specialties have demonstrated as much progress and success as has neonatology.
With regionalization of perinatal care, improved technology, and better understanding of their pathophysiology
and specific needs, the survival of ELBW infants has improved dramatically (4,5,6,7,8,9,10,11,12). In fact, in
most well organized perinatal centers in North America and Europe, neonatal deaths are uncommon for infants
with birth weights more than 1,000 g, in the absence of congenital anomalies. Recent reports demonstrate the
improvement in overall perinatal and neonatal mortality and the increasing survival of VLBW and ELBW infants
over time. Joseph and colleagues describe the reduction in infant mortality rates in Canada between the years
1985 to 1987 and 1992 to 1994, with the magnitude of reductions ranging from 14% (95% CI, 7-21) at 24 to
25 weeks gestation to 40% (95% CI, 31-47) at 28 to 31 weeks (13). Data in Table 25-1 reflect the progress as
experienced in our own perinatal center over the last 15 years for infants born weighing less than 1,000 g. Our
own perinatal mortality statistics at the Sir Mortimer B. Davis-Jewish General Hospital, a McGill University
tertiary care perinatal referral center with near 4,000 deliveries per year and a catchment area of 12,000
deliveries, show a neonatal mortality consistently between 0.3 and 0.5 per 1,000 live births for infants weighing
more than 1,000 g, including those who died from lethal congenital anomalies.
Thus, the care of VLBW infants, and particularly of ELBW infants, occupies an important part of the daily
activities of all NICUs and contributes heavily to the cost of neonatal care (14,15,16,17).
As mortality has much decreased, concerns have been expressed regarding whether morbidity has followed the
same rate of improvement (12,16,18,19,20,21,22,23,24). There is current evidence that for infants born
weighing more than 750 g, the decline in morbidity is a significant one, although not parallel to mortality.
However, for infants with birth weights less than 750 g, their long-term prognosis remains less favorable.
Although the incidence of cerebral palsy and other physical impairments is relatively low, the incidence of laterappearing cerebral dysfunction is quite elevated, with requirement for additional resources to manage behavioral
and school difficulties in later childhood.
TABLE 25-1 DECLINE IN MORTALITY BY BIRTHWEIGHT AT THE JEWISH GENERAL HOSPITAL, McGILL
UNIVERSITY, FROM 1984 TO 2002
19841985
20012002
Birth Weight(g)
Births
Mortality
Births
Mortality
Percent Improvement
500750
7511,000
5001,000
25
36
61
14 (56.0%)
10 (27.8%)
24 (39.3%)
41
40
81
18 (43.9%)
4 (10.0%)
22 (27.2%)
22
64
31
P.460
Major problems related to ELBW infants are listed in Table 25-2.
The aim of this chapter is to present a global approach to the care of ELBW infants, with emphasis on the
problems and management issues particular to them. The reader is referred to the specific chapters in the
textbook for a more comprehensive review of each problem.
Much of what is written in this chapter is based on our own experience in the management of ELBW infants, with
appropriate reference to the most recent published data. We expect that our experience may be different from
many in other parts of the world. It is important to appreciate that the Canadian health care system, which
provides universal access to health care, emphasizes prevention and has a very successful antenatal referral
policy, with the vast majority of VLBW infants being inborn. Table 25-3 indicates the number of infants weighing
less than 1,500 g born in the five level III maternity hospitals in the province of Quebec in comparison with the
number born in levels II and I (25). The success of the regionalization of perinatal care can be better
appreciated by knowing the size of the province (approximately four times the size of France) and the weather
conditions.
TABLE 25-2 MAJOR PROBLEMS IN EXTREMELY-LOW-BIRTH-WEIGHT INFANTS
Respiratory
Respiratory failure
Apnea
Air leaks
Chronic lung disease
Cardiovascular
Central nervous system
Intraventricular hemorrhage
Periventricular leukomalacia
Seizures
Renal
Electrolyte imbalance
Acidbase disturbances
Renal failure
Ophthalmologic
Retinopathy of prematurity
Strabismus
Myopia
Gastrointestinalnutritional
Feeding intolerance
Inguinal hernias
Cholostatic jaundice
Postnatal growth retardation

Immunologic
Poor defense to infection
EPIDEMIOLOGY
Until recently, statistics on ELBW infants were analyzed exclusively by birth weight. Although this method offers
the advantage of an objective measurement, it does not take into account the effect of gestational age (26). In
other words, many infants born weighing less than 1,000 g are more mature than their birth weight may
indicate, hence denoting the problems of intrauterine growth restriction (IUGR) superimposed on prematurity.
The neonatal problems and long-term prognosis can be quite different for the more mature, SGA infant than for
the less mature, appropriate for gestational age (AGA) infant of the same birth weight. The distinction between
AGA and SGA infants born before 28 weeks of gestation became possible only in recent years, thanks to the
introduction of early pregnancy ultrasonography. In Canada, and particularly in the Province
P.461
of Quebec, systematic ultrasonography between 16 and 18 weeks of gestation has permitted not only the early
detection, and the potential for termination of major congenital anomalies but, at the same time, it has provided
a reasonably accurate dating of almost all pregnancies. Precise gestational age assignment, along with the birth
weight of premature infants, has made it possible to relate specific problems, diagnoses and prognoses to the
degree of immaturity, and to recognize the implications of IUGR at a very early gestational age (26). In our
perinatal center, in the last 6 years, the incidence of IUGR, defined as a birth weight beyond two standard
deviations (2SDs) below the mean for a given gestational age, has been 24.5% for infants born weighing less
than 1,000 g (Table 25-4). It is hoped that as gestational age dating becomes universal and more accurate, the
current method of reporting perinatal statistics based on birth weight will be complemented by the gestational
age, thus reflecting both the degree of maturity and the degree of appropriateness of intrauterine growth.
Hence, neonatal pathology and prognosis can be based on both gestational age and birth weight.
TABLE 25-3 LIVE BIRTHS ACCORDING TO THE LEVEL OF HOSPITAL CARE IN THE PROVINCE OF
QUBEC, 1998
Level of Care
500999 g
n = 244
%
I
II
III
7
35
202
2.9
14.3
82.8
1,0001,499 g
n = 345
%
3
37
305
0.9
10.7
88.4
TABLE 25-4 IMPACT OF BIRTH WEIGHT ON OUTCOME IN THE CANADIAN COLLABORATIVE STUDY
Gestational Age
Weight (g)
Mortality
24 wk
<700
63.3%
n = 241
>700
25 wk
<760
n = 364
>760
From SB Effer, unpublished data.
37.2%
43.3%
35.9%
Although mortality rates of VLBW and ELBW infants are declining, the incidence of these births has not changed
significantly. In the province of Quebec, the rate of live births for infants weighing 500 to 999 g has increased
slightly from 0.3 in 1992 to 0.4% in 1998 and for those weighing 1,000 to 1,500 g from 0.4 to 0.5%. Similarly,
the incidence of births less than 26 weeks increased from 0.1 to 0.2% and that of births between 26 and 28
weeks from 0.4 to 0.5% (25).
Factors that have long been recognized as being associated with prematurity include extremes of maternal age,
socioeconomic status, low level of education, adverse social habits, maternal diseases, gynecologic infections
and, more recently, multiple pregnancies secondary to in vitro fertilization (27).
Significant predictors for the survival of ELBW infants are older gestational age, heavier birth weight, female
gender, African American race, singleton birth, and the absence of severe fetal growth restriction (28). The
importance of birth weight for the survival of infants born at 24 and 25 weeks of gestation has been
demonstrated clearly in a multicenter study of Canadian tertiary care centers (Table 25-5). In this particular
study, all infants were inborn, and the gestational ages were confirmed by early ultrasonography. Likewise,
maturity by only a few days has been shown to add significant chances of survival, as shown in Table 25-6.
Whether analysis is done by increments of 100 g or by increments of a few days, the impact of these two factors
on the survival of ELBW infants is very important. Tables 25-4 and 25-7 indicate the survival rate of infants born
weighing less than 1,000 g in our institution between 1995 and 2002, analyzed by weight and gestational age.
In our experience, infants born before 27 weeks of gestation with a birth weight beyond 2SD below the mean
are at a disadvantage compared to appropriate-for-gestational-age infants of the same gestational age in terms
of acute and chronic problems, the most striking complication being the higher incidence of retinopathy of
prematurity (ROP) (26).
TABLE 25-5 IMPACT OF GESTATIONAL AGE ON SURVIVAL OF 533 INFANTS AGED 24 TO 25 WEEKS IN
THE CANADIAN COLLABORATIVE STUDY
Gestational
Age (d)
No. of
N.B.
Neonatal
Mortality (%)
168171
172176
177181
125
177
171
55.2
47.4
34.5
From SB Effer et al., unpublished data.
In terms of global epidemiologic evaluation of outcomes for ELBW infants, many factors contribute to the
inaccuracy of data. A number of countries, and particularly some developing ones, do not keep statistics for
infants born before 28 weeks of gestation. In other countries,
P.462
when death occurs rapidly in the first day of life, the death is not recorded as a neonatal death. Also,
information originating from small private institutions may be inaccurate and difficult to control. National and
regional data can also be seriously affected by the ratio of inborn to outborn infants and the number of
extremely premature infants who are resuscitated. Indeed, great variations do exist in terms of intervention and
resuscitation in the delivery room between institutions and countries, and they reflect not only differences in the
capacity of some institutions to manage newborns near the limits of viability, but also differences in philosophy.
Tables 25-4, 25-7 to 25-9 indicate the survival, management, and complications of infants weighing less than
1,000 g born in our center over a 8-year period (1995-2002).
TABLE 25-6 POPULATION PROFILE OF INFANTS WITH BW <1,000 g BORN IN 19952002, N = 369
Survivors
Birth Weight (g)
Live Births # infants
< 500
500750
751999
500999
< 1000
17
151
201
352
369
Gestational age
Birth weight
Apgar 1 min.
Apgar 5 min.
SGA rate
C/Section rate
Days in hospital
(for survivors)
26 1.3 wks
758 142 g
4.3 2.2
6.5 1.9
24.6%
47.7%
88 36
mean, 1 s.d.
# infants
4
87
166
253
257
(23.5%)
(57.6%)
(82.6%)
(71.9%)
(69.6%)
TABLE 25-7 SURVIVAL RATE BY GESTATIONAL AGE OF 352 INBORN INFANTS WEIGHING 5001,000
gAT THE SMBD-JEWISH GENERAL HOSPITAL, McGILL UNIVERSITY, FROM 1995 TO 2002
Gestational Age (wk)
Total Births
Survivors
< 23
2324
2526
10
105
129
2 (20.0%)
50 (47.6%)
98 (76.0%)
2728
2930
3132
All ages
76
25
7
352
73 (96.1%)
23 (92.0%)
7 (100%)
253 (71.9%)
TABLE 25-8 OUTCOME OF 352 INBORN INFANTS WEIGHING 5001,000 gAT THE SMBD-JEWISH
GENERAL HOSPITAL, McGILL UNIVERSITY, 19952002
No. of Infants(n = 352)
Survived
Antenatal betamethasone
Cesarean section
Oxygen for at least 24h
253
265
168
318
71.9
75.3
47.7
90.3
Ventilation
296
239
84.1
67.9
197/239
19
51
86
27
220
9
82.4
5.4
14.5
24.4
7.7
62.5
2.6
249
70.7
Surfactant for respiratory distress syndrome

Drained pneumothorax
Pulmonary interstitial emphysema only
Intraventricular hemorrhage all grades
Grades IIIIV
Surgical necrotizing enterocolitis
Apnea
TABLE 25-9 COMPLICATIONS AMONG 253 SURVIVORS 5001,000 gAT THE SMBD-JEWISH GENERAL
HOSPITAL, McGILL UNIVERSITY, FROM 19952002
Oxygen 28 d
Oxygen 36 wks PCA
Ventilation
IVH all grades
IVH grades IIIIV
Periventricular leukomalacia
Ventriculomegaly
Retinopathy of prematurity all stages
stage III
Threshold
Laser
No. of Infants 253
161
79
193
42
12
63.6
31.2
76.3
16.6
4.7
14
82
5.5
32.4
159
49
15
62.8
19.4
5.9
15
5.9
162
64.0
Closure with indomethacin

Surgery
Sepsis
Surgical necrotizing enterocolitis
Home O2
135/162
27/162
70
4
15
Days in hospital
88 36
83.3
16.7
27.7
1.6
5.9
36
PERINATAL MANAGEMENT
Prenatal
With the advent of routine early ultrasonography, the gestational age is fairly well established on admission to
the Obstetrics unit for the vast majority of women presenting with premature labor, premature rupture of
membranes or other problems diagnosed in the second trimester of pregnancy. Such patients need to be
immediately placed in the charge of a specialist in high risk Obstetrics (Obstetrical Perinatologist or Specialist in
Maternal Fetal Medicine) to coordinate the evaluation and management plans and to ensure appropriate
communication and counseling. Based on the investigation for the causes of the problem at hand, the evaluation
of the degree of cervical dilatation, the condition of the membranes, the presence or absence of chorioamnionitis
and, if possible, the most recent evaluation of fetal well-being by bedside or formal ultrasonography, the
Perinatologist can decide on the best course to be taken, such as an estimate of the likelihood of controlling
labor with tocolysis to allow adequate time for antenatal corticosteroid therapy (7,29), often with consultation
with their colleagues in neonatology.
The prospective parents should receive accurate information regarding all facets of the proposed management,
including the possible need for cesarean delivery, and information regarding the subsequent management of the
newborn infant, including the potential risks related to both the degree of prematurity and the therapeutic
interventions that may be necessary to keep the infant alive.
P.463
Ideally, this information should be provided conjointly by both the obstetric perinatologist and the neonatologist
and should be based not only on general statistical information, but also on the specific institutional experience
with outcomes of newborn infants of similar gestational age. In our center, the attending neonatologist provides
a written consultation on all patients admitted to the obstetric high-risk unit. We meet with the family, offer an
extensive review of our experience with similar cases, and answer their questions regarding risks and outcomes.
The father and mother are invited to visit the NICU and to familiarize themselves with the environment and the
personnel. If the mother is not able to visit for medical reasons, we show her a book with pictures explaining
each step of the baby's treatments, from the delivery room to the time of discharge.
The lowest gestational age at which resuscitation should be initiated has long been the subject of debate.
Guidelines are available from both the American and the Canadian Fetus and Newborn Committees (30,31).
Based on our experience, we offer an optimistic opinion in terms of survival and potential morbidity for
pregnancies of 25 weeks' gestation and over. Between 24 and 25 weeks, although we underline that the chances
of survival are quite good, we also emphasize the increased risk of potential complications, such as
intraventricular hemorrhage (IVH), periventricular leukomalacia (PVL), retinopathy of prematurity (ROP), chronic
lung disease (CLD), neurosensory impairments, and later school and behavioral difficulties. For pregnancies
between 23 and 24 weeks of gestation, we describe the higher incidence of complications previously mentioned,
and the lower survival rate; however we also mention the possibility of intact survival or survival with minimal
handicaps. Finally, for pregnancies below 23 weeks of gestation, we do not recommend intervention. For parents
who request full intervention, we strongly advise that resuscitation will be undertaken only if the newborn has at
least the degree of maturity predicted by dates and/or ultrasonography, and if, in the judgment of the
neonatologist present in the delivery room, the newborn has reasonable chances of responding to resuscitation.
We always make it clear to the parents that initiation of resuscitation and subsequent treatments in the NICU do
not preclude discontinuation of therapy if a major complication such as severe IVH is detected in the hours or
days following birth. The presence of a staff neonatologist in the delivery room is an integral part of our protocol
for the management of ELBW infants.
One of the most difficult questions that parents ask, and which our obstetric colleagues continuously debate, is
the safest route of delivery in the presence of either a breech presentation or evidence of fetal distress (32). In
our institution, based on our own results, we advise cesarean delivery in such situations as of 25 weeks'
gestation. Between 24 and 25 weeks, we feel less inclined to recommend cesarean delivery, particularly in view
of the fact that many may require a classical incision. The decision to proceed with such an intervention is taken
with a clear understanding by the parents of all the medical implications for both the mother and the infant.
Finally, at less than 24 weeks of gestation, cesarean delivery is performed strictly for maternal indications. Our
incidence of cesarean section by gestational age is indicated in Table 25-10. It is obvious that a large number of
cesarean deliveries, and particularly those between 22 and 24 weeks of gestation, are performed strictly for
maternal indications, i.e., severe abruption, preeclampsia, etc.
TABLE 25-10 CESAREAN SECTION RATE IN 352 INFANTS WEIGHING 5001,000 gAT THE SMBDJEWISH GENERAL HOSPITAL, McGILL UNIVERSITY, 19952002
GA by Wks
No. Deliveries
No. C/Sections
<23
2324
2526
10
105
129
0
31
71
0
29.5
55.0
2728
2930
3132
All ages
76
25
7
352
39
20
7
168
51.3
80.0
100
47.7
Another difficult management situation relates to ruptured membranes between 18 and 22 weeks of gestation,
resulting in severe oligohydramnios, with the inherent risk of lung underdevelopment (33,34). Serial ultrasound
studies can evaluate the degree of reaccumulation of amniotic fluid and allow for a better-educated decision
regarding whether continuation of pregnancy is advisable (35). However, in the vast majority of these cases, the
outcome is very poor, and termination of pregnancy constitutes reasonable advice, particularly if rupture of
membranes occurred before 20 weeks of gestation with poor reaccumulation of amniotic fluid. Amnioinfusion has
been proposed and attempted as a means of overcoming the problem of chest compression, with limited success
thus far (36).
Impending Delivery
The management of a patient with impending premature delivery should include the following: evaluation of
gestational age by dates and/or early ultrasound, fetal size and position, condition of the fetal membranes,
amniotic fluid volume, and evidence of chorioamnionitis and other obstetric complications such as bleeding,
toxemia, and so on. Vaginorectal cultures for the detection of group B streptococcal colonization and initiation of
therapy with penicillin are also in order (37). If culture results return negative, penicillin can be discontinued. In
all patients from 24 weeks of gestation who are not infected and for whom there is no maternal indication for
immediate delivery, such as massive bleeding, and in whom the cervix is
P.464
dilated less than 5 cm, we propose tocolysis with magnesium sulfate and administration of betamethasone
(29,38). Our experience over the years with the combination of tocolysis and betamethasone has been fully
supported by the 1994 NIH Statement on Antenatal Use of Corticosteroids (39). For patients between 24 and 34
weeks' gestation, we administer two doses of 12 mg of betamethasone, 24 hours apart. Beyond 34 weeks, we
use steroids only when an amniocentesis indicates lung immaturity. Multiple pregnancies are offered similar
therapy (40). We also monitor body temperature and changes in leukocyte count, keeping in mind the possible
transient leukocytosis after the administration of betamethasone.
If a patient has fever or demonstrates other signs of chorioamnionitis, broad-spectrum antibiotics are initiated.
In the presence of ruptured membranes, a number of obstetricians use a combination of ampicillin and
erythromycin in an attempt to temporarily stop labor and administer steroids (41,42,43). This seems a
reasonable approach, because between 30% and 50% of premature births are believed to be precipitated by
common genital tract infections.
Delivery Room Management

The successful management of the ELBW infant begins in the delivery room (Table 25-11). A well-organized and
equipped delivery room and the presence of a competent team headed by an experienced neonatologist are
essential ingredients to the proper reception of these very fragile newborns. The basic principle guiding
successful management is directed toward prevention of any physiologic deviation from normality, such as
hypothermia, acidosis, or hypoxia. At the same time, it is important that each intervention during the
resuscitation process be carefully adapted to the size and to the needs of the tiny infant. Brisk maneuvers,
excessive positive pressure with bagging, or inappropriate administration of drugs and fluids may induce
permanent central nervous system (CNS) or lung injuries.
TABLE 25-11 THE FIRST 60 MINUTES OF LIFE
1. Expert resuscitation in the delivery room.
2. Good thermoregulation.
a. Keep the infant warm and dry in the delivery room
b. Provide high-humidity environment in the incubator
3. Minimum handling and avoidance of brisk maneuvers.
4. Expert cardiorespiratory support.
a. Liberal use of nasopharyngeal CPAP.
b. Intubation when indicated, avoiding excessive ventilatory pressures.
c. Continuous monitoring of oxygenation with pulse oximetry.
d. Monitoring of blood pressure. Prudent administration of volume expanders.
e. Catheterization of umbilical vessels, when indicated.
f. Close monitoring of blood gases, Hb, WBC + diff, blood glucose.
g. Radiographic evaluation of lung pathology and position of catheters.
h. Administration of surfactant, when indicated. Rapid adjustment of ventilatory support.
5. Intravenous DIOW and antibiotics when indicated.
6. Parental information
It seems particularly inappropriate when high-risk mothers are referred to a tertiary care center for specialized
perinatal care, to have their premature newborn infants cared for in the delivery room and during the critical
first hours of their lives by unsupervised and inexperienced in-training personnel. Major decisions, such as
whether to initiate resuscitation and for how long, often need to be made in extremely short periods of time and
under heavy pressure for infants at the limit of viability. This can be done only by experienced personnel (44).
In our center, the birth of an ELBW infant is always attended by a neonatologist in addition to the pediatric
house staff and a trained delivery room nurse. Appropriate equipment is used according to the American Heart
Association and American Academy of Pediatrics (AAP) guidelines for neonatal resuscitation, with particular
emphasis on temperature control i.e., radiant heater set at maximum temperature and prewarmed blankets
(45).
During the initial steps of stabilization, the condition of the infant is rapidly assessed. After drying, positioning on
warm blankets, and suctioning, most ELBW infants require immediate initiation of positive pressure ventilation
with a bag and mask. Resuscitation is initiated using an inspired oxygen concentration of 100%, which is rapidly
reduced as the infant's condition improves. Although concerns have recently been raised that such a practice
may result in potential exposure to oxygen free radical species (46,47), the available data is limited to
asphyxiated newborn infants at term, and there is currently insufficient evidence to support a change in practice.
We found that, for ELBW infants, ventilation is more effective if performed at a higher ventilatory rate than for
the term infant. We use anesthesia bags and ventilate at a rate of 60 to 80 breaths per minute, adjusting the
pressure to provide adequate bilateral air entry and chest wall excursion. For extremely premature infants,
intubation in the delivery room may rapidly follow.
With proper ventilation, in our experience, rarely will an infant require chest compressions or epinephrine. The
prognosis of ELBW infants requiring this extent of resuscitation is very guarded, particularly if their birth weight
is below 750 g. Fluid resuscitation is reserved only for those infants in which significant blood loss is suspected.
Even following optimal resuscitation, the Apgar scores of ELBW infants rarely exceed 6 or 7 in view of their
decreased tone and reactivity, poor respiratory effort, and initially poor peripheral perfusion (48). The infant's
heart rate is thus the best measure of the effectiveness of resuscitation efforts.
The topic of delivery room management would not be properly covered without mentioning the ethical dilemmas
faced by the neonatologist when parental and
P.465
medical opinions regarding resuscitation differ, or when an ELBW infant is severely asphyxiated and requires
prolonged resuscitation. It is our view that reasonable parental opinions must be respected after full and honest
discussion of the infant's chances of meaningful survival.
Admission to the Neonatal Intensive Care Unit

Expert management in the delivery room and during the first hours after admission to the NICU is of paramount
importance to prevent immediate and long-term complications in the ELBW infant. It is well established that the
majority of cerebral injuries occur around the time of delivery or in the immediate postnatal period. Acute
changes in cerebral blood flow may predispose the very fragile network of periventricular vessels to rupture.
Hence, it is essential to handle these very fragile infants with extreme care, avoiding unnecessary disturbances,
and preventing, rather than correcting physiological deviations in acid-base balance, blood gases, blood
pressure, or body temperature. Also, overly aggressive ventilation either in the delivery room or in the NICU
may predispose to significant acute or chronic pulmonary problems such as hyperinflation and loss of elasticity
of the alveoli, pulmonary interstitial emphysema, pneumothorax, and eventually CLD.
The vast majority of our ELBW infants who require assisted ventilation are intubated in the NICU. Only in
exceptional situations, when the infant does not respond to bag and mask ventilation, is intubation performed in
the delivery room. We use the nasotracheal route, and an endotracheal tube (ETT) of 2.5 mm for infants with a
birth weight less than 1000 g. We believe that it is important to use a low-caliber ET tube to avoid subglottic
trauma, strictures, and eventually stenosis. We have never had to perform a tracheostomy in an infant, and
stridor has been very rare among our patients.
If the newborn infant is in distress, an umbilical arterial catheter is inserted. The arterial line is used exclusively
for blood sampling. We favor the high position of the tip of the catheter, just above the level of the diaphragm.
After each blood sampling, the catheter is flushed with a heparinized solution of 0.45% saline. In infants
weighing less than 750 g, we also insert an umbilical venous line. We use the venous line to infuse fluids, thus
avoiding excessive handling and disturbance to the newborn infant during the first 24 to 48 hours of life. The tip
of the catheter is positioned at the junction of the inferior vena cava and the right atrium, hence avoiding the
liver. This is important, particularly when infusing hypertonic solutions, such as sodium bicarbonate or calcium
gluconate.
Blood is analyzed for glucose, electrolytes, blood gases, hemoglobin, and leukocytes. An intravenous with 10%
dextrose is initiated at a rate of 65 to 85 mL/kg/d and the infant is placed in a high humidity closed incubator.
Serum glucose levels are closely monitored, and the concentration of dextrose administered is adjusted
accordingly. In very tiny babies, when more than ten percent of the baby's blood volume has been removed, we
replace it with packed red cells. We use a single donor, collecting the blood in small packs, which may be used
for up to several weeks (49). Donors are extensively screened for all viral illnesses. We encourage parents who
wish to donate blood to their infants to do so, as long as they are of a compatible blood group and are free of
viral and other infections. Very sick infants receive 1:1 nursing care until the condition is stabilized, at which
point the ratio of nurse-to-baby becomes 1:2.
A percutaneous central venous catheter is usually inserted between the second and third day of life for
intravenous alimentation (50,51). As portal of entry, we use the upper extremities of the infant, and we aim for
the tip of the catheter to be in the superior vena cava, being careful to avoid intracardiac positioning, with its
attendant risks of erosion into the pericardial space (52,53). In case of failure to properly position a central line,
peripheral venous access is maintained using an extremity or scalp vein. Parenteral nutrition (TPN) is generally
introduced within the first 24 hours of life. When the mother receives intravenous fluids during her labor, a
baseline electrolytic profile of the newborn shortly after birth appears to be the proper way to follow subsequent
changes. Electrolytes are repeated between 12 and 18 hours of age. During the first 72 hours, the body weight
is recorded every 8 hours, and fluid intake is adjusted accordingly. The new incubators have incorporated scales,
allowing recordings without excessive handling and disturbance of the newborn infant. One other major
advantage of the new incubators is that they can provide a high level of humidity, hence substantially reducing
the need for large volumes of fluid.
To establish prognostic criteria, it is important to obtain a cranial ultrasound in the first 24 hours of life (54,55).
This ultrasound needs to be repeated at least one week later, or as often as necessary, depending on the
pathology detected on admission or, if the infant's condition has deteriorated, suggesting CNS involvement. It is
also important, before discharge from the hospital, to repeat the cranial ultrasound to evaluate the presence or
absence of periventricular leukomalacia (PVL) (56). Ideally, this last ultrasound should be performed at 35 to 36
weeks of postmenstrual age.
Respiratory Support
The vast majority of infants with a birth weight less than 1,000 g will need some form of respiratory assistance
to survive. For vigorous infants, nasal continuous positive airway pressure (CPAP) or recently, nasal ventilation,
is the preferred mode of support (57). Some controversy surrounds the timing and criteria for the initiation of
assisted ventilation. Likewise, controversy also exists regarding whether these tiny infants should receive
prophylactic exogenous surfactant in the delivery room (58,59,60,61,62). We do not systematically intubate
infants born weighing less than 1,000 g, and we do not administer surfactant unless the infant requires assisted
ventilation and a minimum FiO2 of 0.30.
P.466
The introduction of exogenous surfactant therapy has reduced significantly the mortality of all newborns
suffering from respiratory failure secondary to respiratory distress syndrome (RDS), but its impact has been
particularly important among the most premature infants (63,64,65,66,67,68,69). Administration of surfactant
in these very tiny infants requires extra care, as rapid changes in lung compliance may not only damage the
lungs by creating overinflation and overdistention, but also may predispose to acute changes in ductal circulation
which, in turn, could lead to both cerebral and/or pulmonary hemorrhage. With rapid improvement in
oxygenation, persistent hyperoxia also may be detrimental to the eyes. Hence, the administration of surfactant
should be performed by an experienced person, under close monitoring of ventilatory parameters, and with
rapid reduction of peak inspiratory pressures (PIP) and oxygen concentrations. If necessary, a second dose of
surfactant may be administered as soon as 6 hours after the first. However, in our experience, if the response to
the second dose is not satisfactory, it is highly unlikely that the condition will improve with additional
administration of surfactant. In our center, 64% of the babies improved rapidly, requiring only a single dose of
surfactant. Natural surfactant preparations are nowadays practically the only ones used (70).
Mechanical ventilation has dramatically improved the survival of infants weighing less than 1,000 g. In the
1970s, very few infants born weighing less than 1,000 g survived. In the early 1980s, survival of infants
weighing 500 to 750 g varied from 3% to 25%, and that of infants weighing 750 to 1,000 g ranged from 30% to
70% (71). In initiating mechanical ventilation, it is imperative that minimal effective settings be used (72).
Studies have shown that hyperventilation and overinflation of the lungs increase the loss of surface active
phospholipids (73). Also, overinflation predisposes to air leaks and particularly to pulmonary interstitial
emphysema (PIE). The latter is a serious complication in the tiny infant, and is a relatively frequent one. It is
probably related to structural immaturity of the lungs, particularly to the relative lack of elastic tissue, which
normally increases progressively throughout gestation (74). Also, the interstitium is larger in the more immature
infant as a result of poor alveolization. Although drainage of a pneumothorax may lead to rapid improvement,
management of PIE is far more complicated. As lung compliance is reduced, there is a need for increased PIPs to
maintain adequate ventilation. This results in increased barotrauma to the small airways. Chorioamnionitis has
been reported as a risk factor predisposing to PIE (75). The highest incidence of PIE in tiny infants has been
observed when intrauterine pneumonia complicates the RDS. To overcome the problems related to PIE, a
number of strategies have been devised. These include acceptance of higher levels of partial pressure of CO2
(PCO2) and lower levels of potential of hydrogen (pH), reduction of the positive end-expiratory pressure (PEEP)
to between 2 and 3 cm water (H2O), selective intubation of the contralateral lung, positioning the infant on the
affected side, increasing the expiratory time, and systemic corticosteroid therapy. The combination of the above
strategies can occasionally produce quite spectacular recovery from this condition. However, high frequency
oscillatory or JET ventilation is probably the most effective therapy (76,77).
A variety of ventilatory strategies have been promoted to maintain satisfactory ventilation and to reduce the risk
of complications (78), such as high PIP-low rates, low PIP-high rates, variation in the I:E ratio, variations in the
flow, permissive hypercapnia, tolerance of lower pH and, more recently, high-frequency oscillation and even
ventilation via nasal prongs. In recent years, however, the general trend is to use the lowest possible PIP to
achieve acceptable ventilation and oxygenation. Of course, the question is what is considered acceptable?
Some neonatologists will tolerate a pH as low as 7.20 and a PCO2 as high as 65 mm Hg. Most centers also aim
for PaO2 values between 50 and 70 mm Hg. Our own approach to the ventilation of tiny infants over the years
has been to favor nasotracheal intubation with a 2.5-mm ET tube. Our PIPs rarely exceed 14 to 15 cm H2O, and
we set the PEEP at 5 cm H2O, with initial rates of 65 to 75 per minute. We aim for arterial oxygen pressure
(PaO2) values between 45 and 50 mm Hg, which is enough to abolish production of lactic acid and, at the same
time, remains relatively close to intrauterine values. Our pulse oximeters (79) are set to alarm at a lower limit of
80% and an upper limit of 93% for the first few weeks of life. We believe that this modest degree of
oxygenation offers the advantage of reducing the need for administration of elevated oxygen concentrations,
thus minimizing lung toxicity, and may help to avoid retinal damage. Our incidence of bronchopulmonary
dysplasia (BPD) and ROP are shown in Table 25-9. We believe that by using the lowest possible PIP and initially,
a relatively rapid respiratory rate, we reduce overdistention and barotrauma and minimize the risk of lung
injury. Because, in RDS, there are compartments in the lung with relatively normal ventilation perfusion ratios
and others with poor ventilation and adequate perfusion, it seems reasonable to attempt to improve ventilation
of the poor ventilation perfusion (V-Q) compartment without overdistention of the normal V-Q compartment.
Raising the ventilatory rate, which raises the mean airway pressure without changing the PIP, appears to
accomplish this (80). We also have observed that with initially higher respiratory rates, the tiny infant very
rapidly stops fighting the respirator, thus making the gas exchange smoother and possibly decreasing the
incidence of air leaks. Relatively high respiratory rates also seem to be more physiologic for the very immature
infant, as observed by Greenough and collaborators (81). For toilet of the airways, we use the Ballard closed
suction circuit, thus avoiding disconnecting the infant from the ventilator (82). We suction sparingly during the
first few days of life, when the volume of secretions is minimal. Analgesia/sedation is given for nonemergent
intubation and for infants who remain agitated while on mechanical ventilation.
Most of our ventilated babies have their umbilical vessels cannulated for blood sampling and for fluid infusion.
P.467
As soon as the procedures of intubation and catheterization of the umbilical vessels are completed, we perform
chest and abdominal radiography to assess the position of the ET tube and the umbilical catheters and, at the
same time, to evaluate the severity of lung pathology. Fifteen minutes after the initiation of ventilation, we
obtain an arterial blood gas and adjust the ventilatory parameters accordingly. We generally aim for a pH above
7.25 and a PCO2 between 45 and 55 mm Hg, but when the PIPs are elevated or in the presence of PIE, we
tolerate partial pressure of CO2 (PCO2) values up to 60 mm Hg as long as the pH is at least 7.20. Our ET tubes
are sutured to the tape placed on the upper lip. We record the level at which sutures were placed on the ET
tube, thus avoiding the need to repeat a chest radiograph to evaluate the tube position when reintubation is
required. Actually, we take very few radiographs, and we rely extensively on clinical assessment, blood gases,
and pulse oximetry. However, a chest radiograph should be taken if there are concerns about the position of the
endotracheal tube or the development of any form of air leak.
Avery and associates (83) reported in 1987 that the incidence of BPD varied between neonatal units. The unit
with the lowest incidence used CPAP much more frequently than did the other units. Epidemiologic data from 36
units in the Vermont-Oxford Trial Network also indicate large differences in the incidence of BPD, from 16% to
70% for infants weighing between 501 and 1,500 g (84). The incidence of BPD was lower in units allowing
higher PCO2 values. More evidence of the association of BPD and PCO2 was provided by Kraybil and associates
(85). More recently, Garland and associates (86) reported the highest incidence of BPD among infants with the
lowest PCO2 before the administration of surfactant.
The concept of permissive hypercapnia for patients requiring mechanical ventilation gives priority to the
prevention or limitation of severe pulmonary hyperinflation over the maintenance of normal ventilation. The
principle consists of allowing the PCO2 to rise by minimizing ventilator pressures and tidal volume (87). Potential
risks of high PCO2 values include increased cerebral perfusion, increased retinal perfusion, increased pulmonary
vascular resistance, and reduction of pH. Based on epidemiologic observations, it appears that respiratory
acidosis, unlike metabolic acidosis, is not associated with poor neurologic outcomes. Vannucci and associates
(88) demonstrated similar findings in animal studies involving rats.
Flow rate also can affect ventilation and increase airway injury. We generally use a flow rate between 3 and 5 L/
min. Only when we need very high pressures, for instance, in the presence of pulmonary hemorrhage, do we
allow the flow rate to exceed 5 L/min.
Several reports in the literature have expressed concern about potential side effects of low PCO2 values (89).
Graziani and associates (90) reported that, along with other factors, marked hypocarbia during the first 3
postnatal days was associated with increased risk of periventricular white matter injury in premature infants.
The theoretical model of ischemic brain injury has been described by Wigglesworth and Pape (91). These
authors hypothesize that cerebral blood flow could be decreased by several factors, including hypotension,
hyperoxia, hypocarbia, and increased venous pressures. Concern also has been expressed in the literature about
high-frequency ventilation, which may lead to low PCO2 values as a result of effective alveolar ventilation (92).
However, the data regarding the development of PVL among infants managed with these devices remain
controversial. Most authors agree, however, that for hypocarbia to be dangerous for the brain, it has to reach
levels below 30 mm Hg. Our policy is to avoid PCO2 values below 40 mm Hg by first reducing PIP before
reducing respiratory rates.
High frequency oscillatory ventilation (HFOV) has been used in recent years in an attempt to reduce the
incidence of early ventilatory complications and to prevent bronchopulmonary dysplasia. Published reports are
often contradictory and, so far, there is no clear evidence that HFOV offers an advantage over conventional
ventilation (93,94). HFOV, however, offers an advantage when treating infants with PIE, pulmonary
hypertension, or diaphragmatic hernias (95).
More recently, the effects of patient-triggered ventilation with volume-guarantee have been explored in the
management of preterm infants (96). This novel technique calls for an automatic adjustment of the peak
inspiratory pressure to ensure a minimum set mechanical tidal volume.
The timing of extubation of ELBW infants is very important, because they are prone to develop severe apnea,
with the potential risk of cerebral injury. Nowadays, with early administration of surfactant and improvement in
lung compliance, rapid extubation and placement on nasal CPAP or nasal ventilation is possible in the majority of
ELBW infants. However, some extremely premature infants develop severe episodes of apnea and desaturation,
requiring frequently reintubation. For this reason, for the tiniest infants, we often favor a more progressive
weaning process by maintaining them for a few extra days at a very low PIP of 10 to 12 mmHg and rates of 15
to 25 per minute, although providing maximum intravenous and oral alimentation (97). When the infant is
stronger and starting to gain weight, we administer caffeine and proceed directly to extubation. The infant then
is placed on nasal CPAP. The CPAP is discontinued when, after periodic trials, the infant can maintain good
oxygenation without significant apnea, bradycardia, and desaturations. If an infant on nasal CPAP shows signs of
fatigue manifested by recurrent apnea and retention of CO2, we try nasal ventilation prior to reintubation. In
many circumstances, this approach provides the extra help that these tiny infants require to avoid reintubation
(98,99).
Cardiovascular Support
By far, the major cardiovascular problem in ELBW infants is the presence of a patent ductus arteriosus (PDA).
More than 50% of infants born weighing less than 1000 g will
P.468
have a PDA diagnosed during the first few days of life (100,101). The onset of clinical manifestations of the PDA
is related to the timing of improvement of the infant's respiratory status, which is associated with a decreasing
pulmonary vascular resistance and a predominantly left-to-right shunt. The patency of the ductus arteriosus can
be easily documented in the first hours of life, with the help of echocardiography. At this early stage of life, the
shunt is either right-to-left or bidirectional, depending on the severity of the infant's respiratory condition. In our
center, the incidence of clinically significant PDA requiring therapy has been around 65% of all infants weighing
less than 1,000 g. The left-to-right ductal shunting can be diagnosed as early as in the first day of life in infants
with RDS who improved following surfactant therapy (102). An active precordium, with bounding pulses and
visible carotid pulse, will often precede auscultation of a murmur. If left untreated, the infant may develop leftsided heart failure and pulmonary edema or hemorrhagic pulmonary edema, with significant deterioration of the
respiratory status. Significant left-to-right ductal shunting may cause decreased peripheral perfusion and oxygen
delivery. ELBW infants with significant PDA are at risk for IVH, necrotizing enterocolitis (NEC), renal failure, CLD,
and metabolic acidosis (103). The size of the ductus arteriosus, and the ratio of the left atrium to aortic root can
be easily measured by echocardiography (104). We consider as significant a PDA of diameter greater than 1.5
mm and/or a ratio of left atrium to aortic root greater than 1.3 for ELBW infants.
The ductus arteriosus of the premature infant is less responsive to the vasoconstrictive effect of oxygen and is
less likely to close spontaneously than that of term infants, especially in infants with RDS. The classic
management of PDA first involves medical and supportive measures, i.e., fluid restriction, diuresis, distending
airway pressure, transfusion of packed red blood cells to keep the hematocrit above 0.4. If these measures fail
to close the ductus arteriosus, pharmacologic closure is possible with a cyclooxygenase inhibitor, namely
indomethacin (105,106,107). In ELBW infants, because closure of the ductus arteriosus is unlikely to occur
spontaneously, the infant is at risk for the short- and long-term complications mentioned previously. Hence,
therapy with indomethacin has become standard practice for the majority of these infants. Different protocols of
treatment have been proposed. We treat most infants with 0.2 mg/kg of indomethacin every 12 hours for three
doses. For infants developing clinical and/or biochemical signs of renal failure, we subsequently use 0.1 mg/kg
per dose (108). A reduction in fluid intake is advisable prior to administration of indomethacin. In our experience
about 80% of the treated infants will respond with a functional closure of the ductus. However, about 30% of
these may reopen, in which case, further indomethacin is administered. If the ductus arteriosus fails to close
after three courses of indomethacin, or if indomethacin cannot be administered because of significant renal
dysfunction on previous treatment, then surgical ligation is necessary. In the last 8 years, 16.7% of ELBW
infants required surgical ligation of the ductus after repeated failures of pharmacologic therapy. The more
premature the infant, and the greater the postnatal age at the time of treatment, the greater the failure of
indomethacin (109). A group particularly resistant to indomethacin therapy is ELBW infants with severe IUGR
(26). Suggested hypotheses for this high failure rate of indomethacin therapy among this group of infants
include chronic hypoxia, altered levels of prostaglandin, and altered number or sensitivity of their receptors
(110).
Contraindications to indomethacin therapy include renal failure, active bleeding, and thrombocytopenia. The
presence of IVH does not appear to be an absolute contraindication to the use of indomethacin. Recent studies
indicate that there is no progression of the severity of IVH after administration of indomethacin for PDA closure
(111). Until additional data are available, it is prudent in the presence of IVH to verify the platelet count prior to
the administration of indomethacin and to eliminate any bleeding diathesis.
The optimal timing of administration of indomethacin has been a matter of debate. Although early or
prophylactic use of indomethacin may result in a higher initial PDA closure rate and in a reduction in the
incidence of severe IVH, it does not appear to ultimately reduce the need for surgical PDA ligation, does not
confer any long-term respiratory advantage, and does not change the rate of survival without neurosensory
impairment at 18 months (112,113).
Another cyclooxygenase inhibitor that seems very promising is ibuprofen (114). It has the theoretical advantage
over indomethacin that it may increase the range of blood pressure at which cerebral blood flow is
autoregulated, but has little effect on cerebral blood flow during normotension (115,116). In a Phase I study
carried out in our center, we observed a dramatic response in 12 ELBW infants treated with ibuprofen within 3
hours of birth (117). All twelve infants had their PDA permanently closed after three doses of ibuprofen. We also
observed a significant trend in the reduction of IVH and practically no side effects. These findings have since
been confirmed by other groups of investigators. The efficacy of ibuprofen for the treatment of PDA appears to
be equivalent to that of indomethacin, with a lower likelihood of oliguria (118). Furthermore, in a series of
patients evaluated by near-infrared spectroscopy and Doppler ultrasonography, ibuprofen therapy did not result
in a significant reduction in cerebral perfusion and oxygen availability (119). One reported side effect of
ibuprofen, namely the development of transient severe pulmonary hypertension and hypoxemia in a few infants
treated within six hours of birth, may still require further evaluation (120).
Finally, in the presence of a clinically significant ductus that fails to respond to pharmacologic closure or in the
presence of contraindications to the use of indomethacin, surgical ligation should not be delayed, as the
presence of a continuous left-to-right shunt may contribute to ventilator and oxygen dependency with well-
known consequences in terms of CLD.

P.469
Regarding early management of low blood pressure, with the exception of hypovolemic shock, when volume
replacement is an urgent matter, we rarely use volume expanders and -adrenergic drugs to maintain blood
pressure. We pay particular attention to skin perfusion. As a rule of thumb, we aim for a mean blood pressure
that is numerically slightly above the gestational age of the infant.
FLUIDS AND ELECTROLYTES

The management of fluids and electrolytes remains one of the most challenging aspects in the care of the ELBW
infant. Knowledge of the body composition of these infants and better understanding of their renal function has
helped to determine their requirements (121,122).
It is important to remember that the body of the ELBW infant is made up of 85 to 90% water, which is
distributed as one-third intracellular water (ICW) and two-thirds extracellular water (ECW). Immediately
following birth, glomerular filtration rate and fractional excretion of sodium (FENa) are low and urine output is
minimal. This is followed by a diuretic phase, which results in a decrease in the ECW compartment. Furthermore,
because of the large body surface area to body weight ratio and the underdeveloped epidermis of the ELBW
infant, evaporative losses, if uncontrolled, may be of very significant magnitude, i.e., 5.7 mL/kg/h (123,124).
Additionally, the immature kidney, having a limited concentrating ability (less than 700 mOsm), produces large
amounts of diluted urine. Thus, without close control of fluid intake and the infant's environment, the infant is
very vulnerable to dehydration and hypertonicity, which may predispose to IVH. One has to be also careful not
to overload the infant with fluids, because this may have an impact on PDA, with possible congestive heart
failure, pulmonary edema, worsening pulmonary function and CLD (125).
Shortly after admission, once the infant's respiratory status has stabilized and the need for ready access to the
infant is not as essential, the infant is transferred from a radiant warmer to an incubator, maintaining 75 to 80%
humidity for the first few days of life. This approach significantly reduces water and heat loss and allows
limitation of fluid intake to between 65 and 85 mL/kg/d for the first day of life. Additionally, we monitor the
infant's weight every 8 hours during the first days of life, and adjust fluid intake consequently. Under these
described conditions, there is rarely a need to exceed an intake of 170 to 180 mL/kg/d even for infants with a
birth weight less than 600 g.
Electrolyte abnormalities such as hypernatremia, hyponatremia, and hyperkalemia are frequently seen in ELBW
infants. Hypernatremia is usually the result of severe insensible water loss, but can be secondary to the
treatment of metabolic acidosis with large amounts of sodium bicarbonate. Hyponatremia (less than 130 mmol/
L) is more frequently seen because of the high FENa during the diuretic phase. It may also be noted during the
first day of life if the mother has received large amounts of hypotonic intravenous fluids. Hyponatremia is also
seen following therapy with indomethacin without prior proper reduction of fluid intake, and later on when
diuretics are used for the treatment of CLD.
On admission, infants receive only a 10% dextrose solution in water. We monitor blood glucose and electrolytes
closely and we test all urine for glucose. We subsequently adjust the intravenous dextrose according to the
blood glucose values. Protein and lipid intake is generally initiated during the first 24 hours of life in the form of
parenteral alimentation. We begin sodium supplementation only when its serum value is less than 140 mmol/L,
which usually occurs between the second and third day of life.
Hyperkalemia (K+greater than 7 mmol/L) is a severe acute problem in this ELBW group of infants, even in the
absence of oliguria and potassium intake (126). A rapid rise in serum potassium may be seen during the first 24
hours of life, especially in the most immature infants. A few mechanisms have been proposed for this
hyperkalemia, i.e., relative hypoaldosteronism, immaturity of the renal distal tubules, and internal potassium
shift from the intracellular space to the extracellular space. Hyperkalemia is also more severe in infants with
intraventricular or pulmonary hemorrhage, extensive bruising, or renal failure. It is prudent to obtain a baseline
measurement of electrolytes following birth, particularly when the mother has received intravenous fluids. We
routinely repeat the electrolytes every twelve to eighteen hours during the first couple days of life, or more
frequently if indicated, especially for the smallest infants.
We treat hyperkalemia when it exceeds 7 mmol/L with insulin and sodium bicarbonate to allow an intracellular
shift of potassium and calcium gluconate for stabilization of the myocardium. Potassium supplements are
introduced only after the serum level has stabilized below 4.5 mmol/L.
We also monitor urine output, and use the urine specific gravity and osmolality as an additional guide to assess
the renal function and hydration status of the infant. Typically, the urine pH of the ELBW infant is greater than 7
in the first few of days of life, then decreases as the tubular reabsorption of bicarbonate increases. Microscopic
hematuria is consistently seen in the ELBW infant during the first few days after birth, regardless of the health
status of the infant. Monitoring for the presence of glucose in the urine can be a good indicator of the
carbohydrate homeostasis of the newborn infant. Some extremely premature infants have a low glucose renal
tubular threshold and may be predisposed to osmotic diuresis.
Management of fluids in the ELBW infants is very much dependent on securing an intravenous line. Intravenous
access may, at times, become very difficult, thus compromising fluid, electrolyte, and glucose homeostasis,
thermoregulation, and physiologic stability as a result of the pain induced by multiple attempts to insert a
venocath. The umbilical vessels usually provide an easy access route for the first days of life. Subsequently,
small neonatal infusion needles (IMP Group International Inc.) are inserted in
P.470
the scalp, or a venocath is inserted in one of the extremities. In our center, we favor the insertion of a
percutaneous central venous catheter, which gives continuous venous access for as long as necessary (50). In
our experience, this technique has not increased the incidence of infection when compared to peripheral
catheters used for prolonged parenteral alimentation. (51)
The Skin of the Extremely Low Birth Weight Infant

The skin of an infant born at 23 to 26 weeks of gestation is extremely immature. Maturity of the epidermis is
present at birth only in infants born after 32 weeks of gestation (127). Prior to this time, the epidermis is
underdeveloped, especially the stratum corneum, predisposing to very high transepidermal water loss, and to
the risk of trauma and percutaneous absorption of toxic agents. The skin also is permeable to gases, allowing for
diffusion of oxygen and carbon dioxide. The tremendous transepidermal water loss of the ELBW infant
predisposes to dehydration, electrolyte imbalance, and evaporative heat loss. Trauma to the skin may provide a
portal of entry for infectious organisms. Fortunately, after birth, there is acceleration of epidermal maturation,
such that, by 2 weeks of age, the skin of the premature infant almost resembles that of the term infant.
Preservation of skin integrity and prevention of transepidermal water loss have been, and still are, areas of
challenge in neonatology. Under a radiant warmer, water loss is extensive. Plastic shields have been used to
decrease this loss, with variable success. When the infant is placed in an incubator, increasing the ambient
humidity to 80% reduces total extracellular loss to a minimum. Initial studies using the topical ointment
Aquaphor; had shown reduced losses for up to 6 hours following application, improved skin condition and
decreased skin bacterial colonization, thus possibly reducing the risk of nosocomial infection (128). However,
more recent studies indicate no advantage with this approach. (129)
Nutrition
Nutrition is an essential part of the care of the ELBW infant. These tiny infants are born with very low reserves of
fat and carbohydrates, and they rapidly develop nutritional deficiencies in calcium, phosphorus, iron, trace
minerals, and vitamins. Their endocrine and enzymatic capability is limited as a result of immaturity. Postnatally,
they rapidly enter a catabolic state unless provided with sufficient nutrients. On the other hand, reversal of this
catabolic state often is difficult because of limited feeding tolerance. The gastrointestinal (GI) tract is immature
in terms of digestive pathways and motor function, increasing the risk of developing NEC.
The first goal of nutrition is to prevent catabolism. Usually, this will be achieved by providing a minimum of 50
kcal/kg/d. Growth will require additional caloric intake. Achieving steady growth is essential for the ELBW infant,
because the intrauterine growth velocity at 25 to 30 weeks' gestation is relatively higher than at term. If
reasonable caloric intake cannot be provided, catch-up growth may never be achieved.
In the early days of life, satisfactory nutrition can never be achieved exclusively with milk. Parenteral nutrition
provides the additional calories (130). In contrast to enteric nutrition, parenteral administration starting at 80 to
85 kcal/kg/d can provide the necessary calories for growth. When the infant no longer is receiving intravenous
nutrition, 100 to 120 kcal/kg/d are needed to maintain growth. However, this level of caloric intake may not be
sufficient in infants suffering from CLD and other high-energy-requiring conditions (131). Appropriate growth, if
one wants to mimic intrauterine growth, should be a 2% daily increase in body weight, slowing to 1% near term.
Both parenteral nutrition and enteral nutrition are not without difficulties and complications in the ELBW infant.
TPN requires intravenous access and, this can be associated with a variety of infections, with Staphylococcus
epidermidis being the most common. The composition of TPN has been, and remains, an area of active research,
especially regarding the composition of essential amino acids and fatty acids. Exclusive intravenous nutrition
affects the mucosal lining of the GI tract, which is bypassed and eventually may lead to villus atrophy. TPN also
requires regular metabolic monitoring for glucose, electrolytes, urea, lipids, and acid-base balance. Cholestatic
jaundice is a frequent complication of TPN (132). In the vast majority of cases, this is a self-limited condition,
the exact etiology of which is not completely understood, but appears to implicate both amino acids and lipids.
Enteral nutrition may consist of either breast milk or premature formulas. During fetal life, the fetus constantly
swallows amniotic fluid, promoting intestinal development. Enteral feeding, even in small amounts, has been
demonstrated to stimulate trophic factors and hormonal maturation of the GI tract, thus improving overall
intestinal function and potentially improving feeding tolerance and preventing mucosal atrophy (133,134).
Whether early introduction of feedings and stimulation of the GI tract could prevent or decrease the incidence of
NEC has not yet been established.
The current standard for postnatal nutrition is to duplicate as much as possible in utero fetal growth rates (135).
To achieve this goal, high intravenous amino acid and lipid intake are necessary. Recent studies indicate that
more aggressive amino acid intakes up to 3 g/Kg/d from the first days of life not only are well tolerated, but
they also suppress protein breakdown and provide increased protein accretion and synthesis. (136)
In our center, we start TPN from day one, as soon as the infant is metabolically stable, with 1 g/kg/d of amino
acids, 1g/kg/d of lipids, and glucose according to tolerance. Calcium, phosphorus, vitamins, and trace minerals
also are added. Sodium and potassium are added according to the electrolytic profile. The intake of amino acids
and lipids is rapidly increased to a maximum of 3 to 3.5 g/kg/d. Lipids
P.471
are restricted in the presence of severe indirect hyperbilirubinemia. We also adjust the amino acid intake
according to urea, pH, and the degree of TPN-related cholestasis. Monitoring of urea, creatinine electrolytes,
glucose, and bilirubin is performed daily for the first 3 to 4 days of life and then is reduced to twice a week.
When the oral intake is half of the total caloric requirement, monitoring is performed only once a week. During
the past 8 years, the mean duration of any amount of TPN administration among 253 survivors born weighing
less than 1,000 g was 31.5 15 days (median 28.5 days, range 9 to 121 days).
In terms of enteral nutrition, we attempt to introduce minimal feedings, starting in stable infants as early as 48
hours of life. In the vast majority, we use breast milk or, if necessary, a premature formula (68 kcal/100 mL).
However, tolerance varies widely from one infant to the other. We encourage mothers to pump their own milk,
which is used exclusively for their own babies. The progression of enteral feedings and the addition of a breast
milk fortifier varies from one neonatal unit to the other. For the smallest infants, we increase the volume per
feeding by 1 mL every 24 hours, up to 10 mL, and then by 1 mL every 12 hours. Generally, we add a fortifier to
the breast milk, thus increasing its caloric content to 81 kcal/100 mL when oral intake is between one-quarter
and one-half of the total daily intake. For those infants on premature formula, we also advance to a more calorie
dense formulation (81 kcal/100 mL). Intolerance to milk feeds is not an infrequent occurrence. Tolerance of full
enteral feeding, in our experience, usually is achieved between 20 and 30 days of life. In the 192 survivors
mentioned previously, birth weight was regained at a mean of 12.2 6.8 days (median 13 days, range 2 to 48
days). These data are comparable to that reported by Berry and associates (137).
Glucose, Calcium, and Phosphorus Homeostasis

Early hypoglycemia frequently is seen in this group of infants, because of poor glycogen reserves and the
immaturity of the postnatal adaptive mechanism of endocrine and enzymatic control of glucose. In particular,
ketogenesis and lipogenesis, which lead to the production of alternate fuels, are limited in very premature
infants, making the infant more dependent on glucose (138,139). Hence, an infusion of dextrose at a rate of 8
to 10 mg/kg/min is necessary to maintain normoglycemia.
Hyperglycemia is a frequent and challenging complication, particularly in extremely immature infants of 23 to 24
weeks' gestation and in those with IUGR. This condition is usually the result of high glucose infusion rates, but
may also occur because of incomplete suppression of hepatic glucose production in the presence of
hyperglycemia, reflecting the immaturity of the regulatory mechanisms mentioned previously (140).
Hyperglycemia poses the risk of osmotic diuresis, and thus of increased water loss, which eventually may have
cerebral implications. Insulin can be used to control hyperglycemia. Although the exact mechanism of its action
in the extremely premature infant is not completely clear, it is thought to work by decreasing hepatic production
of glucose and increasing glucose utilization by peripheral tissues.

We monitor blood glucose within 1 hour of birth and as often as required, until stabilization. We also test all
urine for glycosuria. We tolerate blood glucose values up to 8 or even 10 mmol/L, provided there is no
glycosuria. If an insulin infusion is used, one must remember to preflush the tubing with the infusate, because
insulin adheres to plastic and, unless the binding sites are saturated prior to infusion, a very erratic infusion of
insulin may occur, making the interpretation of glucose values difficult (141,142). The sudden onset of
glycosuria in a previously stable infant may be an early sign of infection. Hyperglycemia also is seen frequently
with initiation of dexamethasone therapy for BPD (143). Because of the slow metabolic adaptation of the ELBW
infant, rapid and significant changes in glucose intake should be avoided to prevent episodes of hypo- or
hyperglycemia, which can become difficult to control. Finally, for the treatment of acute and severe
hypoglycemia, a bolus of no more than 200 mg/kg of dextrose can be administered as needed, although, at the
same time, increasing the concentration of the intravenous glucose infusion.
Additional calcium is necessary in ELBW infants both in the early days of life and later, when their limited
calcium reserves are rapidly depleted during this period of rapid growth. However, it is important to appreciate
that, because of the low serum albumin of the immature infant, total serum calcium rarely exceeds 1.75 mmol/
L. Measurement of ionized calcium is, of course, the ideal way to evaluate hypocalcemia. Because hypocalcemia
also can induce apnea, we always verify calcium levels after the first 24 hours of life. It is important to
remember that metabolic acidosis can give falsely reassuring values of serum calcium, which decline rapidly with
the improvement of acid-base balance. Mechanisms involved in the early manifestations of hypocalcemia include
parathyroid dysfunction, renal immaturity, and calcitonin stimulation. Treatment of hypocalcemia consists of
administering 500 mg/kg/d of calcium gluconate. As soon as we introduce TPN, 300 mg/kg/d of calcium
gluconate is added daily to the solution, together with multivitamins. Because maternal milk has an inadequate
mineral and vitamin content, we use a human milk fortifier and vitamins. In the presence of persistent
hypocalcemia, it is important to verify the levels of magnesium. The daily recommended dose of Vitamin D in
ELBW infants is between 800 and 1,000 IU. (144)
Acid-Base Balance
No other investigation is ordered more frequently in ELBW infants than measurements of acid-base balance.
Both the respiratory and metabolic components need frequent adjustments. Acid-base homeostasis varies in
relation to the degree of renal maturity. The renal threshold of bicarbonate can be as low as 15 mEq/L. Hence,
in ELBW infants,
P.472
there is often a need for additional buffering with sodium bicarbonate. The need for supplemental sodium
bicarbonate also is frequent with high amounts of amino acids in the TPN. We initiate correction of the acid-base
balance as soon as the base deficit exceeds 5 to 6 mEq/L. Because TPN contains calcium, the addition of sodium
bicarbonate may induce precipitation. For this reason, we have elected to administer sodium bicarbonate via
slow push of 0.5 mEq/kg every 2 to 6 hours, according to the severity of the metabolic acidosis, while evaluating
the progress of correction and adjusting the frequency of administration accordingly. With this approach, we do
not need to discontinue intravenous alimentation, nor do we need to start a second intravenous line. Sodium
acetate added to the TPN as an additional buffer, is an alternative treatment to acidosis (145). The late
metabolic acidosis of prematurity also is related to a combination of increased nitrogen load and low renal
threshold.
Although acidosis is the main concern in the early days of life, later on, many of these tiny babies may develop a
metabolic alkalosis as a result of the administration of diuretics, in combination with fluid restriction, as part of
the management of CLD. On occasion, we have found it helpful to administer acetazolamide at a dosage of 2.5
mg/kg every 12 hours to overcome a significant metabolic alkalosis (pH more than 7.45).
Jaundice
Rarely will an infant weighing less than 1,000 g escape the need for phototherapy. Hepatic immaturity and
reduced erythrocyte lifespan, blood group incompatibilities, extensive extravasation of blood, and increased
enterohepatic circulation as a result of poor bowel motility all contribute to the fact that ELBW infants are very
prone to develop jaundice. Because the serum bilirubin binding capacity is decreased in premature infants as a
result of the lower serum albumin, the level at which toxicity for the brain and acoustic nerves may occur is
much lower than that of the more mature infant. Guidelines for the initiation of phototherapy have been
proposed in the past and have undergone frequent revisions. However, some basic principles are universally
accepted and govern the management of jaundice. These include age of the baby in hours or days from birth,
gestational age, presence of a hemolytic disorder, degree of bruising or other extravasation of blood, and level
of serum albumin. We find it helpful in our decision to initiate or discontinue phototherapy to estimate the serum
binding capacity from the level of serum albumin (146). Based on this principle, we have not seen a single case
of kernicterus clinically or on autopsy material. Although we do not initiate phototherapy immediately after birth,
we believe that relatively early phototherapy can decrease significantly the need for exchange transfusion, which
incidentally is poorly tolerated in the very immature infant. We generally initiate phototherapy when the bilirubin
level has reached 80 micromol/L in the first 24 hours of life, or if we note an increment of more than 40
micromol/L/d. When phototherapy is used, it is important to increase the fluid intake by 15% to 20% to avoid
excessive insensible water loss. When the bilirubin level approaches the exchange transfusion level, it is
important to avoid variations in acid-base balance, high levels of lipid infusion, hypothermia, and certain
medications, which may compete with and displace bilirubin from albumin, thus precipitating kernicterus.
MAJOR MORBIDITIES OF THE EXTREMELY LOW-BIRTH-WEIGHT

INFANT
Neurologic Disorders
The neurologic examination of the newborn infant is related to gestational age and is affected greatly by any
disturbance in the CNS. The ELBW infant is typically hypotonic. The primitive reflexes are absent but brain stem
function (corneal, gag, oculocephalic reflexes, facial grimacing, nasal tickle) can be tested. Neuronal migration
usually is completed by 24 weeks of gestation, but synaptic development and myelinization are just beginning at
this age.
Intraventricular Hemorrhage
IVH is a major morbidity for the ELBW infant, with serious potential sequelae in surviving infants, which include
hemorrhagic periventricular infarction, posthemorrhagic hydrocephalus, seizures, PVL, and in the long-term,
neurosensory and neurodevelopmental impairments. Despite modern advances, this remains a common
problem, with an incidence of up to 44% of ELBW cohorts (147,148). Fortunately, the majority of ELBW infants
develop less severe IVH (Grade I or II). There is evidence that the overall incidence of IVH may be declining in
recent years (149). In a study of a large cohort of infants with birth weight of 500-1500 g from the National
Institute of Child Health and Human Development (NICHD) Neonatal Network, Shankaran and associates (147)
observed a significant decrease in severe intracranial hemorrhage (Grades III and IV) from 19% to 15% over a
3-year period. Significant variation in the rates of all grades of IVH has been observed between centers.
The degree of prematurity is a very strong predictor for IVH, with gestational age and birth weight inversely
correlated with the incidence and severity of IVH. Thus, the smallest and youngest infants are at greatest risk
for more severe IVH. Recent reports described the incidence of Grades III and IV IVH at 9% to 13% for ELBW
infants, in contrast to 2% to 5% for infants with birth weights greater than 1,000 g. Severe IVH was found in
16% of infants with gestational age at birth less than 25 weeks, in contrast to about 1% to 2% in which
gestation was greater than 25 weeks.
In our center, among 253 ELBW infants born between 1995 and 2002, we observed an overall incidence of
24.4%
P.473
for any IVH and 7.6% for severe IVH (Grades III and IV) (Table 25-8). It is important in assessing the incidence
of IVH to know whether it is representative of the whole population born weighing less than 1,000 grams or only
among survivors, in which the incidence of IVH would be much lower. Indeed, among our survivors, the
incidence of IVH was 16.6% and for grades III and IV, 4.7% (Table 25-9).
The variations in the incidence of IVH could also be explained by the multifactorial pathogenesis proposed by
Volpe (150,151), describing intravascular, vascular and extravascular factors, superimposed on the fragility of
the germinal matrix and the limited cerebral blood flow autoregulation in ELBW infants.
We are particularly attentive to rapid stabilization, avoidance of hypo- and hyperoxia, and hypo-and
hypercarbia, maintenance of normoglycemia, and control of the environment with the use of high humidity in
the first few days of life to prevent excessive water loss, hypernatremia and hyperosmolarity. We attempt to
maximize synchronization of the ventilator breaths to the infant's own respiratory efforts. We use the Ballard
closed suction circuit, thus avoiding frequent disconnection of the infant from the ventilator. We avoid volume
expanders, unless there is documented blood loss or significant hypotension. We favor early medical PDA
closure. Finally, we minimize handling of the infant, with clustering of care or use of sedative/analgesics as
warranted to decrease significant discomfort and pain. We also avoid significant apnea by initiating prompt
supportive management, such as early nasal CPAP/ventilation, methylxanthines or mechanical ventilation, as
necessary, and by rapid diagnosis and treatment of causative factors, such as infection, hypoglycemia, and
hypocalcemia.
IVH may present acutely, leading to shock and death. It may be clinically silent or, more commonly, it may
present with worsening cardiorespiratory instability. The time of occurrence of IVH has been well investigated
and our experience is similar to published data. About 50% of bleeds occur during the first day of life, 25%
during the second day, and 15% on the third day of life. (152). It is unusual for an infant to develop IVH after 7
days of life.
Ultrasonography is the most reliable and safest neuroimaging technique to diagnose IVH in premature infants
(153). We obtain a cranial ultrasound in the first 24 hours of life. If no IVH is detected, the study is repeated 1
week later, or sooner if the infant suffers from any acute event in the interim. If pathology is present,
ultrasonography may be repeated at intervals of 48 to 72 hours, until stabilization of the intracranial pathology
and as clinically warranted to facilitate parental counseling and decisions for care.
The immediate management of IVH involves stabilization of the cardiovascular system, correction of any
bleeding diathesis, and monitoring for hyperbilirubinemia and hyperkalemia. Careful neurologic examination and
serial measurements of the head circumference along with serial cranial ultrasounds must be planned for the
early detection and management of progressive posthemorrhagic hydrocephalus. If there is rapid dilatation of
the ventricles, neurosurgical intervention may be necessary for temporary or permanent drainage of the
cerebrospinal fluid. We have not found repeated lumbar punctures to be effective as a temporizing technique to
control progressive posthemorrhagic hydrocephalus.
Prognosis for mortality and long-term morbidity is related to the extent of the brain injury, hallmarked primarily
by the extent of the bleed. Severe IVH with periventricular hemorrhagic ischemia in the ELBW cohort has a
mortality of more than 50% and leads to progressive ventricular dilatation in 80% of infants. Many studies have
confirmed the significant association between Grade IV IVH, periventricular leukomalacia and ventriculomegaly
with cerebral palsy. One study gave an odds ratio (OR) of 15.4 (95% CI, 7.6-31.1) for a diagnosis of CP at 2 to
9 years of age in the presence of Grade IV IVH, PVL, or ventriculomegaly. Furthermore, Grade IV IVH and
moderate to severe ventriculomegaly were strongly associated with the risk of mental retardation or
neuropsychiatric disorders at 2 to 9 years, with the odds ratio ranging from 9.97 to 19.0 (154,155).
We believe that transport in utero, antenatal steroids, judicious obstetrical management, and expert delivery
room stabilization and NICU care are important preventative measures against IVH. Pharmacologic strategies,
specifically indomethacin administered within the first 12 hours of birth in ELBW infants, can reduce the
incidence of severe IVH (112), but without significant long-term benefit at 18 months and hence is not current
or recommended practice in our center (148). Further innovative research is needed to impact this serious
complication for ELBW infants.
Periventricular Leukomalacia
PVL is the other significant injury of the developing premature brain, with an incidence estimated to range from
4% to 15%. In our center, PVL is uncommon, with an incidence of 5.5%. It is believed to be a consequence of
hypoxic-ischemic events, leading to necrosis of the white matter (156). The most commonly affected areas are
the white matter near the trigone of the lateral ventricles and around the Foramen of Monro. Although often
diagnosed in association with IVH, PVL may also occur independently as an isolated lesion. Sometimes, the
origin is clearly intrauterine. Chorioamnionitis is recognized as a risk factor for PVL (156). Postnatally acquired
PVL is seen more frequently in male infants, in infants with severe RDS, in infants with septicemia, and in those
with significant cardiovascular instability or apnea (156).
Cystic PVL among preterm infants is the single best predictor of adverse long-term neurologic outcomes (157).
The frequency of cerebral palsy developing after cystic PVL has been reported to vary between 62 and 100%.
(158). Its characteristic manifestation is a spastic paresis involving predominantly the lower extremities (spastic
diplegia), and it is usually diagnosed in the first 2 years of life.
P.474
Although uncommon, severe disability with spastic quadriplegia, strabismus and impaired visual acuity,
developmental delays, cognitive impairment, and seizures may be observed in early childhood.
The diagnosis of PVL is made primarily by cranial ultrasonography. When the lesions have occurred in utero, it is
possible to make the diagnosis soon after birth, with the first or second studies. However, prenatally or
postnatally acquired PVL is usually diagnosed by 3 to 6 weeks or later, as time is necessary to develop the
cavitation of the injured periventricular white matter. This ultrasound, usually performed prior to discharge
home, provides important prognostic information for the neurodevelopment of the infant, facilitating parental
counseling and planning for long-term multidisciplinary follow-up.
Magnetic resonance imaging (MRI) may allow greater detection of white matter abnormalities than ultrasound.
However, there have been insufficient follow-up studies to indicate whether the additional findings provide more
information about neurodevelopmental prognosis. Therefore, the routine use of MRI and other techniques of
neuroimaging for the detection of PVL and other brain injury in the ELBW is not yet a recommendation, and
should be an avenue for future research. (155). Because of the later appearance of cystic lesions or residual
ventricular dilatation secondary to involution of the periventricular white matter and resorption of the cysts, it is
important to repeat the cranial ultrasound study between 36 and 40 weeks' postmenstrual age (155).
Seizures
Seizures are relatively rare in the ELBW infant, despite many potential risk factors, such as intracranial
hemorrhage, hypoglycemia, and electrolyte disturbances. Compared to full-term infants, seizures in very
premature infants are even more challenging to diagnose, due primarily to cortical immaturity coherent with the
gestational age. Subtle, tonic or myoclonic seizures may be difficult to differentiate from the general
uncoordinated movements, tremulousness and myoclonic jerks often seen in extremely premature infants. The
etiologies of seizures in the ELBW infant, as for the term infant, include CNS pathology, metabolic derangements
(e.g. hypoglycemia, hypocalcemia, severe hyponatremia), infection, and drug withdrawal.
The investigation and management of seizures is described comprehensively elsewhere. Suffice it to say that the
electroencephalogram (EEG) can be difficult to interpret, because a conventional surface EEG may not detect the
electrical activity of deeper cortical and subcortical structures. The treatment of seizures in ELBW infants
involves correction of any metabolic derangements, appropriate antibiotic therapy for infection as warranted,
and control of seizure activity to avoid brain injury as a result of alteration in cerebral energy metabolism. As for
term infants, the prognosis in the face of neonatal seizures depends primarily on the underlying cause. However,
the prognosis is generally worse for ELBW with seizures, compared to term infants. Our preference for medical
control of seizures is with phenobarbital at a loading dose of 20 mg/kg, which can be increased by another 10
mg/kg if control is not achieved (maximum loading dose 30-40 mg/kg). If control is not achieved, phenytoin is
added, at a loading dose of 15 mg/kg (maximum loading dose 30-40 mg/kg). Rarely, the use of sedativehypnotic anticonvulsants, such as lorazepam (0.05-0.1 mg/kg/dose) or diazepam (0.1-0.2 mg/kg/dose), and
rectal paraldehyde (0.3 ml/kg, diluted 1:2 in mineral oil) may be considered for resistant seizures. There should
be proper attention to adverse effects of these agents on very premature infants, such as the potential
exacerbation of jaundice secondary to bilirubin displacement from its albumin-binding site by the sodium
benzoate in diazepam and the abnormal movements (muscle twitchiness, myoclonus) seen with lorazepam
(159,160).
Hearing Impairment
ELBW infants are at increased risk for hearing impairment because of multisystemic illness and the frequent use
of potentially ototoxic medications, such as aminoglycosides and diuretics. The estimated prevalence of hearing
impairments in extremely immature (< 26 weeks gestation) survivors ranges from 1.7% to 3.8%. (161). In the
TIPP study, the prevalence of hearing loss requiring amplification for surviving ELBW infants was 2% (20/876
infants). (162)
Reporting on the outcomes of VLBW infants at age 19 years, Ericson and Kallen found high rates of persistent
hearing impairment, with the odds ratio for severe hearing loss being 2.5 (95% CI, 1.2-5.0) (163). In 1994, the
Joint Committee on Infant Hearing issued a Position Statement recommending that all infants born with birth
weight less than 1500 grams undergo auditory screening. (164). Early diagnosis of hearing loss and
amplification with hearing aids as early as 6 months of age, together with speech therapy, are essential to
reduce the progressive disability in speech and language development caused by hearing impairment. The tests
commonly used to detect hearing loss in the ELBW include the evoked auditory brainstem responses (ABR) and
otoacoustic emission (OAE). (165). In our unit, a hearing screen is obtained prior to discharge using the OAE,
with a diagnostic ABR confirming abnormal screening results.
Hematologic Disorders
Anemia
Low iron stores, multiple blood tests, blood loss as a result of either organ hemorrhage or hemolysis, and rapid
growth are some of the factors that make anemia a practically unavoidable hematologic complication for any
ELBW infant. The ELBW infant usually has a hemoglobin concentration of 140 to 160 g/L at birth. Those who
have suffered from IUGR may have a hemoglobin concentration as high
P.475
as 200 g/L. Their blood volume is 85 to 90 mL/kg. However, these values can be affected by the extent of
placental transfusion at delivery (166). We generally allow for a 10-second placental transfusion in ELBW infants
not suffering from IUGR.
The need for transfusion of blood products is a source of anxiety for parents. Any measure that can decrease the
frequency and severity of anemia should be implemented. These measures include limiting blood tests to those
essential for appropriate management of the infant, use of microtechniques, and use of pulse oximetry and
other transcutaneous devices to monitor partial pressure of oxygen (PO2) and PCO2.
Although the administration of erythropoietin with iron supplementation does not eliminate completely the need
for blood transfusion, it can reduce the number of transfusions (167,168). The lack of universal use of
erythropoietin is probably as a result of the fact that it is expensive and that, at least in sick infants, it does not
eliminate the need for blood transfusions.
To decrease the risk of infection, one unit of packed red cells from a single, properly screened donor, divided
into several small bags (satellite bags), could be used for the same infant for several weeks (169). We have
found this approach useful and more reassuring to parents.
We also have implemented a protocol for direct blood donation from compatible parents (father or mother) who
are cytomegalovirus, human immunodeficiency virus, and hepatitis B and C negative. Blood is irradiated prior to
transfusion to avoid graft-vs.-host disease. However, preparation of such blood requires time. The issue of blood
transfusion is discussed with the family, antenatally whenever possible, or soon after admission of the infant to
the NICU, and the parental decision is documented in the chart.
Our guidelines for the management of anemia with blood transfusion in ELBW infants are as follows: (a) infants
born with severe anemia and/or hypovolemic shock; (b) replacement of blood taken from an umbilical line in the
first days of life for frequent blood monitoring and exceeding 10% of the baby's blood volume; (c) maintenance
of hematocrit between 0.35 and 0.40 during the first week of life and between 0.30 and 0.35 during the second
week; (d) maintenance of hematocrit greater than 0.35 in infants with PDA and in those still having severe lung
disease in the second week of life; (e) in infants with CLD, we maintain the hematocrit between 0.30 and 0.35;
(f) after the second week of life, the hemoglobin is allowed to decrease as long as the baby has no signs of
symptomatic anemia such as poor feedings, high output cardiac failure, apnea, edema, failure to gain weight,
tachycardia, and tachypnea. Finally, 4 to 6 mg/kg/d of elemental iron is commenced at 4 to 6 weeks of life.
When transfusion is necessary, it should be administered slowly, especially during the first weeks of life, when
any acute change in blood volume may be translated into changes in cerebral blood flow, thus predisposing to
IVH, and when there is cardiorespiratory instability. We usually transfuse a volume of 10 mL/kg of packed red
blood cells, which may be repeated at 12 hours, according to the need. Furosemide 1 mg/kg can be given with
transfusions.
Hemostasis and Bleeding Diathesis

Vitamin K and vitamin-K-dependent coagulation factors are present at low concentrations at birth (170). As a
result, all of our ELBW infants receive 0.5 mg of vitamin K intramuscularly immediately after delivery.
Additionally, parenteral nutrition is supplemented with vitamin K.
Conditions requiring immediate administration of additional vitamin K include hemorrhagic pulmonary edema,
pulmonary or gastric hemorrhage, and disseminated intravascular coagulation. The management of these
conditions often will require, additionally to vitamin K, administration of fresh frozen plasma, transfusion of
platelets, and treatment of the underlying condition.
Thrombocytopenia (platelet count less than 100 109/L) is commonly observed in ELBW infants and, if severe
enough, may put the infant at risk for IVH. Thrombocytopenia is more frequent in infants born to preeclamptic
mothers and infants with IUGR (171). Accelerated platelet destruction is seen in infants with sepsis, indwelling
catheters, or active bleeding, or following exchange transfusion. In practice we transfuse platelets to infants with
a platelet count below 30 to 40 109/L. However, in case of active bleeding and a platelet count less than 60
109/L, transfusion of platelets should be considered.
Chronic Lung Disease

A large number of ELBW infants continue to require oxygen supplementation at one month of life, and many of
them will remain oxygen dependent beyond 36 weeks of postmenstrual age. Both dates have been proposed in
the literature to define CLD. It is also not rare to see a number of infants discharged on home oxygen programs.
Up to 70% of infants weighing less than 1000 g have been reported to develop CLD, and an important number
of survivors are discharged home on supplemental oxygen (172). Thus, for many ELBW infants, CLD seems
practically unavoidable.
It is beyond the scope of this chapter to discuss the pathogenesis and pathophysiology of CLD, and we refer the
reader to the appropriate chapter. Hence, we will limit our discussion to a few aspects of the problem that are
more specific to the ELBW infant. CLD encompasses more than the classic BPD described by Northway and
associates in 1969 (173). Over the years, we have observed two distinct groups of infants developing CLD (97).
The first group consists of infants with severe RDS, who require early intubation and mechanical ventilation. In
our center, close to 65% of these infants receive exogenous surfactant. This is followed by a rapid decrease in
oxygen requirement and need for ventilatory support. By day 3 of life, the majority of these infants are
breathing room air or require a minimal amount of supplemental oxygen. However, in most of these
P.476
infants, this improvement is transient, as ventilatory support and oxygen supplementation become necessary
after 7 to 10 days of life, with eventual resolution of the respiratory problem over the ensuing weeks. The period
of deterioration is usually accompanied by the presence of increased airway secretions, which require frequent
toilet of the airways. The second group is composed of infants who had no initial lung pathology, and who
required intubation for immaturity and/or apnea, but needed only minimal oxygen supplementation, if any. A
number of these infants go on to develop clinical and radiological signs of CLD. They follow the same pattern of
deterioration, and eventual improvement as the group with initial lung pathology. Fig. 25-1 indicates the
bimodal fluctuation in oxygen requirement among 120 survivors of birth weight less than 1000 g in our center
between 1993 and 1997.
Figure 25-1 Variation in O2 requirement in 120 survivors. RA, room air. 500-1000 g (1993-97).
In our population of ELBW survivors, 63.6% required oxygen supplementation at 28 days of age, and 31.2% at
36 weeks' postmenstrual age, with 5.9% of infants requiring supplemental oxygen after discharge home (Table
25-9).
The proposed management of the infant with CLD is based on a combination of the following interventions:
Respiratory support.
Acceptance of a PCO2 value up to 65 mm Hg, provided that the pH is at least around 7.25. Once the infant
attains a postmenstrual age of 35 weeks, we aim for a steady oxygen saturation above 90% on pulse oximetry
to prevent cor pulmonale. The hematocrit is generally kept above 0.35.
Prevention of fluid overload.

Early closure of the PDA, fluid restriction to 120 to 130 mL/kg/d, or even less in some cases, administration of
diuretics on a chronic basis, i.e., hydrochlorothiazide 1 mg/kg/dose and spironolactone 1 mg/kg/dose every 12
hours. Additional diuresis can be obtained with furosemide, when necessary.
Nutritional support.
It has been increasingly recognized that infants with CLD require particular attention to nutritional support.
These infants may fail to thrive as a result of an increase in energy expenditure associated with an increased
work of breathing, and their enteric intake may be suboptimal as a result of frequently associated feeding
intolerance and gastroesophageal reflux. Lung growth and repair of damaged pulmonary tissue requires an
adequate intake of all nutrients (174). We thus aim for an energy intake of 120 to 140 kcal/kg/d, encourage
early initiation of enteral feeds, and use diuretics as needed to eliminate excess fluid.
Control of nosocomial infections.

As ETT tubes frequently become colonized, it is important to periodically send secretions for culture and
sensitivity to antibiotics. However, we do not treat the colonized infant if the condition remains stable, but rather
monitor the infant carefully for clinical and laboratory signs compatible with infection. We may treat, however,
infants colonized with ureaplasma, mycoplasma, or chlamydia with erythromycin if the respiratory status
presents an unusual progression.
Control of inflammation.
Because an inflammatory response seems to be an important mechanism leading to CLD, corticosteroids had
been used extensively during the 1990s to decrease pulmonary edema, prevent inflammation, and increase
surfactant and antioxidant production. Systemic dexamethasone had been the drug of choice, and various
protocols were proposed in terms of onset and duration of treatment (175,176). Most infants responded
favorably, with rapid extubation and a significant decrease in oxygen requirement. Initially, the side-effects of
this therapy were believed to be limited to transient hypertension and hyperglycemia, an increased risk of
infection, reversible hypertrophic cardiomyopathy, and a flattening of the growth curve.
With the onset of the new millennium, new data became available regarding the long-term outcomes of steroidtreated infants. Of note were some serious concerns that corticosteroid therapy may be independently
associated with poor brain growth and with the
P.477
development of cerebral palsy (177). This new information has resulted in a joint statement by the AAP and
Canadian Pediatric Society, in which the routine use of systemic dexamethasone for the prevention or treatment
of CLD in VLBW infants is no longer recommended (178). We currently limit the administration of systemic
steroids to the most unstable ventilator-dependent infants with severe, life-threatening CLD, and the duration of
therapy is limited to a three-day pulse of dexamethasone of 0.1 to 0.2 mg/kg/day, after parental consultation.
It remains to be clarified whether less potent systemic corticosteroids, such as hydrocortisone, administered at
very low doses, may have a role to play in the future treatment of CLD (179). The same can be said of inhaled
steroid preparations such as budesonide, the efficacy of which has likely been limited to date by inadequate
methods of pulmonary delivery (180,181).
Bronchodilators.
In infants with decreased air entry and wheezing, we have found that the administration of nebulized salbutamol
at a dose of 1.25 mg diluted in 2.5mL physiologic saline solution, can be helpful (182). We have often observed
that, immediately after a treatment, air entry improves and clearance of secretions becomes easier, particularly
in combination with chest physiotherapy.
Home Oxygen Therapy

Despite maximum therapy, some infants remain dependent on supplemental oxygen beyond 40 weeks'
postmenstrual age. In our center, such infants are evaluated and followed after discharge by a pediatric
pulmonologist. The parents undergo specific training for home oxygen therapy. Prior to discharge, participation
is organized with a multidisciplinary follow-up program, which include in its membership the pediatric
pulmonologist, nutritionist and occupational therapist. For the years 1995 to 2002, the incidence of home
oxygen therapy in our surviving ELBW infants was 5.9%. In comparison, the incidence of supplemental oxygen
at discharge to home in the TIPP study was 18% to 20% (148).
Necrotizing Enterocolitis
NEC is the major GI disorder that selectively affects the sick premature infant. Its etiology is multifactorial and
includes suspected predisposing factors such as intestinal immaturity, poor intestinal motility, hypoxemia,
ischemia, PDA, umbilical catheter placement, IUGR, feeding practices, exchange transfusion, and systemic
infections (183).
Antenatal administration of steroids appears to accelerate intestinal maturation and provides additional
protection against NEC to the prematurely born infant (184). The incidence of NEC varies widely, from center to
center, and is estimated at between 9% and 25% for ELBW infants (185). Our incidence of surgical NEC
between 1995 and 2002 was 2.6% (Table 25-8). Variation in the incidence probably reflects differences in
diagnostic criteria and clinical practices. Indeed, the successful reduction in the incidence of NEC is related
essentially to prevention, by avoiding all known predisposing factors and by intervening and interrupting the
cascade of progression of the disease at the earliest signs, which many of us like to call pre-NEC. These signs
include increased gastric residuals, abdominal distension, cardiovascular instability, deterioration of skin
perfusion, clusters of apneic spells, and unexplained glycosuria and lipemia. In the presence of these signs and
depending on their severity, it is our practice either temporarily to discontinue feedings or to decrease the
volume by 50% and reevaluate the situation in a few hours. In the presence of further abdominal distension, we
do not hesitate to discontinue oral feeds and insert an orogastric catheter under continuous low suction. In the
majority of cases of benign distension, the intestinal decompression reestablishes the intestinal vascular supply
and, within 2 to 4 hours, the abdomen returns to normal. In an otherwise healthy looking and active infant, we
do not necessarily initiate therapy with antibiotics unless there is blood in the stools or the abdominal radiograph
demonstrates, besides dilated bowel loops, signs compatible with evolving NEC, such as pneumatosis
intestinalis. In these cases, a septic workup is performed and antibiotics are begun immediately. Our initial
treatment is a combination of gentamicin and ampicillin, to which clindamycin is added in case of additional
deterioration, or when the presentation is that of overwhelming NEC. One of the most distressing situations in
neonatology occurs when a previously stable premature infant, in whom milk intake has progressed well,
suddenly develops fulminant abdominal distension accompanied by sepsis, profound metabolic acidosis,
neutropenia, thrombocytopenia, and cardiovascular shock followed by rapid death. Fortunately, this dramatic
clinical presentation is rare, as it is difficult to anticipate or to prevent, let alone to treat.
The management of NEC, besides antibiotics and orogastric suction, requires close monitoring of vital signs,
frequent abdominal radiographs, with lateral views to visualize the possible presence of free air in the peritoneal
cavity, correction of metabolic abnormalities, and cardiovascular support with volume expanders and
vasopressors. Many infants with NEC may require assisted ventilation, as apnea frequently complicates the
situation. TPN is also an integral part of the management of these infants. Persistent intractable metabolic
acidosis, severe neutropenia and thrombocytopenia are ominous prognostic signs and usually reflect extensive
intestinal necrosis.
For infants managed medically, the time of reintroduction of oral feeds is a critical one. We usually reinitiate
enteric feeds after 7 to 14 days of therapy, depending on the rapidity of resolution of clinical and radiologic
signs. The presence of bowel sounds, a stable condition, and a well-perfused infant without significant apnea are
the basic requirements for starting oral feeds. We initially use an elemental formula and progress slowly to full
enteric feeds over a period of 7 days or longer, as tolerated. For those infants on maternal milk, mother's own
fresh milk is used. Post-NEC intestinal strictures are not unusual and can
P.478
present several weeks after the initial episode, with milk intolerance, vomiting, and abdominal distension.
Strictures may also be the result of subclinical injury of the intestine and can be seen in infants in whom the
diagnosis of NEC was never made before.
In summary, much still needs to be learned about NEC, its prevention, and its management. In the meantime,
every neonatal unit needs to develop its own philosophy and approach to reduce the risks of NEC. In our view,
strategies that decrease the risk of NEC include antenatal administration of steroids, use of mother's own milk,
early priming of the GI system with a very small volume of milk prior to the advancement of feedings, and,
above all, careful observation of the infant's condition, with reduction or temporary discontinuation of feeds and
intestinal decompression at the earliest suspicion.
Inguinal Hernias
Inguinal hernias occur frequently in ELBW infants, with a reported incidence between 14 and 30% (186,187). In
our center, the incidence of inguinal hernia diagnosed prior to discharge home ranges from 10% to 15%.
Predisposing factors are weakness of the abdominal musculature and tissues of the inguinal canal and increased
intraabdominal pressure, especially in the presence of CLD. They can present as early as 2 weeks of age, often
becoming very large and causing intermittent feeding intolerance, abdominal distension or bouts of crying and
irritability. Surgical repair must be planned in consultation with a Pediatric Surgeon, preferably prior to discharge
home where appropriate. Recurrence in the first year of life is a common observation.
Retinopathy of Prematurity
ROP remains a frequently diagnosed morbidity for ELBW infants and may result in significant visual impairment,
ranging from correctible myopia and astigmatism to bilateral blindness. The incidence and severity of ROP is
inversely proportional to birth weight and gestational age (188,189). Thus, as survival has increased, so has the
number of surviving ELBW infants with severe ROP, particularly in infants born at the limits of viability at 23 and
24 weeks. Severe ROP is defined as unilateral or bilateral stage 4 or 5 disease or disease requiring laser or
cryotherapy in at least one eye. Schmidt et al found an incidence of about 7% in a surviving ELBW cohort (162).
However, blindness has become a rare outcome, with an estimated prevalence of 2% in ELBW survivors (162).
The incidence and severity of ROP in our center is shown in Table 25-9.
The etiology and pathogenesis are complex and discussed in greater detail elsewhere in this book. Arterial
oxygen tension remains a great risk factor, despite the introduction of noninvasive continuous O2 monitors, such
as pulse oximetry, and the presumed tighter control of the arterial pO2. In addition to immaturity, we have
recently reported that the combination of IUGR and severe prematurity further increases the risk of developing
severe ROP (26).
Based on current knowledge of predisposing factors and years of careful observation, we have developed a
policy of keeping stringent control of the PaO2 below 50 mm Hg for all infants with birth weight less than 1,000
g in the first weeks of life. We have established guidelines for pulse oximetry, primarily based on gestational age
and birth weight. Our long-standing practice to accept lower limits of arterial oxygen saturation (SaO2) and to
avoid brisk variations is in keeping with recent publications supporting the safety of this practice in an attempt
to decrease the incidence of ROP (190). Furthermore, avoidance of swings from normoxemia to hyperoxemia
and hypoxemia has been reported to decrease the incidence of ROP (191). For ELBW infants, the upper limit of
SaO2 is set at 93% and we try to avoid rapid fluctuations in PaO2. We favor permissive hypercapnia, although
we avoid prolonged hypercapnia, with values of pCO2 greater than 55 mm Hg in the first weeks of life. We have
a low tolerance of recurrent apnea associated with significant oxygen desaturation. In such cases, we do not
hesitate to use respiratory support (nasal CPAP/ventilation; endotracheal ventilation) if the episodes are not
readily controlled by respiratory stimulant medication, e.g. caffeine.
Early diagnosis and treatment of prethreshold and threshold ROP is required to preserve maximal visual acuity
and prevent progression to blindness. This is achieved by expert serial ophthalmologic examinations as
described by the joint statement of the AAP, the American Association of Pediatric Ophthalmology and
Strabismus, and the Academy of Ophthalmology, such that all ELBW infants should be screened for ROP, with
the first examination at 4 weeks of postnatal age or between 31 and 33 weeks postmenstrual age (192).
Like others, we have observed a few unpredicted cases of rush disease developing early, between 30 and 33
weeks, or progressing at later corrected gestational ages after a period of quiescence at stage 1 or 2.
Consequently, in our center, we schedule the first ophthalmologic screen at 4 weeks after birth or at 30 weeks
of postmenstrual age, whichever comes first. This practice is coherent with a recent publication by Subhani and
associates (193), which recommended earlier screening in ELBW infants, beginning by 5 to 6 weeks of age using
the postnatal age criterion and not waiting for the corrected postmenstrual age. Follow-up examinations are
scheduled depending on the findings (e.g., degree of vascularization, zone and stage of retinal changes,
tortuosity, plus disease) at intervals of 1 to 2 weeks. If the disease process seems to accelerate, examinations
are performed twice a week. Infants with threshold and plus disease are candidates for laser therapy. In our
center, 6% of ELBW underwent laser therapy in the past few years with satisfactory results.
Apnea of Prematurity
Apnea of prematurity is a feature of nearly all infants with a birth weight less than 1,000 g. Its incidence and
frequency decrease with advancing gestational age, but at
P.479
times it may be seen up to 42 weeks of postmenstrual age (194). In the ELBW population, it is a frequent
indication for mechanical ventilation, thus exposing these infants to potential complications related to respiratory
support.
Apnea usually is defined as a cessation of breathing for 20 seconds or more, or of a shorter duration if
associated with cyanosis or bradycardia. Different patterns have been observed in premature infants: central
apnea (absent breathing movements), obstructive apnea (breathing movements, but no airflow) or mixed apnea
(central and obstructive) (195). ELBW infants are particularly prone to obstructive apnea, especially when in
supine position with the neck in the midline, because of the weakness of the muscles of the oropharynx. Apnea
as a result of obstruction of the lower airways also has been reported, suggesting immaturity of lung mechanics.
The cessation of gas exchange during a significant apneic episode is manifested by hypoxemia and/or
bradycardia. Recurrent episodes of apnea may affect neurodevelopmental outcome. Although it is difficult to
relate frequency and severity of apnea to outcome, one can only stress the importance of monitoring these
infants by pulse oximetry. Because apneic episodes can occur in premature infants as a result a variety of
underlying diseases, investigation of other pathologic causes must be undertaken before diagnosing apnea of
prematurity.
Patient management will depend on the severity and frequency of apneic episodes. Methylxanthines, which
stimulate the respiratory center, are the most effective pharmacologic treatment for apnea of prematurity.
Methylxanthines, aside from reducing the frequency of apneic pauses, have other actions that are equally
important. They increase respiratory rate, tidal volume, and minute ventilation, and they decrease
diaphragmatic fatigue. They also increase the sensitivity of the chemoreceptors to carbon dioxide and improve
blood pressure and cardiac output (196,197,198). Treatment with either aminophylline or theophylline is
effective but needs to be repeated two to four times per day. In our center, we treat apnea of prematurity with
caffeine, which is the metabolite of theophylline (199). We found that caffeine has fewer GI side effects and
causes less CNS irritability. It also has a much broader therapeutic index and achieves more stable plasma
levels. In view of its long half-life, it only needs to be given once daily. We use caffeine base at a loading dose of
10 mg/kg, followed 24 hours later by a single daily dose of 2.5 mg/kg. Caffeine can be administered
intravenously or orally. When given intravenously, the injection should be performed slowly, as it may otherwise
be quite painful. We verify the serum level of caffeine in cases of intractable apnea or in the presence of clinical
signs of toxicity. Caffeine is also the drug of choice when we start weaning an infant from the mechanical
ventilator. The therapeutic range of serum caffeine levels is very broad, between 23 and 104 micromol/L, and
the mean half-life is 102 hours.
If we cannot control apnea in a satisfactory manner with caffeine, we rapidly initiate nasal CPAP or nasal
ventilation, which, in combination with caffeine, offers very good stabilization in the vast majority of cases.
Finally, if apnea persists, we do not hesitate to intubate and ventilate such an infant with low pressures and
rates. In infants with persisting apnea and desaturations after 40 weeks' postmenstrual age, we perform a
respirogram prior to discharging them home on xanthines and a home monitoring program. It is important,
before closing the discussion on apnea, to underline the fact that pharmacologic treatment of apnea should be
considered only after proper investigation has eliminated any underlying condition requiring specific treatment,
such as anemia, infection, or metabolic disorders.
Neonatal Infections
The ELBW infant is particularly vulnerable to bacterial, viral, and fungal infections. A significant number of
premature labors are likely precipitated by infection. Chorioamnionitis is a frequent finding after a premature
birth, particularly in the presence of prolonged rupture of membranes. As the clinical signs of infection are often
nonspecific, the index of suspicion and the concern about the possibility of intrauterine infection should be very
high in the presence of a premature birth. Hence, screening for infection should be an integral part of the
evaluation of the ELBW infant. The diagnosis of neonatal infection can sometimes be difficult, as early neonatal
infection often manifests with respiratory symptomatology, which is also the overwhelming pathology of
prematurity. This is particularly true in the presence of group B streptococcal pneumonia, which is often
indistinguishable clinically and radiologically from RDS (200). However, early appearance of recurrent apnea,
poor perfusion, hypotension, and significant metabolic acidosis, often in the presence of an abnormal leukocyte
count, are very strong elements in favor of infection.
In symptomatic infants, we obtain a skin surface culture, blood culture, and leukocyte count, and we initiate
broad-spectrum antibiotic coverage with ampicillin and gentamicin. We do not routinely perform a lumbar
puncture on admission. However, if the blood culture is positive or if there is clinical evidence of deterioration
compatible with meningitis, then a lumbar puncture is performed. If the result suggests the presence of
meningitis, we adjust the duration of therapy, antibiotic coverage and dosage accordingly. If the infant's
condition improves rapidly, the blood culture is negative, and the acute phase reactants are normal, we
discontinue antibiotics after 2 to 5 days.
Nosocomial infections are not rare among infants whose hospitalization can be as long as 3 to 4 months. Aside
from the immaturity of the immune system, predisposing factors include ventilator care, intravenous
alimentation via central or peripheral lines, and exposure to extensive handling. In recent years, S. epidermidis
has emerged as the most common organism (201,202). However, in ventilated infants and in those with CLD,
Pseudomonas, Klebsiella, and S. aureus are the predominant organisms found. Fungal infections are not rare
and should be suspected in the presence
P.480
of unexplained thrombocytopenia, hyperthermia (203), and clinical signs of progressive deterioration.
It is common practice to initiate intravenous therapy with vancomycin and a third-generation cephalosporin on
suspicion of nosocomial sepsis, particularly when the infant is critically ill. One must, however, bear in mind that
the widespread use of antibiotics in any neonatal intensive care unit may rapidly result in the development of
drug-resistant bacterial strains (204,205,206). Another acceptable approach in more stable infants is to use a
combination of cloxacillin and an aminoglycoside such as gentamicin, with subsequent adjustment of antibiotic
therapy according to bacterial susceptibility testing results.
It is our policy not to treat colonization of ET tubes with antibiotics unless there are signs of pneumonia or
systemic infection. In the absence of signs of infection, the vast majority of our ventilated infants are left
without antibiotic coverage under close observation, with weekly monitoring of ETT colonization and sensitivity
to antibiotics. This information can be used in case of clinical deterioration suggesting infection.
Concern has been expressed regarding the risk of infection in infants treated in incubators with high humidity.
Our experience, using the new incubators providing up to 80 percent humidity, has generally been positive.
However, after the first 48 hours of life, we do reduce the humidity level to 60% to 65%.
In the past, considerable effort was placed on measures believed to protect the newborn infant from nosocomial
infections. These efforts included the restriction of visitors, and the use of gowns, gloves, masks, and hair nets.
However, no studies have shown any evidence supporting these measures (207). In our unit, despite no longer
using gowns for several years, we have not observed any change in the incidence of infection. We recommend
gowning only when parents handle their babies in their arms or in cases requiring strict isolation. Parental
visiting is unrestricted, and siblings are allowed to visit if they are healthy. In our view, the cornerstone of a
relatively low incidence of nosocomial infection has been the enforcement of a very strict program of hand
hygiene for both visitors and staff. In addition to routine hand-washing before and after patient contacts, an
alcohol-based hand gel is also available at the bedside for unanticipated interventions (208).
Pain Control, Sensory Overload, and Developmental Care
During their NICU stay, preterm infants are exposed to external stimuli that are very different from those
experienced by the fetus in utero: noise, lights, frequent disturbances, a nonliquid environment promoting
different body movements and postures, and pain. Although neonatal pain was largely ignored in the past, it is
now widely accepted that all newborns, including the ELBW, do experience pain. Furthermore, because
ascending pain pathways are well developed by 24 weeks, but because descending pathways with endogenous
opiates that may modulate incoming pain impulses are not present before 32 weeks, it is probable that the pain
experienced by the very premature infant may actually be more intense. (209). Appropriate management of
neonatal pain and stress is a very important aspect of modern neonatal care (210). Many nonpharmacologic
strategies have been proposed to minimize the pain and stress experienced by infants for minor and major
procedures, such as bundling, nonnutritive sucking, clustering of interventions and expertise in performing the
procedures. Pharmacologic options include the use of opiate and nonopiate analgesics and local anesthetics.
Topical anesthetics have not been fully evaluated for use in ELBW infants. Sedation and analgesia for elective
intubation maintains physiologic stability for the neonate and facilitates the procedure for both the neonate and
the operator (211). The safety of continuous opiate infusion early in life and for prolonged periods of time in
very premature infants is still in the phase of evaluation (212). Although oral sucrose does attenuate the pain
response for less invasive procedures such as heel sticks, i.v. insertion, and endotracheal suctioning, it remains
unclear if there are adverse long-term effects secondary to repeated sucrose treatments of premature infants
(213). Clearly, further research is required to better manage the stress and pain experienced by premature
infants, to understand the long-term effects on the growth and development of the premature brain from
episodes of stress and pain and exposure to exogenous opiates, and to demonstrate the long-term safety and
utility of strategies to control neonatal stress and pain. Our efforts to reduce neonatal pain should succeed
without inadvertent harm to the infants.
How the ex utero sensory input experienced by the premature infant affects the development of cerebral
pathways and systems, including programmed cell death or apoptosis, is not known. These pathways are
established after the period of neuronal migration, which is often completed around the 24th week of gestation.
Unfortunately this time is often marked by clinical instability in the newborn very premature infant, e.g., from
RDS, electrolyte disturbances, PDA, and limited energy and nutrient intakes, which may compound an adverse
effect on the development of the premature brain.
These concerns have been the impetus for the wide implementation by most NICUs of global developmental care
practices (e.g., control of light and noise, kangaroo care, massage and music therapy) and individualized
developmental care (e.g. Newborn Individualized Developmental Care and Assessment Program [NIDCAP]) to
minimize the stress, limit the sensory overload, and thereby optimize the health and neurodevelopment of each
infant and facilitate parent-child bonding (214). Many studies have been published about developmental care,
but because most have been limited by small samples, differing reported outcomes and other methodological
problems, there continues to be controversy regarding the utility of these practices. Systematic reviews of the
effectiveness of developmental care do confirm short-term benefits, such as improved growth outcomes, shorter
duration of
P.481
mechanical ventilation and oxygen supplementation, decreased length of stay and cost of hospitalization, and
improved neurodevelopmental outcomes to 9 to 12 months corrected age, but not at 24 months; effects at older
ages have not been reported (215,216).
In our nursery, the medical and nursing personnel have been sensitized to these issues over the years. The
bedside nurse is responsible for implementing and maintaining the developmental care and comfort of the
infant, with clustering of care to minimize external interventions and eliminate unnecessary disturbances,
although promoting close parent-infant bonding. Particular attention is given to the parents' needs during
hospitalization. Preparation for discharge involves the multi-disciplinary NICU and Neonatal Follow-Up teams, in
collaboration with the family. Parents' group meetings are held regularly, allowing free discussion of common
problems and concerns, and providing a ready support group.
Planning for Hospital Discharge

Preparing for hospital discharge is a momentous time for ELBW infants and their parents. This very much
anticipated step is also fraught with emotional upheavals, with some parents pushing for early discharge and
other parents clearly afraid to take their child home. It is the role of the NICU staff not only to judge the medical
readiness of the infant for discharge, but also the readiness of the parents and family, and to support and
educate the parents such that they are well prepared to take their infant home and that the discharge process is
orderly. The AAP recently published guidelines for hospital discharge of high-risk infants (217). Discharge
planning should be a multidisciplinary process with participation of the Neonatal Follow-Up team. The medical
complications should be reviewed with the parents and plans for specific follow-up organized. Parents should be
clear about who will be involved in their infant's care after they leave the NICU, who to call in emergencies, and
should be taught about worrisome signs that should prompt assessment of their infant by a health care
professional. Children with special needs, such as home oxygen therapy, will require even more planning,
preparation and clear instructions for the parents. We plan the first Follow-Up Clinic visit within 1 to 2 weeks of
discharge mainly to reassure parents that they are doing well, as we review the infant's weight gain, feeding
problems, and check the hemoglobin for late anemia. Thus, transition from the NICU to home should be planned
such that parents do not feel cut off from the security of the NICU, and such that they feel confident in their
ability to safeguard and nurture their precious infant.
FOLLOW-UP OF THE EXTREMELY LOW BIRTH WEIGHT INFANT

The ultimate goal of neonatal intensive care for ELBW infants is to improve survival without impairment or
disability attributable to the very premature birth. Fortunately, as survival has improved, follow-up studies have
not shown an increase in the rates of neurodevelopmental delays or neurosensory impairments, such as cerebral
palsy, blindness, deafness, or mental retardation (218,219,220). Although the majority of survivors do well, the
improved survival rate translates to an increase in the absolute number of infants with neurodevelopmental and/
or neurosensory problems and for whom long-term follow-up would be of benefit. Recent studies of outcomes of
ELBW and VLBW infants at school age, adolescence and early adulthood reveal high rates of school difficulties,
persistent neurosensory impairments, and health-related problems (221,222,223,224). Other studies describe
the positive perceptions and values of these survivors and their families (225,226,227,228,229,230).
In 1993, Allen and associates reported the 36 month corrected age survival and morbidity rates for infants born
between 22 and 25 weeks gestation (7). There were no survivors at 22 weeks, whereas 15%, 56%, and 80% of
infants survived at 23, 24, and 25 weeks, respectively. Only 2% of those born at 23 weeks survived without
severe intracranial abnormality. Intact survival was 21% at 24 weeks and 69% at 25 weeks. The authors
concluded that aggressive resuscitation is warranted for infants born at 25 weeks or greater, but not for those
born at 22 weeks or less. The Canadian Pediatric Society forwarded similar recommendations, with great
deference to parental wishes at 23 and 24 weeks gestation (231). The Neonatal Resuscitation Program (NRP)
emphasizes that the limit of viability is a moving target and up-to-date outcome data will assist in decisionmaking and care (232). An example of this is the change in the limit of viability in Japan, as stipulated in the
Eugenic Protection Law, from 24 to 22 completed weeks of gestation. Nishida and associates reported that ELBW
infants treated at the Tokyo Women's Medical College from 1984 to 2001 had a survival rate of 82%, with 50%
survival for infants with birth weights of 400 to 499 grams (9/18 infants), 50% survival at 22 weeks gestation
(3/6 infants) and low rates of neurosensory complications, such as cerebral palsy in 6.3% and mental
retardation in 17.3%, at the 6-year follow-up (233).
There are many publications about the early neonatal outcomes based on birth weight, which encompass inborn,
outborn, center-specific, and regional or population-based cohorts. With a regionalized perinatal program and
population-based data from Australia, Yu described significant improvements in the survival of ELBW infants,
increasing from 25% in 1979 to 56% by 1992 (234). Decreases in neurosensory deficits at 2 years of age were
also observed, with severe disability reduced from 18 to 11%, severe cerebral palsy from 3 to 1%, blindness
from 7 to 3% and deafness from 6 to 0.5%. The TIPP study described 1202 infants with birth weights of 500 to
999 grams. Survival to 18 months of age corrected for prematurity was 80%, with the following rates of
neurosensory impairments: cerebral palsy 12%, severe cognitive delay 27%, hearing loss requiring amplification
2%, and bilateral blindness 2%. Death after a postmenstrual age of 36 weeks occurred in 4% of the sample
(148). A review of the current literature by
P.482
Msall and associates provides a general summary of the rates of the major neurodevelopmental outcomes of
ELBW infants in the first 2 to 3 years: cerebral palsy 9% to 26%, blindness 1% to 15%, deafness 1% to 9%,
and cognitive disability 10% to 42% (235).
Early outcome data by gestational age is less encouraging. A regional study from Ontario, Canada, published in
1993, observed an absence of any handicap at 2 years of age in 50% of surviving infants born at 23 weeks,
41% of those born at 24 weeks, and 59% of those born at 25 weeks (6). In the same year, a regional study
from Oxford, England reported moderate-to-severe functional handicaps at 4 years of age in 80% of surviving
infants born at less than 25 weeks of age and in 66% of those born at 25 weeks (236). The large EPICure study
described the outcomes of infants born at 25 or fewer completed weeks in the United Kingdom and Ireland in
1995 (237,238). This study added more fuel to the controversy regarding care of the extremely premature
infants as they reported dismal survival and neurodevelopmental outcomes based on total births by gestational
age, including stillbirths and cases not resuscitated in the delivery room. The outcomes of 283 surviving infants
(92% follow-up rate) were described at a median age of 30 months: 49% had any disability, 23% had severe
disability, and 18% had cerebral palsy.
Studies at school age consistently show significant difficulties for the ELBW population. Hack and Fanaroff (219)
described the school-age outcomes in infants born weighing less than 750 grams, found that 56% of these
extremely small infants required special education classes, and concluded that infants with birth weights less
than 750 grams are expected to have significant problems with cognitive ability, psychomotor skills and
academic achievements. Gross and associates (239) found that ELBW at 10 years did less well in school than
term controls. They found that family factors, such as stable family units, were stronger predictors of school
performance than were perinatal complications. This study illustrates the importance of variables other than
prematurity and perinatal events on eventual cognitive outcome and also points to new directions for potential
intervention to impact on behavior and learning for ELBW infants.
Studies of outcomes into early adulthood primarily describe VLBW cohorts. Consistent findings were persistence
of chronic health problems and neurosensory impairments, smaller and lighter stature with an impact on
physical work capacity, difficulties in educational attainment and lower IQ even in the absence of neurosensory
impairment. (240,241) Bjerager and associates (228) describe the quality of life (QoL) at 18 to 20 years of age
for VLBW cohort born in Copenhagen between 1971 and 1974 and found that VLBW subjects free of handicaps
had a quality of life comparable to normal birth weight controls. Dinesen and associates (229) found that the
VLBW cohort born in 1980 to 1982 achieved higher QoL scores in comparison to the VLBW cohort born in 1971
to 1974.
Mazurier and associates (242) described a Canadian ELBW cohort at 16 to 21 years of age. They found that
ELBW subjects had a mean intelligence quotient (IQ) in the normal range (95 +/- 11) but lower than that of the
control group (107 +/-14), had significantly more school failures, and a lower rate of completion of secondary
school education (61% vs. 87%).
Monset-Couchard (243) and associates reported on the mid- and long-term outcomes of 166 premature SGA,
ELBW infants born in Paris, France. Language delays were observed at some stage in 31% of cases; behavioral
disturbances in 42% (with 12% having severe problems); cerebral palsy in 2%; visual deficits increased to 68%
with age; hearing losses after otitis media in 8%; 47% entered high school at their proper age and only 50% of
the older individuals obtained their baccalaureate in their 19th/20th year.
The limited number of studies describing outcomes of ELBW and VLBW infants into adolescence and adulthood
indicate that problems with health, education and adaptation may persist. It is unclear if the long-term
outcomes of ELBW and VLBW individuals without neurosensory impairment, whether SGA or not, differ
significantly from those of the normal-birth weight population. Furthermore, the current studies describe the
outcomes of infants born in the 1970s and 1980s, prior to the current innovations in neonatal practices such as
surfactant, alternative ventilation strategies, intensive nutrition and developmental care, and recent changes in
obstetrical practices that may have a positive impact on neonatal outcomes, which include higher utilization of
antenatal steroids, planned maternal transfers to tertiary care perinatal centers and intrapartum antibiotics.
Whether the current ELBW and VLBW graduates have the same risks of long-term difficulties as their
predecessors is not known. The importance of longitudinal follow-up of ELBW and VLBW infants and long-term
follow-up studies cannot be overstated.
Our own longitudinal follow-up of infants born between 22 and 25 weeks of gestation is on-going. From April
1988 to March 1994, 110 infants born between 22 and 25 weeks gestation were admitted to our NICU. There
were 58 survivors (survival rate 53%). In the first two years of life, problems with growth and general health
were noted, with 56% requiring at least one rehospitalization for hernia repair, ear tubes, eye surgery,
bronchiolitis-asthma or pneumonia. Reactive airway disease was seen in 59% and recurrent otitis media in 57%,
with the need for hearing aid in 6%. The Griffiths Scales of Mental Development scores at a corrected age of 24
months were in the normal range (mean GQ 96 4) and no infant scored in the mentally retarded range of less
than 70. Outcomes at 5 years of age were available for 53 infants. For this cohort of 53 infants (91% follow-up
rate), the mean gestational age was 24.9 0.7 weeks and the mean birthweight was 715 112 gms. There
was a progressive improvement of their general health, as shown in Table 25-12. The rates of neurosensory
impairments at 5 years of age were as follows: 5.2% had significant visual impairment, 5.5% required hearing
aids and 9.4% were diagnosed with cerebral palsy (1 child had quadriplegia and 4 had monoplegia). Normal
cognition was seen in 67%, with 12 children (23%) having mild deficits
P.483
and 5 children (9%) having severe deficits. Behavioral difficulties such as hyperactivity and attention-deficit
features were present in 28%, and speech delay in 37%. Overall, a higher incidence of deficits was present
among SGA children compared to AGA (67% vs. 37%). Microcephaly (head circumference <5th centile) was
found to be an important prognostic factor and was a predominant finding in SGA children (244).
TABLE 25-12 HEALTH IMPROVEMENT OVER 5-YEAR PERIOD AMONG 53 CHILDREN BORN BETWEEN
2225 wks GA
Re-hospitalization
Asthma
Otitis media
NICU Discharge to 2 Years
25 Years
Difference
p Value
56%
56%
53%
10.5%
31%
31%
44
25.5
25
0.01
0.04
0.06
The overall experience to date with ELBW infants allows us to come to the following conclusions:
The vast majority of the surviving ELBW infants can hope for a very meaningful life.
As survival of infants born less than 26 weeks has increased, there is an increase in the absolute numbers
of ELBW infants who are healthy, and also of infants with long-term complications.
A substantial number of ELBW infants will have physical, intellectual and behavioural impairments
resulting in significant functional disability, which may persist into adolescence and adulthood.
Because of the sequelae of extremely premature birth, it is important to have an organized, long-term
Follow-Up Program as part of the postdischarge care of ELBW infants, to insure appropriate diagnosis and
assist with therapy, resources and support to the child and family.
Evidence of significant brain injury, such as Grade III or IV IVH, cystic PVL and ventricular dilatation, is an
ominous predictor of future handicaps. Concomitant diagnosis of chronic lung disease at 36 weeks
corrected age and severe retinopathy of prematurity increase the likelihood of neurodevelopmental
problems. These provide important information to counsel the parents during care in the NICU and prior to
discharge home.
One of the on-going challenges for Neonatal Follow-up of ELBW infants is to advocate for appropriate resource
and service supports to these high-risk children and their families, and to insure longer-term follow-up, ideally
into school entry.
In the past, the exceptional circumstances of birth at extreme prematurity offered no chance for survival.
Although a very difficult road, fraught with major medical complications, parental stress and unanswerable
ethical dilemmas, modern neonatal care for these extremely premature and ELBW infants has afforded many
parents the opportunity to realize their hope of taking home a healthy infant with the potential for a good future.
FUTURE DIRECTIONS AND ETHICAL ISSUES
Although the birth weight barrier of 500 grams has been broken (2), the debate regarding the lower limit of
viability remains open (30,31).
Early follow-up of infants born with a weight less than 500 grams is not very reassuring (1). It is obvious that
beyond the heavy cost in human and financial resources in the NICU, these children will require extensive and
careful long-term evaluation and support. Their performance at school age and as young adults is not yet clear.
However, it seems unlikely at this point that their management will cease. Therefore, it is our responsibility as
neonatologists to provide short-and long-term accurate information to both parents and society, if we want to
defend our interventions with credibility. We need to establish general and institutional guidelines and to address
the ethical, financial and philosophical issues related to our interventions. Parental decisions need to be
respected, but parents need also to be well informed of the potential risks. It is likely that with improved medical
knowledge and technology, avoidance of major complications, and with better
P.484
nutrition, the outcome of all ELBW infants will improve. However, as we follow with great interest the future of
these children born at the limit of viability, it is important to focus our attention on improving their management
in our NICUs rather than on trying to break new barriers.
TABLE 25-13 MEASUREMENT OF DEVELOPMENT AND COGNITIVE FUNCTION OVER TIME FOR
INFANTS BORN BETWEEN 2225 WEEKS GA
Griffiths
GQ
6 Mo
Mean
Median
Range
a
103 7
103
78115
Griffiths
GQ
a
12 Mo
97 7
97
80121
Griffiths
Stanford-Binet
GQ
IQ
5 Years
24 Mo
96 4
97
78115
92 5
94
70110
Age corrected for prematurity; Griffiths GQ: Griffiths Mental Development Scales General Quotient
It is also evident that not all neonatologists, nor all NICUs, should undertake the treatment of ELBW infants. The
management of these infants requires great expertise and ample resources. A critical mass is essential to
maintain high medical and nursing standards. Ideally, only regional perinatal centers should be involved in the
care of ELBW infants, and treatment should be undertaken only when adequate resources are available, without
compromising the care of more mature infants having better chances of intact survival. Finally, the ultimate
decision regarding the management and degree of intervention in the delivery room remains the responsibility of
the neonatologist because, very often, there may not be time for multiple consultations and because, after all,
the neonatologist is the person with the wider experience to make such decisions. However, it is also important
that neonatologists involved in the care of the ELBW infant be sensitive to the wishes of the family and be
objective in their presentation of information and interpretation of outcome data to the family. The neonatologist
also should be prepared to advise parents when therapy has reached its limits, and should avoid heroic
interventions when the expected outcome, based on current scientific knowledge, is definitely unfavorable. In
our view, as we enter the new millennium, the moving target in the management of the ELBW infant should no
longer be the gestational age or the birth weight, but rather the condition of the infant at birth, his or her
potential to survive, the parents' desire, and our own honest evaluation of the capabilities, commitment, and
resources of our own environment to provide lengthy support to the infant and to his or her family.
ACKNOWLEDGMENTS
We would like to express our profound gratitude to the NICU nursing personnel for many years of competent
and devoted care of our ELBW infants. Our deep appreciation to Mrs. Judi Garon for her dedication and
secretarial expertise.
REFERENCES
1. Rowan CA, Lucey JF, Shiono P, et al. Fetal infants: the fate of 4172 inborn infants with birth weights of 401500 grams. The experience of the Vermont Oxford Network (1996-2000). Pediatr Res 2003;53:397A.
2. Fanaroff AA, Poole K, Duara S, et al. Micronates: 401-500 grams: the NICHD Neonatal Research Network
Experience 1996-2001. Pediatr Res 2003;53:398A.
3. Whitelaw A, Yu VYH. Ethics of selective non treatment in extremely tiny babies. Semin Neonatol
1996;1:297.
4. Fanaroff AA, Wright LL, Stevenson DK, et al. Very low birth weight outcomes of the National Institute of
Child Health and Human Development Neonatal Research Network, May 1995-December 1992. Am J Obstet
Gynecol 1995;173:1423.
5. Tudhope D, Burns YR, Grey TA, et al. Changing patterns of survival and outcomes at four years of children
who weighed 500-999 grams at birth. J Pediatr Child Health 1995;31:451.
6. Whyte HE, Fitzhardinge PM, Shennan AT, et al. Extreme immaturity: outcome of 568 pregnancies of 23-26
weeks gestation. Obstet Gynecol 1993;82:1.
7. Papageorgiou AN, Doray JL, Ardila R, et al. Reduction of mortality, morbidity and respiratory distress
syndrome in infants weighing less than 1000 grams by treatment with betamethasone and ritodrine. Pediatrics
1989;83:493.
8. Piecuch RE, Leonard CA, Cooper BA, et al. Outcome of extremely low birth weight infants (500-999 grams)
over a twelve year period. Pediatrics 1997;100:633.
9. Allen MC, Donohoe PK, Dusman AE. The limit of viability-neonatal outcome of infants born at 22-25 weeks
gestation. N Engl J Med 1993;329:1597.
10. Synnes AR, Ling EWY, Whitfield MF, et al. Perinatal outcomes of a large cohort of extremely low
gestational age infants (23-28 weeks gestation). J Pediatr 1994;125:925.
11. Ferrara TB, Hoekstra RE, Couser RJ, et al. Survival and follow-up of infants born at 23-26 weeks of
gestational age (effects of surfactant therapy). J Pediatr 1994;124:119.
12. Kitchen WH, Doyle LW, Ford WG, et al. Changing two year outcome in infants weighing 500-999 grams at
birth: a hospital study. J Pediatr 1991;118:938.
13. Joseph KS, Kramer MS, Allen AC, et al. Gestational age and birth weight-specific declines in infant
mortality in Canada, 1985-1994. Fetal and Infant Health Study Group of the Canadian Perinatal Surveillance
System. Paediatr Perinat Epidemiol 2000;14:332.
14. Pomerance JJ, Pomerance LJ, Gottlieb JA. Cost of caring for infants weighing 500-749 grams at birth.
Pediatr Res 1993;4: 231A.
15. McCormick MC, Bernabol JC, Eisenberg JM, et al. Cost incurred by parents of very low birth weight infants
after the initial neonatal hospitalization. Pediatrics 1991;88:533.
16. Bennett-Britton S, Fitzhardinge PM, Asby S. Is intensive care justified for infants weighing less than 801
grams at birth? J Pediatr 1981;99:937.
17. Bohin S, Draper ES, Field I. Impact of extremely immature infants on neonatal services. Arch Dis Child
1996;74:F110.
18. Saigal S, Zsatmari P, Rosenbaum P, et al. Cognitive abilities and school performance of extremely low
birth weight infants and matched term control children at eight years: regional study. J Pediatr 1991;118:751.
19. Harpe M, Taylor JG, Kline N, et al. School age outcomes of children with birth weights under 750 grams. N
Engl J Med 1994; 331:753.
20. Lahayne LA, Pine TR, Jackson C, et al. Outcome of infants weighing less than 800 grams at birth: 15 years
experience. Pediatrics 1995;96:479.
21. Saigal S, Rosenbaum P, Hattersley B, et al. Decreased disability rate among 3-year old survivors weighing
between 501 to 1000 gms at birth and born to residents of a geographically defined region from 1981 to 1984
compared with 1977 to 1980. J Pediatr 1989;114:839.
22. Halsey CR, Collin MF, Anderson CL. Extremely low birth weight children and their peers: a comparison of
preschool performance. Pediatrics 1993;91:807.
23. Halsey CR, Collin MF, Anderson CL. Extremely low birth weight children and their peers: a comparison of
school age outcomes. Arch Pediatr Adolesc Med 1996;150:790.
24. Zelkowitz P, Papageorgiou A, Zelazo P, et al. Behavioral adjustment of very low birth weight and normal
birth weight children. J Clin Child Psychol 1995;24:25.
25. Comit d'Enqute sur la mortalit et morbidit prinatale, Rapport pour 1998. Collge des Mdecins du
Qubec, 2002.
26. Bardin C, Zelkowitz P, Papageorgiou A. Comparison of outcomes of AGA and SGA infants born between 24
and 27 weeks gestation. Pediatrics 1997;100:1.
27. Vasa R, Vidyasagar D, Winegar A, et al. Perinatal factors influencing the outcome of 501 to 1000 gram
newborns. Clin Perinatol 1986;13:267.
28. Ott WJ. Small for gestational age fetus and neonatal outcome: reevaluation of the relationship. Am J
29. Crowley P, Chalmers I, Keirse MJ. The effects of corticosteroid administration before preterm delivery: an
overview of evidence from controlled trials. Br J Obstet Gynecol 1990;97:11.
P.485
30. American Academy of PediatricsCommittee on Fetus and Newborn, and American College of
Obstetricians and GynecologistsCommittee on Obstetric Practice. Perinatal care at the threshold of viability.
31. Fetus and Newborn Committee, Canadian Pediatric SocietyMaternal Fetal Medicine Committee, Society
of Obstetricians and Gynecologists of Canada. Management of the woman with threatened birth of an infant of
extremely low gestational age. CMAJ 1994;151:547.
32. Bottoms SF, Paul RH, Iams JD, et al. Obstetric determinants of neonatal survival: influence of willingness
to perform cesarean delivery on survival of extremely low birth weight infants. Am J Obstet Gynecol
1997;176:960.
33. Thibeault DW, Beatty EC, Hall RT, et al. Neonatal pulmonary hypoplasia with premature rupture of fetal
membranes and oligohydramnios. J Pediatr 1985;107:273.
34. Blott M, Greenough A. Neonatal outcome after prolonged rupture of the membranes starting in the second
trimester. Arch Dis Child 1988;63:1146.
35. Nimrod C, Davies D, Iwaniski S, et al. Ultrasound prediction of pulmonary hypoplasia. Obstet Gynecol
1986;68:495.
36. Vergani P, Locatelli A, Strobelt N, et al. Amnioinfusion for prevention of pulmonary hypoplasia in secondtrimester rupture of membranes. Am J Perinatol 1997;14:325.
37. Baker CJ. Group B Streptococcal infections. Clin Perinatol 1997;24:59.
38. Papageorgiou A, Desgranges MF, Masson M, et al. The antenatal use of betamethasone in the prevention
of RDS: a controlled double-blind study. Pediatrics 1979;63:83.
39. National Institutes of Health. Effects of corticosteroids for fetal maturation on perinatal outcomes.
Bethesda, MD: National Institutes of Health, 1994;12:1.
40. Ardila J, Le Guennec JC, Papageorgiou A. Influence of antenatal betamethasone and gender cohabitation
on outcome of twin pregnancies 24-34 weeks gestation. Semin Perinatol 1994; 18:15.
41. Mercer BM, Arheart KL. Antimicrobial therapy in expectant management of preterm premature rupture of
the membranes. Lancet 1995;346:1271.
42. Hillier SL, Nugent RR, Eschenbach DA, et al. Association between bacterial vaginosis and preterm delivery
of a low birth weight infant. N Engl J Med 1995;333:1737.
43. John C, Hauth MD, Robert L, et al. Reduced incidence of preterm delivery with metronidazole and
erythromycin in women with bacterial vaginosis. N Engl J Med 1995;333:1732.
44. Bhat R, Zikos-Labropoulou E. Resuscitation and respiratory management of infants weighing less than
1000 grams. Clin Perinatol 1986;13:285.
45. Kattwinkel J, ed. Textbook of neonatal resuscitation, 4th ed. American Academy of Pediatrics and
American Heart Association, 2000.
46. Vento M, Asensi M, Sastre J, et al. Resuscitation with room air instead of 100% oxygen prevents oxidative
stress in moderately asphyxiated term neonates. Pediatrics 2001;107:642.
47. Saugstad OD, Rootwelt T, Aalen O. Resuscitation of asphyxiated newborn infants with room air or oxygen:
an international controlled trial: the Resair 2 study. Pediatrics 1998;102:e1.
48. Hegyi T, Carbone T, Anwar M, et al. The Apgar score and its components in the preterm infant. Pediatrics
1998;107:77.
49. Strauss RG. AS-1 red cells for neonatal transfusions: a randomized trial assessing donor exposure and
safety. Transfusion 1996; 36:873.
50. Chathas MK, Paton JB, Fisher DE. Percutaneous central venous catheterization. Am J Dis Child
1990;144:1246.
51. Liossis G, Bardin C, Papageorgiou A. Comparison of risks from percutaneous central venous catheters and
peripheral lines in infants of extremely low birth weight: a cohort controlled study of infants > 1000 g. J
Matern Fetal Neonatal Med 2003;13:1.
52. Nowlen TT, Rosenthal GL, Johnson GL, et al. Pericardial effusion and tamponade in infants with central
catheters. Pediatrics 2002;110:137.
53. Nadroo AM, Lin J, Green RS, et al. Death as a complication of peripherally inserted central catheters in
neonates. J Pediatr 2001;138:599.
54. Pinto-Martin JA, Riolo S, Cnaan A, et al. Cranial ultrasound prediction of disabling and nondisabling
cerebral palsy at age two in a low birth weight population. Pediatrics 1995;95:249.
55. Nwaesei CG, Pape KE, Martin DJ, et al. Periventricular infarction diagnosed by ultrasound: a postmortem
correlation. J Pediatr 1984;105:106.
56. Rodriguez J, Claus D, Verellen G, Lyon G. Periventricular leukomalacia: ultrasonic and neuropathological
correlations. Dev Med Child Neurol 1990;32:347.
57. Gittermann MK, Fusch C, Gittermann AR, et al. Early nasal continuous positive airway pressure treatment
reduces the need for intubation in very low birth weight infants. Eur J Pediatr 1997;156.
58. Soll RF. Prophylactic synthetic surfactant for preventing morbidity and mortality in preterm infants
(Cochrane Review). In: The Cochrane Library, Issue 4, 2004. Chichester, UK: John Wiley & Sons, Ltd.
59. Sagstad OD, Bevilacqua G, Katona M, et al. Surfactant therapy in the newborn. Prenat Neonat Med
2001;6:56.
60. Merritt TA, Hallman M, Berry C, et al. Randomized, placebo- controlled trial of human surfactant given at
birth versus rescue administration in very low birth weight infants. J Pediatrics 1991;118:581.
61. Kendig JW, Notter RH, Cox C, et al. A comparison of surfactant as immediate prophylaxis and as rescue
therapy in newborns of less than 30 weeks' gestation. N Engl J Med 1991;324:865.
62. Dunn MS, Shennan AT, Zayack D, et al. Bovine surfactant replacement therapy in neonates of less than
30 weeks' gestation: a randomized controlled trial of prophylaxis versus treatment. Pediatrics 1991;87:377.
63. Morley CJ. Surfactant treatment for premature babies: a review of clinical trials. Arch Dis Child
1991;66:445.
64. TA, Hallman M, Vaucher Y, et al. Impact of surfactant treatment in cost of neonatal intensive care. J
65. Corbet A, Bucciarelli R, Zoldan S, et al. Decreased mortality rate among small premature infants treated
at birth with a single dose of synthetic surfactant: a multicenter controlled trial. J Pediatr 1001;118:277.
66. Kwong M, Egan E, Nutter RH, et al. Double blind clinical trial of calf lung surfactant extract for the
prevention of hyaline membrane disease in extremely premature infants. Pediatrics 1985; 76:585.
67. Kovacs L, Bardin C, Rossignol M, et al. Reduction in mortality but not in chronic lung disease after
surfactant therapy in infants < 1000 grams. Pediatr Res 1995;37:339A.
68. Ferrara TB, Hoekstra RE, Couser RJ. Effect of surfactant on outcome of infants with birth weight of 600750 grams. Pediatr Res 1991;27:243A.
69. Horbar ID, Wright EC, Oustand L, et al. Decreasing mortality associated with the introduction of surfactant
therapy: an observational study of neonates weighing 601-1300 grams at birth. Pediatrics 1993;92:191.
70. Vermont-Oxford Neonatal Network. A multicenter, randomized trial comparing synthetic surfactant with
modified bovine surfactant extract in the treatment of neonatal respiratory distress syndrome. Pediatrics
1996;97:1.
71. Brans YW, Escobedo MB, Hayashi RH, et al. Perinatal mortality in a large perinatal center: five year review
of 31,000 births. Am J Obstet Gynecol 1984;148:284.
72. Carlo WA, Stark AR, Wright LL, et al. Minimal ventilation to prevent Bronchopulmonary dysplasia in
extremely low birth-weight infants. J Pediatr 2002;141:370.
73. Wyszogrodski I, Kyei-Aboagye K, Taeush HW, et al. Surfactant inactivation by hyperventilation:

conservation by end-expiratory pressure. J Appl Physiol 1975;38:461.
74. Avery ME, Fletcher BD, Williams RG. The lung and its disorders in the newborn infants. Philadelphia: WB
Saunders, 1981.
75. Heneghan MA, Sosulski R, Alarcon MB. Early pulmonary interstitial emphysema in the newborn: a grave
prognostic sign. Clin Pediatr 1987;26:361.
76. Andreou A. One-sided high frequency oscillatory ventilation in the management of an acquired neonatal
lobar emphysema: a case report and review. J Perinatol 2001;25:61.
77. Cotten M. The science of neonatal high-frequency ventilation. Respir Care Clin N Am 2001;7:611.
78. Goldsmith JP, Karotkin E. Assisted ventilation of the neonate. Philadelphia: WB Saunders, 1996:25.
79. Thilo EH, Andersen D, Wesserstein MC, et al. Saturation by pulse oximetry: comparison of the results
obtained by instruments of different brands. J Pediatr 1993;122:620.
P.486
80. Heicher DA, Kasting DS, Harrod JR. Prospective clinical comparison of two methods for mechanical
ventilation of neonates: rapid rate and short inspiratory time versus slow rate and long inspiratory time. J
Pediatr 1981;98:957.
81. Greenough A, Greenall F, Gamsu H. Synchronous respiration: which ventilator rates are better? Acta
Paediatr Scand 1987; 76:813.
82. Castling D, Greenough A, Giffin F. Neonatal endotracheal suction: comparison of open and closed suction
techniques. Br J Intens Care 1995;5:258.
83. Avery ME, Tooley WH, Keller JB, et al. Is chronic lung disease in low birth weight infants preventable? A
survey of eight centers. Pediatrics 1987;73:20.
84. Vermont-Oxford Network Database Project. Very low birth weight outcomes for 1990. Pediatrics
1993;91:540.
85. Kraybil EN, Runyan DK, Bose CL, et al. Risk factors for chronic lung disease in infants with birth weight of
751 to 1000 grams. J Pediatr 1989;115:115.
86. Garland JS, Buck RK, Allred EN, et al. Hypocarbia before surfactant therapy appears to increase the
bronchopulmonary dysplasia risk in infants with respiratory distress syndrome. Arch Ped Adolesc Med
1995;149:617.
87. Feihl F, Perret C. Permissive hypercapnia. Am J Respir Crit Care 1994;150:1722.
88. Vannucci RC, Towfigh J, Heitjan DF, et al. Carbon dioxide protects the perinatal brain from hypoxic
ischemic damage: an experimental study in the immature rat. Pediatrics 1995;95:868.
89. Fujimoto S, Togari H, Yamanuchi N, et al. Hypocarbia and cystic periventricular leukomalacia in premature
infants. Arch Dis Child 1994;71:F107.
90. Graziani LJ, Spitzer AR, Mitchell DG, et al. Mechanical ventilation in preterm infants: neurosonographic
and developmental studies. Pediatrics 1992;90:515.
91. Wigglesworth JS, Pape KE. An integrated model for hemorrhagic and ischaemic lesions in the newborn
brain. Early Hum Dev 1978;2:179.
92. Wiswell TE, Graziani LS, Kornhaurser MS, et al. Effects of hypocarbia on the development of cystic
periventricular leukomalacia in premature infants treated with high frequency jet ventilation. Pediatrics
1996;98:918.
93. Courtney SE, Durand DJ, Asselin JM, et al. High- frequency oscillatory ventilation versus conventional
mechanical ventilation for very low birth weight infants. N Eng J Med 2002;347:643.
94. Johnson AH, Peacock JL, Greennough A, et al. High-frequency oscillatory ventilation for prevention of
chronic lung disease of prematurity. N Eng J Med 2002;347:633.
95. Priebe GP. High-frequency oscillatory ventilation in pediatric patients. Respir Care Clin Am 2001;7:633.
96. Herrera CM, Gerhardt T, Claure N, et al. Effects of volume- guaranteed synchronized intermittent
mandatory ventilation in preterm infants recovering from respiratory failure. Pediatrics 2002;110:529.
97. Le Guennec JC, Rufai M, Papageorgiou A. Spectrum of oxygen dependency in surviving infants weighing
600 to 1000 grams: decreased incidence of severe chronic lung disease. Am J Perinatol 1993;10:292.
98. Barrington KJ, Bull D, Finer NN. Randomized trial of nasal synchronized intermittent mandatory ventilation
compared with continuous positive airway pressure after extubation of very low birth weight infants. Pediatrics
2001;107:638.
99. DePaoli AG, Davis PG, Lemyre B. Nasal continuous airway pressure versus nasal intermittent positive
ventilation for preterm neonates: a systematic review and meta-analysis. Acta Paediatr 2003;92:70.
100. Clyman RI. Medical treatment of patent ductus arteriosus in premature infants. In: Long WA, ed. Fetal
and neonatal cardiology, Philadelphia: WB Saunders, 1990;682.
101. Gonzalez A, Ventura-Junca P. Incidence of clinically apparent ductus arteriosus in premature infants less
than 2000 g. Rev Chil Pediatr 1991;62:354.
102. Hoekstra RE, Jackson JC, Myers TF, et al. Improved neonatal survival following multiple doses of bovine
surfactant in very premature neonates at risk for respiratory distress syndrome. Pediatrics 1991;88:10.
103. Cotten RB, Stahlman MT, Kovar I, et al. Medical management of small preterm infants with symptomatic
patent ductus arteriosus. J Pediatr 1978;92:467.
104. Mellander M, Larsson LE, Ekstrm-Jodal B, et al. Prediction of symptomatic patent ductus arteriosus in
preterm infants using Doppler and M-mode echocardiography. Acta Paediatr Scand 1987;76:553.
105. Mahony L, Carnero V, Brett C, et al. Prophylactic indomethacin therapy for patent ductus arteriosus in
very-low-birth-weight infants. N Engl J Med 1982;306:506.
106. Rennie JM, Doyle J, Cooke RW. Early administration of indomethacin to preterm infants. Arch Dis Child
1986;61:233.
107. Fowlie PW. Prophylactic indomethacin: systematic review and meta-analysis. Arch Dis Child Fetal
Neonatal Ed 1996;74:F81.
108. Ment LR, Oh W, Ehrenkranz RA, et al. Low-dose indomethacin and prevention of intraventricular
hemorrhage: a multicenter randomized trial. Pediatrics 1994;93:543.
109. Robie DK, Waltrip T, Garcia-Prats JA, et al. Is surgical ligation of a patent ductus arteriosus the preferred
initial approach for the neonate with extremely low birth weight? J Pediatr Surg 1996; 31:1134.
110. Heymann MA. Prostaglandins and leukotrienes in the perinatal period. Clin Perinatol 1987;14:857.
111. Ment LR, Oh W, Ehrenkranz RA, et al. Low-dose indomethacin therapy and extension of intraventricular
hemorrhage: a multicenter study. J Pediatr 1994;124:951.
112. Van Overmeire B, Van de Broek H, Van Laer P, et al. Early versus late indomethacin treatment for patent
ductus arteriosus in premature infants with respiratory distress syndrome. J Pediatr 2001;138:205.
113. Schmidt B, Davis P, Moddemann D, et al. Long-term effects of indomethacin prophylaxis in extremelylow-birth-weight infants. N Engl J Med 2001;344:1966.
114. Mitchell JA, Akarasereenout P, Thiemermann C, et al. Selectivity of non-steroidal anti-inflammatory drugs
as inhibitors of constitutive and inducible cyclo-oxygenase. Proc Natl Acad Sci U S A 1993;90:11693.
115. Mosca F, Bray M, Lattanzio M, et al. Comparative evaluation of the effects of indomethacin and ibuprofen
on cerebral perfusion and oxygenation in preterm infants with patent ductus arteriosus. J Pediatr
1997;131:549.
116. Aranda JV, Varvarigou A, Beharry K, et al. Pharmacokinetics and protein binding of intravenous ibuprofen
in the premature newborn infant. Acta Paediatr 1997;86:289.
117. Varvarigou A, Bardin CL, Beharry K, et al. Early ibuprofen administration to prevent patent ductus
arteriosus in premature newborn infants. JAMA 1996;275:539.
118. Van Overmeire B. A comparison of ibuprofen and indomethacin for closure of patent ductus arteriosus. N
Engl J Med 2000; 343:674.
119. Mosca F. Comparative evaluation of the effects of indomethacin and ibuprofen on cerebral perfusion and
oxygenation in preterm infants with patent ductus arteriosus. J Pediatr 1997;131:549.
120. Gournay V, Savagner C, Thiriez G, et al. Pulmonary hypertension after ibuprofen prophylaxis in very
preterm infants. Lancet 2002;359:1486.
121. Hellerstein S. Fluids and electrolytes: physiology. Pediatr Rev 1993;14:70.
122. Guignard JP, John EG. Renal function in the tiny, premature infant. Clin Perinatol 1986;13:377.
123. Shaffer SG, Bradt SK, Meade VM, et al. Extracellular fluid volume changes in very low birth weight
infants during first 2 postnatal months. J Pediatr 1987;111:124.
124. Lorenz JM, Kleinman LI, Ahmed G, et al. Phases of fluid and electrolyte homeostasis in the extremely low
birth weight infant. Pediatrics 1995;96:484.
125. Takahashi N, Hoshi J, Nishida H. Water balance, electrolytes and acid-base balance in extremely
premature infants. Acta Paediatr Jpn 1994;36:250.
126. Sato K, Kondo T, Iwao H, et al. Internal potassium shift in premature infants: cause of nonoliguric
hyperkalemia. J Pediatr 1995;126:109.
127. Rutter N. The immature skin. Eur J Pediatr 1996;155:18.
128. Nopper AJ, Horil KA, Sookdeo-Drost S, et al. Topical ointment therapy benefits premature infants. J
Pediatr 1996;128:660.
129. Campbell JR. Systemic candidiasis in extremely low birth weight infants receiving topical petrolatum
ointment for skin care: a case-control study. Pediatrics 2000;105:1041.
130. Heird LW, Gomez MR. Parenteral nutrition in low birth weight infants. Annu Rev Nutr 1996;16:471.
P.487
131. Yeh TF, McClenan DA, Ayahi OA, et al. Metabolic rate and energy balance in infants with
bronchopulmonary dysplasia. J Pediatr 1989;114:448.
132. Merritt RJ. Cholestatic jaundice with total parenteral nutrition. J Pediatr Gastroenterol Nutr 1980;5:9.
133. Carver JD, Barness LA. Trophic factors for the gastrointestinal tract. Clin Perinatol 1996;23:265.
134. Berseth CL. Effect of early feeding on maturation of the preterm infant's small intestine. J Pediatr
1992;120:947.
135. American Academy of Pediatrics, Committee on Nutrition, 1998. Nutritional needs of preterm infants. In:
Kleinman R, ed. Pediatric nutrition handbook. Elk Grove Village, IL: American Academy of Pediatrics, 55.
136. Thurreen PJ, Melara D, Fennessey PV, et al. Effect of low versus high intravenous aminoacid intake on
very low birth weight infants in the early neonatal period. Pediatr Res 2003;53:24.
137. Berry MA, Conrod H, Usher RH. Growth of very premature infants fed intravenous hyperalimentation and
calcium-supplemented formula. Pediatrics 1997;100;647.
138. Ogata E. Carbohydrate metabolism in the fetus and neonate and altered neonatal glucoregulation.
Pediatr Clin North Am 1989; 33:25.
139. Pildes RS, Pyatis P. Hypoglycemia and hyperglycemia in tiny infants. Clin Perinatol 1986;13:351.
140. Pildes RS. Neonatal hyperglycemia. J Pediatr 1986;5:905.
141. Hewson M, Nawadra V, Oliver J, et al. Insulin infusion in the neonatal unit: delivery variation due to
adsorption. J Pediatr Child Health 2000;36:256.
142. Fuloria M, Friedburg MA, DuRant RH, et al. Effect of flow rate and insulin priming on the recovery of
insulin from microbore infusion tubing. Pediatrics 2000;105:915.
143. Ferrara TB, Couser RJ, Hoekstra RE. Side effects and long term follow-up of corticosteroid therapy in
very low birth weight infants with bronchopulmonary dysplasia. J Perinatol 1990;10:137.
144. Salle BL, Delvin EE, Lapillone A, et al. Perinatal metabolism of vitamin D. Am J Clin Nutr 2000;71:1317S.
145. Peters O, Ryan S, Matthew L, et al. Randomized controlled trial of acetate in preterm neonates receiving
parenteral nutrition. Arch Dis Child Fetal Neonatal Ed 1997;77:F12.
146. Odell GB, Storey GNB, Rosenberg LA. Studies in kernicterus: the staturation of serum protein bilirubin
during neonatal life and its relationship to brain damage at 5 years. J Pediatr 1970;76:12.
147. Shankaran S, Bauer CR, Bain R, et al. Prenatal and perinatal risk and protective factors for neonatal
intracranial hemorrhage. Arch Pediatr Adolesc Med 1996;150;491.
148. Schmidt B, Davis P, Moddemann D, et al. Long-term effects of indomethacin prophylaxis in extremelylow-birth-weight infants. N Eng J Med 2001;344:1966.
149. Batton DG, Holtrop P, DeWitte D, et al. Current gestational age-related incidence of major
intraventricular hemorrhage. J Pediatr 1994;125:623.
150. Volpe JJ. Current concepts of brain injury in the premature infant. Am J Radiol 1989;153:243.
151. Volpe JJ. Neurology of the newborn. Philadelphia: WB Saunders, 1995.
152. Dolfin T, Skidmore MB, Fongk W, et al. Incidence, severity and timing of subependymal and
intraventricular hemorrhages in preterm infants born in a perinatal unit as detected by serial real-time
ultrasound. Pediatrics 1983;71:541.
153. Perlman JM, Rollins N. Surveillance protocol for the detection of intracranial abnormalities in premature
neonates. Arch Pediatr Adolesc Med 2000;154:822.
154. Pinto-Martin JA, Whitaker AG, Feldman J, et al. Relationship of cranial ultrasound abnormalities in lowbirth-weight infants to motor or cognitive performance at 2, 6 and 9 years. Dev Med Child Neurol
1999;41:826.
155. Ment L, Bada HS, Barnes P, et al. Practice Parameters: neuroimaging of the neonate: report of the
Quality Standards Subcommittee of the American Academy of Neurology and the Practice Committee of the
Child Neurology Society. Neurology 2002;58:1726.
156. Perlman JM, Risser R, Broyles RS. Bilateral cystic periventricular leukomalacia in premature infants:
associated risk factors. Pediatrics 1996;97:822.
157. Levene MI. Cerebral ultrasound and neurologic impairment: telling the future. Arch Dis Child
1990;65:469.
158. Kuban KCK, Leviton A. Medical progress: cerebral palsy. N Engl J Med 1994;330:188.
159. Painter MJ, Alvin J. Neonatal Seizures. Cur Treat Options Neurol 1;3:237.
160. Sexson WR, Thigpen J, Stajich GV. Stereotypic movements after lorazepam administration in premature
neonates: a series and review of the literature. J Perinatol 1995;15:146.
161. Lorenz JM, Wooliever DE, Jetton JR, et al. A qualitative review of mortality and developmental disability
in extremely premature newborns. Arch Pediatr Adolesc Med 1998;152:425.
162. Schmidt B, Asztalos EV, Toberts RS, et al. Impact of bronchopulmonary dysplasia, brain injury and
severe retinopathy of prematurity on the outcome of extremely low-birth-weight infants at 18 months. JAMA
2003;289:1124.
163. Ericson A, Kallen B. Very Low birthweight boys at the age of 19. Arch Dis Child Fetal Neonatal Ed
1998;78:F171.
164. Joint Committee on Infant Hearing 1994. Position statement. Pediatrics 1995;95:152.
165. Kennedy CR, Kim L, Dees DC, et al. Otoacoustic emission and auditory brain stem responses in the
newborn. Arch Dis Child 1991;66:1124.
166. Berry MA, Conrod H, Usher RH. Growth of very premature infants fed intraveous hyperalimentation and
calcium-supplemented formula. Pediatrics 1997;100:647.
167. Phibbs RH. Erythropoietin therapy for the extremely premature infant. J Perinat Med 1995;23:127.
168. Soubasi V, Kremenopoulos G, Diamandi E, et al. In which neonates does early recombinant human
erythropoietin treament prevent anemia of prematurity? Results of a randomized controlled study. Pediatr Res
1993;34:675.
169. Strauss RG, Villhauer PJ, Cordle DG. A method to collect, store and issue multiple aliquots of packed red
blood cells for neonatal transfusion. Vox Sang 1995;68:77.
170. Lane PA, Hathawy WE. Vitamin K in infancy. J Pediatr 1985; 106:351.
171. Burrows RF, Andrew M. Neonatal thrombocytopenia in hypertensive disorders of pregnancy. Obstet
Gynecol 1990;76:234.
172. Hudak BB, Allen MC, Hudak ML, et al. Home oxygen therapy for chronic lung disease in extremely lowbirth-weight infants. Am J Dis Child 1989;143:357.
173. Northway WH Jr, Rosan RC, Porter DY. Pulmonary disease following respirator therapy of hyalinemembrane disease. Bronchopulmonary dysplasia. N Engl J Med 1967;276:357.
174. Vaucher YE. Bronchopulmonary dysplasia: an enduring challenge. Pediatr Rev 2002;23:349.
175. Avery GB, Fletcher AB, Kaplan M, et al. Controlled trial of dexamethasone in respirator-dependent infants
with bronchopulmonary dysplasia. Pediatrics 1985;75:106.
176. JJ, D'Eugenio DB, Gross SJ. A controlled trial of dexamethasone in preterm infants at high risk for
bronchopulmonary dysplasia. N Eng J Med 1989;320:1505.
177. Shinwell ES. Early postnatal dexamethasone treatment and increased incidence of cerebral palsy. Arch
Dis Child Fetal Neonatal Ed 2000;83:F177.
178. American Academy of Pediatrics, Committee on Fetus and Newborn and Canadian Paediatric Society,
Fetus and Newborn Committee. Postnatal corticosteroids to treat or prevent chronic lung disease in preterm
infants. Pediatrics 2002;109:330.
179. Watterberg KL, Gerdes JS, Gifford KL, et al. Prophylaxis against early adrenal insufficiency to prevent
chronic lung disease in premature infants. Pediatrics 1999;104:1258.
180. Kovacs L, Davis GM, Faucher D, et al. Efficacy of sequential early systemic and inhaled corticosteroid
therapy in the prevention of chronic lung disease of prematurity. Acta Paediatr 1998;87:792.
181. Shah V, Ohlsson A, Halliday HL, et al. Early administration of inhaled corticosteroids for preventing
chronic lung disease in ventilated very low birth weight preterm neonates (Cochrane Review). Cochrane
182. Kao LC, Warburton D, Platzker AC, et al. Effect of isoproterenol inhalation on airway resistance in chronic
bronchopulmonary dysplasia. Pediatrics 1984;73:509.
183. Covert RF, Neu J, Elliot MJ, et al. Factors associated with age of onset of necrotizing enterocolitis. Am J
184. Bauer CR, Morrison JC, Poole WK, et al. A decreased incidence of necrotizing enterocolitis after prenatal
glucocorticoid therapy. Pediatrics 1984;73:682.
185. Mauy RD, Fanaroff AA, Korones SB, et al. Necrotizing enterocolitis in very low birth weight infants.
Biodemographic and clinical correlates. J Pediatr 1991;119:630.
P.488
186. Uemura S, Woodward AA, Amerena R, et al. Early repair of inguinal hernia in premature babies. Pediatr
Surg Int 1999;15:36.
187. Kumar VH, Clive J, Rosenkrantz TS, et al. Inguinal hernia in preterm infants (< or + 32-week gestation).
Pediatr Surg Int 2002;18:147.
188. Ng YK, Fielder AR, Shaw DE, Levene M. Epidemiology of retinopathy of prematurity. Lancet 1988;2:1235.
189. Palmer EA, Flynn JT, Hardy RJ, et al. Incidence and early course of retinopathy of prematurity. The
Cryotherapy for Retinopathy of Prematurity Cooperative Group. Ophthalmology 1991;98:1628.
190. Tin W, Milligan DWA, Pennefather P, et al. Pulse oximetry, severe retinopathy, and outcome at one year
in babies of less than 28 weeks gestation. Arch Dis Child Fetal Neonatal Ed 2001; 84:F106.
191. Chow LC, Wright KW, Sola A. CSMC Oxygen Administration Study Group. Can changes in clinical practice
decrease the incidence of severe retinopaty of prematurity in very low birth weight infants? Pediatrics
2003;111:139.
192. Fierson WM, Palmer EA, Biglan AW, et al. Screening examination of premature infants for retinopathy of
prematurity. Pediatrics 1997;100:273.
193. Subhani M, Combs A, Weber P, et al. Screening guidelines for retinopathy of prematurity: the need for
revision in extremely low birth weight infants. Pediatrics 2001;107:656.
194. Poets C, Samuels M, Southall DP. Epidemiology and pathophysiology of apnea of prematurity. Biol
Neonate 1994;65;251.
195. Finer NN, Barrington KJ, Hayes BJ, et al. Obstructive mixed and central apnea in the neonate:
physiologic correlates. J Pediatr 1992;125:943.
196. Davi M, Shankaran K, Simons KJ, et al. Physiologic changes induced by theophylline in the treatment of
apnea in preterm infants. J Pediatr 1978;92:91.
197. Gerhardt T, McCarthy J, Bancalari E. Effect of aminophylline on respiratory center activity and metabolic
rate with idiopathic apnea. Pediatrics 1979;63:537.
198. Fesslova V, Caccano ML, Salice P, et al. Assessment of cardiovascular effects of theophylline in
premature newborns by means of echocardiography. Acta Paediatr Scand 1984;73:404.
199. Aranda JV, Turmen T, Davis J, et al. Effects of caffeine on control of breathing in infantile apnea. J
Pediatr 1983;103:975.
200. Baker CJ. Group B streptococcal infections. Clin Perinatol 1997;24:59.
201. Rubin LG, Sanchez PJ, Siegel J, et al. Evaluation and treatment of neonates with suspected late-onset
sepsis: a survey of neonatologists' practices. Pediatrics 2002;110:e42.
202. Stoll BJ, Hansen N, Fanaroff AA, et al. Late-onset sepsis in very low birth weight neonates: the
experience of the NICHD Neonatal Research Network. Pediatrics 2002;110:285.
203. Makhoul IR, Kassis I, Smolkin T, et al. Review of 49 neonates with acquired fungal sepsis: further
characterization. Pediatrics 2001;107:61.
204. Karlowicz MG, Buescher ES, Surka AE. Fulminant late-onset sepsis in a neonatal intensive care unit,
1988-1997, and the impact of avoiding empiric vancomycin therapy. Pediatrics 2000;106:1387.
205. Du B, Chen D, Liu D, et al. Restriction of third-generation cephalosporin use decreases infection-related
mortality. Crit Care Med 2003;31:1088.
206. Bryan CS, John JF Jr, Pai MS, et al. Gentamicin vs cefotaxime for therapy of neonatal sepsis. Relationship
to drug resistance. Arch Pediatr Adolesc Med 1985;139:1086.
207. Donowitz LG. Failure of the overgown to prevent nosocomial infection in a pediatric intensive care unit.
208. Harbarth S, Pittet D, Grady L, et al. Interventional study to evaluate the impact of an alcohol-based hand
gel in improving hand hygiene compliance. Pediatr Infect Dis J 2002;25:489.
209. Anand KJ. Clinical importance of pain and stress in preterm neonates. Biol Neonate 1998;73:1.
210. Joint Statement of the CPS and AAP. Prevention and management of pain and stress in the neonate.
Paediatr Child Health 2000;5:31.
211. DeBoer SL, Peterson LV. Sedation for nonemergent neonatal intubation. Neonatal Netw 2001;20:19.
212. Anand KJ, Barton BA, McIntosh N, et al. Analgesia and sedation in preterm neonates who require
ventilatory support: results from the NOPAIN trial. Neonatal Outcome and Prolonged Analgesia in Neonates.
Arch Pediatr Adolesc Med 1999;153:331.
213. Johnston CC, Filion F, Snider L, et al. Routine sucrose analgesia during the first week of life in neonates
younger than 31 weeks' postconceptional age. Pediatrics 2002;110:523.
214. Als H, Lawhon G, Duffy FH, et al. Individualized developmental care for the very low birth weight preterm
infant-medical and neuro-functional effects. JAMA 1994;272:853.
215. Symington A, Pinelli J. Developmental care for promoting de-velopment and preventing morbidity in
preterm infants (Cochrane Review). In: The Cochrane Library, Issue 4, 2004. Chichester, UK: John Wiley &
Sons, Ltd.
216. Jacobs SE, Sokol J, Ohlsson A. The Newborn Individualized Developmental Care and Assessment
Program is not supported by meta-analyses of the data. J Pediatr 2002;140:699.
217. Committee on Fetus and Newborn, American Academy of Pediatrics. Hospital discharge of the high-risk
neonateproposed guidelines. Pediatrics 1998;102:411.
218. Lee KS, Kim BI, Khoshnood B, et al. Outcomes of very low birth weight infants in industrialized countries:
1947-1987. Am J Epidemiol 1995;141:1188.
219. Hack M, Fanaroff AA. Outcomes of children of extremely low birthweight and gestational age in the
1990s. Semin Neonatol 2000;5:89.
220. O'Shea TM, Klinepeter KL, Goldstein DJ, et al. Survival and developmental disability in infants with birth
weights of 501 to 800 grams, born between 1979 and 1994. Pediatrics 1997;100:982.
221. Leonard CH, Piecuch RE. School age outcome in low birth weight preterm infants. Semin Perinatol
1997;21:240.
222. Hack M, Taylor HG, Klein N, et al. Functional limitations and special health care needs of 10- to 14-yearold children weight less than 750 grams at birth. Pediatrics 2000;106:554.
223. Hille ET, den Ouden AL, Saigal S, et al. Behavioural problems in children who weight 1000 g or less at
birth in four countries. Lancet 2001;357:1641.
224. Saigal, S, Hoult L, Streiner D, et al. School difficulties at adolescence in a regional cohort of children who
were extremely low birth weight. Pediatrics 2000;105:325.
225. Saigal S, Lambert M, Russ C, et al. Self-esteem of adolescents who were born prematurely. Pediatrics
2002;109:429.
226. Saigal S, Rosenbaum PL, Feeny D, et al. Parental Perspectives of the health status and health-related
quality of life of teen-aged children who were extremely low birth weight and term controls. Pediatrics
2000;105:569.
227. Taylor HG, Klein N, Minich NM, et al. Long-term family outcomes for children with very low birth weights.
Arch Pediatr Adolesc Med 2001;155:155.
228. Bjerager M, Steensberg J, Greisen G. Quality of life among young adults born with very low birthweights.
Acta Paediatr 1995;84:1339.
229. Dinesen SJ, Greisen G. Quality of life in young adults with very low birth weight. Arch Dis Child Fetal
230. Saigal S, Feeny D, Furlong W, et al. Comparison of the health-related quality of life of extremely low
birth weight children and a reference group of children at age eight years. J Pediatr 1994; 125:418.
231. Canadian Pediatric Society Statement on Resuscitation Management of the woman with threatened birth
of an infant of extremely low gestational age. A Joint Statement with the Society of Obstetricians and
Gynaecologists of Canada and the Canadian Paediatric Society. CMAJ 1994;151:547.
232. Kattwinkel J, ed. Textbook of neonatal resuscitation, 4th ed. American Academy of Pediatrics and
American Heart Association, 2000:7.
233. Nishida H. Marginally viable, fetal infants-who is too young or small to live: Japanese experience. Ross
Special Conference. Hot Topics 2002 in Neonatology Syllabus, Dec 8-10, 2002, Washington DC.
234. Yu VYH. Developmental outcomes of extremely preterm infants. Am J Perinat 2000;17:57.
235. Msall ME, Tremont MR. Measuring functional outcomes after prematurity: developmental impact of very
low birth weight and extremely low birth weight status on childhood disability. Ment Retard Dev Disabil Res
Rev 2002;8:258.
236. Johnson A, Townshed P, Yudkin P, et al. Functional abilities at age 4 years of children born before 29
weeks of gestation. Brit Med J 1993;306:1715.
237. Costeloe K, Hennessy E, the Epicure Study Group, et al. The EPICURE Study: outcomes to discharge
from hospital for infants born at the threshold of viability. Pediatrics 2000;104:659.
P.489
238. Wood NS, Marlow N, the EPICure Study Group, et al. Neurologic and Developmental Disability after
Extremely Preterm Birth. N Eng J Med 2000;343:378.
239. Gross S, Mettelman BB, Dye TD, et al. Impact of family structure and stability on academic outcome in
preterm children at 10 years of age. J Pediatr 2001;138:169.
240. Hack M, Flannery DJ, Schuchter M, et al. Outcomes in young adulthood for very-low-birth-weight infants.
N Eng J Med 2002;346:149.
241. Ericson A, Kallen B. Very low birthweight boys at the age of 19. Arch Dis Child Fetal Neonatal Ed
1998;78:F171.
242. Mazurier E, Lefebvre F, Tessier R. Educational achievement and intelligence at 16-25 years of exprematures born at <1000 g. Pediatr Res 1999;45:25-A(abst).
243. Monset-Couchard M, de Bethmann O, Kastler B. Mid- and long-term outcome of 166 premature infants
weighing less than 1000 g at birth, all small for age. Biol Neonate 2002;81: 244.
244. Piuze G, Bardin C, Papageorgiou A. Comparison of outcome at 5 years of age of SGA & AGA infants <28
weeks of gestation: a case control study. Pediatr Res 2001;49:334A.
Chapter 26
Intrauterine Growth Restriction and
the Small-For-Gestational-Age Infant
Marianne S. Anderson
William W. Hay
INTRODUCTION
Much of the interest in infants who are small for gestational age (SGA) at
birth, and much of the impetus for studying intrauterine growth
restriction (IUGR) that produces SGA infants, began with the observation
by pediatricians and neonatologists that newborn infants who were
classified according to birth weight as small, average, or large for
gestational age (SGA, AGA, and LGA, respectively) showed specific
morbidities and rates of death that were unique to each of these birth
weight-gestational age classifications (1) (See Appendix D for
intrauterine growth charts). SGA infants were recognized as having more
frequent problems with perinatal depression (asphyxia), hypothermia,
hypoglycemia, polycythemia, long-term deficits in growth,
neurodevelopmental handicaps, and higher rates of fetal and neonatal
mortality (2) (Fig. 26-1). Although there have been tremendous
improvements in perinatal diagnosis and treatment, severe IUGR and the
birth of markedly SGA infants continue to be frequent problems, and the
perinatal morbidity and mortality rates of IUGR fetuses and SGA infants
continue to exceed those of normal fetuses and infants.
DEFINITIONS
Small for Gestational Age
SGA infants are classically defined as having a birth weight that is more
than two standard deviations below the mean or less than the 10th
percentile of a population-specific birth weight vs. gestational age plot.
Broader definitions include less than normal anthropometric indexes,
such as length and head circumference, and marked differences between
growth parameters, even when they are within the normal range. For
example, an infant can be considered relatively SGA when its weight is

at the 25th percentile, but its length and head circumference are at the
75th percentile. In this case, the weight/length ratio (or the Ponderal
index = [weight (g)]/[length (cm)]3) is less than normal, indicating that
growth rates of adipose tissue and skeletal muscle, the principal
determinants of weight, were less than normal (Fig. 26-2) (3).
Intrauterine Growth Restriction

IUGR is defined as a rate of fetal growth that is less than normal for the
population and for the growth potential of a specific infant. IUGR
therefore produces infants who are SGA. SGA infants can be the result of
normal but slower than average rates of fetal growth, such as those
constitutionally small but not abnormal infants whose parents, siblings,
and more distant relatives are small (4). SGA infants also can be the
result of abnormally slow fetal growth that is caused by pathophysiologic
conditions or diseases. Because growth is one of the essential features of
the fetus, nearly any aberration of biologic activity in the fetus can lead
to growth failure. Thus, small size at birth can be either a normal
outcome or one that is a result of intrinsic or extrinsic factors that limit
fetal growth potential.
Birth Weight Classification of Growth

Many terms are used to describe variations in fetal growth. For example,
human newborns are classified as having normal birth weight (greater
than 2,500 g), low birth weight (LBW, less than 2,500 g), very low birth
weight (VLBW, less than 1,500 g), or extremely low birth weight (ELBW,
less
P.491
than 1,000 g). Obviously, classification by weight alone says little about
fetal growth rate, as most infants with less than normal birth weights are
the result of a shorter than normal gestation, i.e., they are preterm.
Similarly, classifying newborns as preterm or term on the basis of birth
weight is erroneous, as infants with IUGR are smaller than normal at any
gestational age.
Figure 26-1 Morbidities specific to small-for-gestational-age infants.

(Adapted from Lubchenco LO. The high risk infant. In: Schaffer AJ,
Markowitz M, eds. Major problems in clinical pediatrics, volume XIV.
Philadelphia: WB Saunders, 1976:6.)
Normal Variations and the Assessment of

Fetal Growth
Normal fetal growth varies almost twofold. For example, mean birth
weight for neonates born in New Guinea is 2,400 g (5), whereas normal
birth weights in other populations can exceed 4,000 g (6). Such
variations are related to genetic and environmental factors, the latter
usually reflecting local diets (e.g., the relatively obese infants of
Polynesian women who eat relatively large amounts of starchy foods in
their normal diet). These and other normal anthropometric variations
must be considered in relation to the diagnosis of IUGR in fetuses and

SGA status of newborns.
Symmetric and Asymmetric Growth

Restriction
SGA infants have been classified as having symmetric or asymmetric
IUGR. Symmetric IUGR implies that brain and body growth both are
limited. Asymmetric growth indicates that body growth is restricted to a
much greater extent than head (and thus, brain) growth. In such cases,
brain growth is considered spared. Mechanisms that allow brain growth
to continue at a faster rate than adipose tissue and skeletal muscle are
not completely known. Contributing factors may include an increased
rate of cerebral blood flow relative to the umbilical and systemic
circulations, which has been observed in some of these infants (7). In
some experimental models, cerebral glucose transporter concentrations
are preserved despite fetal hypoglycemia, indicating preservation of
cerebral glucose uptake capacity (8). The most severely affected infants
have marked reductions in both brain and body growth. Even moderately
IUGR infants have growth restrictions of both brain and body, but to
varying degrees that depend on the duration and severity of the insults
that inhibit growth. Asymmetric and symmetric growth restriction are
best thought of, therefore, as extreme examples (Fig. 26-3). Asymmetric
fetal growth differentially affects organs other than the brain. As shown
in Fig. 26-4, the heart also is larger for body weight in these infants,
whereas the liver, probably representing glycogen deficit, and thymus,
perhaps indicating a response to stress but also showing a potential for
immunologic inadequacy, are smaller for body weight.
In general, factors intrinsic to the fetus cause symmetric growth
restriction, whereas external factors cause asymmetric growth. Patterns
of symmetric growth restriction develop early during fetal life, reflecting
their intrinsic nature. Asymmetric patterns also can develop as early as
the second trimester (9) and as much as 30% to 50% of extremely preterm neonates (less than 1,000 g) are SGA, probably reflecting pathology
that produced growth restriction and led to preterm birth. Factors that
limit the growth of both the fetal brain and body include chromosomal
anomalies (e.g., parti-cularly trisomy conditions), congenital infections

(toxoplasmosis, rubella, cytomegalovirus), dwarf syndromes, some
inborn errors of metabolism, and some drugs (10). The mechanisms by
which these factors limit fetal growth are multifactorial.
Asymmetric growth restriction classically develops during the late second
and third trimesters when fetal nutrients, particularly glucose and lipids,
increasingly contribute to energy storage in the form of glycogen and fat
(in both brown and white adipose tissue) (11). Slight reductions of
energy substrate supply to the fetus limit fat and glycogen storage and
the growth of skeletal muscle, but allow for continued bone and brain
growth. More extreme limitations of energy substrates, for longer
periods, affect both growth and energy storage. Timing is important;
with decreased nutrient supply early in gestation, growth of all body
organs is restricted, whereas decreased fetal nutrient supply later in
gestation primarily restricts growth of glycogen content, adipose tissue,
and skeletal muscle.
INTERPRETATION OF FETAL GROWTH CURVES

Growth Curves Based on Neonatal
Measurements
Cross-sectional growth curves have been developed from anthropometric
data in populations of infants born at
P.492
different gestational ages (12). Such curves have been used to
demonstrate whether an infant's weight is within the normal range for a
given gestational age and thus to estimate whether that infant's in utero
growth was greater or less than normal. The normal range is defined as
birth weights between the 10th and 90th percentile of the populationspecific birth weight vs. gestational age relationship. Fetuses and
neonates who are within the 10th and 90th percentiles for weight vs.
gestational age are considered AGA. Those who are less than the 10th
percentile are considered SGA. Those who are greater than the 90th
percentile are considered LGA. Other terms for SGA infants include light
for dates and small for dates.
Figure 26-2 The Colorado Intrauterine Growth Charts, including symbols that define
the anthropometric measurements for the three infants shown in Fig. 26-2. (O)
Preterm infant at 34 weeks of gestation, showing asymmetry of weight (15th
percentile) versus length and head circumference (75th percentile) and weight-tolength ratio (85th percentile); () severely but symmetrically small-for-gestationalage infant at 39 weeks, showing weight, length, head circumference, and weight-tolength ratio all about equally and markedly (<10th percentile); and () symmetric
average-for-gestational-age infant at 40 weeks, showing weight, length, head
circumference, and weight-to-length ratio about the 65th to 75th percentile. (Growth
charts from Lubchenco LO, Hansman C, Boyd E. Intrauterine growth in length and
head circumference as estimated from live births at gestational ages from 26 to 42
weeks. Pediatrics 1966;37:403 (ref. 13).
Most growth curves usually are confined to the third trimester. Each
curve is based on local populations with variable composition of maternal
age, parity, socioeconomic status, race, ethnic background, body size,
degree of obesity or thinness, health, pregnancy-related problems, and
nutrition,
P.493
and the number of fetuses per mother, the number of infants included in
the study, and by what methods and how accurately measurements of
body size and gestational age were made. Estimating gestational age, in
particular, has considerable error. Such error is derived from variability
in dating conception because of maternal postimplantation bleeding and
irregular menses, wide variability in the development physical features of
maturation in the infant, and interobserver variability in assessing an
infant's developmental stage. The growth curves shown in Fig. 26-2 are
those of Lubchenco and colleagues (13) in Denver, Colorado, published in
1966. They are biased to slightly lower birth weights compared with
many other growth curves, especially close to term, as a result of the
unique mix of racial and ethnic groups in the population of babies who
were born at Colorado General Hospital, Denver, Colorado, at the time
the data were collected. Although the higher altitude of Denver (5,280
feet or ~1,600 m) has been considered a factor in the smaller birth
weights shown in these curves, the independent effect of high altitude on
restricting fetal growth is not clearly demonstrable at 1 mile (1.6 km). In
fact, growth curves similar to those of Lubchenco et al. have been
produced at sea level among lower socioeconomic groups with a high
proportion of blacks and Hispanics in the population (Fig. 26-5).
Mathematical analyses of various fetal growth curves have been used to
determine growth rates over relatively short gestational periods or at
discrete gestational ages (12). For example, the data used in the
Lubchenco growth curves (Figs. 26-2 and 26-5) can be approximated by
a simple exponential function showing fetal weight increasing at about 15
g/day/kg. This rate will vary from the smallest to the largest infants. For
a given weight percentile, however, there are only differences of 1% to
2% for this exponential function among different populations and
studies.
Figure 26-3 Preterm, small-for-gestational-age infant at 34 weeks of

gestation (left), severely small-for-gestational-age infant at 39 weeks
(middle), and average-for-gestational-age infant at 40 weeks (right).
Growth Curves Based on Fetal Measurements

Fetal growth curves also have been developed from serial ultrasound
measurements of fetuses who subsequently were born at term in healthy
condition and with normal
P.494
measurements, providing continuous rather than cross-sectional indexes
of fetal growth. These curves can be better correlated with the expected
fetal growth rate of a particular fetus than can cross-sectional,
population-based growth curves. Serial ultrasound measurements of fetal
growth also can more accurately determine how environmental factors,
such as severe maternal illness and under nutrition, can inhibit fetal
growth. Figure 26-6 shows an example of ultrasound data for fetal
growth curves, including evidence for a particular fetus whose growth
rate clearly was affected adversely by poor maternal health and

beneficially by improved maternal nutrition. Further, fetal ultrasoundderived growth curves show less of the midgestational exponential
increase in fetal growth rate that is typical of cross-sectional growth
curves derived from neonatal measurements at different gestational
ages. This observation strengthens the concept that, because preterm
birth is not a normal outcome, one must be suspicious of growth
parameters and growth curves derived from cross-sectional
measurements of preterm infants assessed at birth. The intrauterine
growth of these preterm infants probably was affected adversely by the
same pathologic factors that led to their preterm birth. Thus, there
probably is no ideal fetal growth curve derived from postbirth, crosssectional measurements. Future growth curves to assess in utero growth
of a specific newborn should be based instead on more thoroughly and
accurately determined fetal growth parameters from ultrasound
measurements in pregnancies with definitely known dates of conception
and birth at term of normally grown and developed infants.
Figure 26-4 Body length and selective organ weight percent changes
from normal preterm infants of similar birth weights in necropsy series of
IUGR fetuses. (After Gruenwald P. Growth patterns in the normal and
deprived fetus. In: Jonxis JHP, Visser HKA, Troelstra JA, eds. Aspects of
prematurity and dysmaturity. Springfield: Charles C. Thomas, 1968:43.)
Figure 26-5 Birth weights by ges-tational age from six sources.

(Adapted from Naeye R, Dixon J. Distortions in fetal growth standards.
Pediatr Res 1978;12:987).
INTRAUTERINE GROWTH RESTRICTION AND

PRETERM BIRTH
Most cases of fetal growth restriction represent only minimal growth
delay and are natural, reproductively successful, although not perfect,
adaptations to nutrient limitation. Most cases of IUGR, therefore, are not
major causes of preterm delivery, and fetal growth rate and length of
gestation usually are not related. In cases of severe IUGR, the
pathophysiologic processes causing the IUGR also can lead to preterm
labor and preterm delivery. Thus, IUGR frequently occurs with a variety
of maternal conditions that are associated with preterm delivery (Table
26-1) (14,15,16,17).
Established maternal conditions that are associated with both IUGR and
preterm delivery include very low maternal prepregnancy weight, prior
preterm delivery, cigarette smoking, indirect effects of very young or
advanced maternal age, and lower maternal socioeconomic status (14).
Regarding race, African-American women who were born in the United
States have a twofold greater incidence of both preterm birth and IUGR
than do white women from the United States or African-American women
who emigrated from Africa. Reasons for this are multifactorial and
include nearly all of the generally associated risks
P.495
P.496
and causes of IUGR and preterm delivery (15). Stretch-activated
mechanisms probably induce preterm labor in cases of multiple
gestation, uterine and placental space-occupying anomalies (e.g.,
fibroids), and polyhydramnios. Insufficient endometrial surface area for
placental invasion and growth, plus abnormal placental perfusion, also
combine to restrict nutrient delivery to the fetus, leading to IUGR. Poor
placental growth and function limit placental supply of growth promoting
hormones to the fetus, e.g., human placental lactogen (hPL), steroid
hormones, and insulin-like growth factor-I (IGF-I) (18,19,20), and limit
effective maternal-fetal nutrient exchange. In cases of polyhydramnios,
IUGR often is related to the primary pathologic processes such as fetal
infection, anemia, cardiac failure, and neuromuscular disorders.
Intrauterine fetal infections can limit fetal growth by damaging the fetal
brain and the neuroendocrine axis that support fetal growth via insulinlike growth factors (IGFs) and insulin. Intrauterine infections also can
damage the fetal heart, leading to diminished cardiac output, poor
placental perfusion, and inadequate nutrient substrate uptake. Fetal
infections and ascending infections of the membranes from the vagina
also are associated with preterm delivery. They probably do this by
enhancing the fetal supply of prostaglandins, which causes fetal and
uterine production of various cytokines that are associated with or cause
the onset of labor (21). Chronic placental and fetal infections also limit
placental perfusion, in some cases by inhibition of nitric oxide
production, which leads to uteroplacental vasoconstriction, placental
insufficiency, and IUGR (22). Preeclamptic women have poor endometrial

vascular support for growth of the placenta, leading to placental growth
failure, fetal nutrient deficit, and IUGR (23). Fetal hypoglycemia,
hypoxemia, and acidosis usually are present in such cases of poor
placental development and perfusion. These factors lead to increased
production of prostaglandins and the activation of labor-promoting
cytokines, leading also to preterm delivery (24). Many of these cases are
delivered preterm to protect the mother from eclampsia or the fetus from
hypoxic-ischemic injury. Very young and very old women both produce
IUGR infants who often are born prematurely. Nutritional, uterine, and
vascular mechanisms may be common in these situations. Young stillgrowing adolescent girls appear less capable of mobilizing fat reserves in
late pregnancy, apparently reserving them instead for their own
continued development (25). Failure to mobilize such reserves can limit
nutrient supply to the fetus and the rate of fetal growth (26). IUGR in
cases of maternal smoking and substance abuse may be as a result of
reduced placental blood flow, inhibition of uteroplacental vascular
development, or direct fetal toxicity.
These examples illustrate that IUGR is commonly associated with
conditions that also are related to preterm delivery. It is not surprising,
therefore, and important to note, that IUGR is an increasingly common
finding among infants born at earlier gestational ages. Whereas at term
10% of infants are classified as SGA, at less than 28 to 30 weeks
gestational age, 30% to 40% of infants may be the result of IUGR (27).
FETAL GROWTH
The period of fetal growth is from the end of embryogenesis, at about
the end of the first third of gestation, until term. During the embryonic
period, growth occurs primarily by increased cell number (hyperplasia)
(28). In the middle third of gestation, cell size also increases
(hypertrophy), although the rate of cell division becomes stable. In the
last third of gestation, the rate of cell division declines, although cell
size continues to increase. Thus, insults that limit fetal growth in the
embryonic period result in global reduction in fetal growth, whereas
insults in the third part of gestation usually limit growth of fetal adipose
tissue and skeletal muscle with less effect on the growth of other organs,
especially the brain and heart (29).
Figure 26-6 Serial fetal body measurements in a mother with severe

ulcerative colitis. Note that fetal growth () begins to decrease
markedly in mid-gestation but returns to normal following the initiation
of central hyperalimentation at approximately 28-30 weeks. Solid lines
represent the mean 2 S.D. of a reference population of human fetuses
with normal fetal growth rates according to ultrasound measurements.
Reproduced with permission from Creasy RK, Resnik R. Intrauterine
growth retardation. In: Creasy RK, Resnik R, (Editors), Maternal-Fetal
Medicine, Second Edition. Philadelphia: W.B. Saunders Co., 1989, p. 558.
TABLE 26-1 MATERNAL CONDITIONS ASSOCIATED WITH

INTRAUTERINE GROWTH RESTRICTION AND PRETERM DELIVERY
Both very young and advanced maternal age

Maternal prepregnancy short stature and thinness
Poor maternal weight gain during the latter third of pregnancy
Maternal illness during pregnancy
Nulliparity
Failure to obtain normal medical care during pregnancy
Lower socioeconomic status
Black race (in the United States)
Multiple gestation
Uterine and placental anomalies
Polyhydramnios
Preeclampsia
Diabetes
Intrauterine infections
Cigarette smoking, cocaine use, and other substance abuse
GROWTH OF BODY COMPONENTS IN THE

FETUS
Water
Fetal body water content, expressed as a fraction of body weight,
decreases over gestation as a result of relative increases in protein and
mineral accretion (Fig. 26-7) (30) and in humans because of the
development of relatively large amounts of adipose tissue in the third
trimester. Fetuses with marked IUGR and SGA neonates who have
decreased body fat content have even lower fractional contents of body
water. Extracellular water also decreases more than intracellular water
as gestation advances, primarily because of increasing cell number and
increasing cell size rather than the intracellular concentration of osmotic
P.497
substances. Measurements of extracellular space in SGA infants usually
are normal for gestational age, as adipose tissue, skeletal muscle, and
mineral accretion all are decreased to about the same extent (31).
Figure 26-7 Composition of weight gain in normal human fetuses. From

Ziegler EE, O'Donnell AM, Nelson SE, et al. Body composition of the
reference fetus. Growth 1976;40:329.
Minerals
Fetal calcium content in SGA and AGA fetuses increases exponentially
with a linear increase in length, because bone density, area, and
circumference increase exponentially in relation to linear growth (12).
Accretion of other minerals varies more directly with body weight and
according to the distribution of the minerals into extracellular (e.g.,
sodium) or intracellular (e.g., potassium) spaces (Table 26-2).
Nitrogen and Protein

There are very few chemical composition studies of normal human
infants. Based on data from 15 studies accounting for 207 infants,
Sparks has shown that nonfat dry weight and
P.498
nitrogen content (predictors of protein content) have a linear
relationship with fetal weight and an exponential relationship with
gestational age (Fig. 26-8) (12,32). Table 26-3 shows nitrogen, protein,
and selected amino acid composition and accretion rates for normal
human fetuses. About 80% of fetal nitrogen content is found in protein;
the rest is found in urea, ammonia, and free amino acids. The data for
fetal protein content and accretion in Table 26-3 thus may be high, as
they are based solely on nitrogen content.
TABLE 26-2 CHEMICAL COMPOSITION OF THE BODY OF THE DEVELOPING FETUS a
Body
Weigh
Approxi
Per Kilogram
mate
Whole Body
Per Kilogram Fat-free Body
Fetal
Age(wk
Water
Fat(
Water
(g)
g)
(g)
N(g) Ca(g)
30
13
900
906
10
100
15
890
894
200
17
885
889
500
23
880
885
14
4.4
1,000
26
860
10
869
14
1,500
31
847
23
17
t(g)
867
Mg
Na
K(m
Cl(m
Fe(
Cu(
Zn(
P(g)
(g)
(mEq)
Eq)
Eq)
mg)
mg)
mg)
3.0
2.0
0.10
20
40
81
10
3.0
2.0
0.10
100
40
70
50
14
4.0
3.0
0.15
100
40
70
50
3.5
18
3.0
0.20
100
44
66
56
3.5
18
6.1
3.4
0.22
90
44
66
65
3.5
18
6.8
3.8
0.24
85
44
66
68
3.8
18
2,000
33
810
50
853
7.
0.
24
85
44
63
84
4.2
18
85
48
56
95
4.3
18
90
49
55
95
4.5
18
95
51
54
95
4.8
18
2,500
35
776
74
838
9.
0.
25
3,000
38
727
120
826
9.
0.
27
3,500
40
686
160
816
0.
0.
27
From Widdowson EM. Changes in body proportions and composition during growth. In: Davis JA, Dobbing J, eds.
Scientific foundations of paediatrics. Philadelphia: WB Saunders, 1974:155, with permission.
Figure 26-8 Nonfat dry weight (A) and nitrogen content (B) are plotted
against gestational age for LGA (,), AGA (O,), and SGA ([black
down-pointing triangle],) infants. Reproduced with permission
from Sparks JW. Intrauterine growth and nutrition. In: Polin RA, Fox
WW, (eds.), Fetal and neonatal physiology. Philadelphia: W. B. Saunders
Co., 1992:184, with permission.
Nitrogen and Protein Accretion in Small-forGestational-Age Infants

Among SGA infants, nitrogen and protein contents are reduced for body
weight, primarily as a result of deficient production of muscle mass. In
fact, they often are reduced below that of fat as a fraction of body
weight (33).
Glycogen
Many tissues in the fetus, including brain, liver, lung, heart, and skeletal
muscle, produce glycogen over the second half of gestation (34). Liver
glycogen content, which increases with gestation (Fig. 26-9), is the most
important store of carbohydrate for systemic glucose needs, because
only the liver contains sufficient glucose-6-phosphatase for release of
glucose into the circulation. Skeletal muscle glycogen content increases
during late gestation and forms a ready source of glucose for glycolysis
within the myocytes. Lung glycogen content decreases in late gestation
with change in
P.499
cell type, leading to loss of glycogen-containing alveolar epithelium,
development of type II pneumocytes, and onset of surfactant production.
Cardiac glycogen concentration decreases with gestation, owing to
cellular hypertrophy, but cardiac glycogen appears essential for postnatal
cardiac energy metabolism and contractile function. Glycogen synthesis
rates are low in human fetuses, about 2 mg/d/g of liver, accounting for
less than 5% of fetal glucose utilization (35). Net synthesis, degradation,
and accumulation rates of fetal glycogen are controlled by the functional
states of two enzymes, glycogen synthase, which promotes glycogen
formation, and glycogen phosphorylase, which promotes glycogen
degradation (34,36,37). The total liver content of these two enzymes is
relatively constant over gestation. Their functional states are regulated
by hormone and substrate concentrations. For example, insulin acts
synergistically with glucose to build hepatic glycogen stores, whereas
close to term, cortisol, epinephrine, and glucagon develop the capacity to
promote glycogenolysis and glucose release into the plasma.
TABLE 26-3 INCREMENTS PER DAY OF NUTRIENTS IN THE FETAL

BODY AT SELECTED INTERVALS DURING GESTATION a
Fetal Age
Range
12-
16-
20-
24-
28-
32-
36-
(wk)
16
20
24
28
32
36
40
Weight
0.02
0.1-
0.3-
0.75-
1.35-
2.0-
2.7-
Range(kg) -0.1
0.3
0.75
1.35
2.0
2.7
3.4
Increments of nitrogen (N) and protein in fetal body per 24h
Total N
Protein
29
93
243
326
386
504
714
0.18
0.5
1.5
2.0
2.4
3.1
4.4
(gm)
(N=6.25)
Increments of individual amino acids in fetal body (mg/d)
ILE
26
53
71
82
109
148
LEU
13
43
111
151
174
231
330
LYS
13
41
107
145
167
222
313
MET
11
28
39
44
59
92
PHE
23
61
83
95
127
184
TYR
17
44
59
68
91
127
THR
23
61
83
95
127
184
VAL
27
70
94
109
145
210
ARG
14
43
114
154
177
236
340
HIS
15
39
53
61
81
112
ALA
13
41
107
145
167
222
319
ASP
17
52
136
183
211
281
392
GLU
23
74
195
263
303
403
568
GLY
21
68
177
240
276
367
513
PRO
15
48
125
168
194
258
300
SER
25
66
89
102
136
191
From Widdowson EM. Chemical composition and nutritional needs
of the fetus at different stages of gestation.

In: Aebi H, Whitehead R, eds. Maternal nutrition during pregnancy
and lactation. Berne: H. Huber, 1980:3948,with permission.
Glycogen Deficiency in Small-for-GestationalAge Infants

Glycogen content is markedly reduced in SGA infants, both in the liver
and in the skeletal muscles (34). This is as a result of lower fetal plasma
concentrations of glucose and insulin, which are the principal regulators
of glycogen synthesis. If the SGA fetus also experiences repeated
episodes of hypoxemia, epinephrine secretion in response will further
deplete glycogen by activating glycogen phosphorylase and increasing
glycogenolysis.
Adipose Tissue
At term, fetal fat content, expressed as a fraction of fetal weight, varies
markedly among species (Fig. 26-10) (37). The fat content of newborns
of almost all land mammals at term is 1% to 3%, which is considerably
less than the 15% to 20% fat content of human term infants. Even in
those species, such as the human, that take up fat from the placenta and
deposit fat in fetal tissues, the rate of fetal fatty acid oxidation is
presumed low. This condition occurs because plasma concentrations of
fatty acids (and keto acid products, such as b-hydroxybutyrate and
acetoacetate) are low, and because the carnitine palmityl transferase
enzyme system is not sufficiently developed to deliver long-chain fatty
acids to the respiration pathway inside the mitochondria. Fat accretion
for the human fetus is shown in Fig. 26-11. Between 26 and 30 weeks of
gestation, nonfat and fat components contribute equally to the carbon
content of the fetal body (12,38). After that period, fat accumulation
exceeds that of the nonfat components. By term, the deposition of fat
accounts for more than 90% of the carbon accumulated by the fetus. The
rate of fat accretion is approximately linear between 36 and 40 weeks of
gestation, and, by the end of gestation, fat accretion ranges between 1.6
and 3.4 g/d/kg. At 28 weeks of gestation, it is slightly less and ranges
between 1.0 and 1.8 g/d/kg.
Figure 26-9 Liver glycogen content (as carbohydrate) in normal human

fetuses and newborn infants with normal birth weights for gestational
age (O O, mean SEM) and infants of low birth weight for
gestational age (X). From Shelley HJ, Neligan GA. Neonatal
hypoglycaemia. Br Med Bull 1966;22:34, with permission.
P.500
Adipose Tissue Deficiency in Small-forGestational-Age Infants

By term, fat content in human fetuses with IUGR may be less than 10%
of body weight (33). In these cases, the smaller placenta limits fetal
fatty acid and triglyceride supply. Similarly, the smaller placenta
decreases fetal glucose supply, which reduces glycerol production and
triglyceride synthesis. Insulin deficiency, a result of decreased glucose
and amino acid supply to the fetus, also limits fat synthesis. The insulin
activation of peripheral lipoprotein lipase is thereby reduced, which
normally is necessary to release fatty acids from circulating lipoproteins
for adipocyte uptake and triglyceride synthesis, and to decrease the
normal insulin stimulation of fatty acid synthase within adipocytes.
Figure 26-10 Fetal fat content at term as a percent of fetal body weight among
species. Reproduced with permission from Hay WW Jr. Nutrition and development of
the fetus: carbohydrate and lipid metabolism. In Walker WA, Watkins JB (Editors),
Nutrition in Pediatrics, Second Edition. Hamilton: BC Decker, 1966, p. 376, with
permission.
Calories (Total Energy Storage)

The energy value of various tissue components is shown in Table 26-4.
Fat has a high energy content, 9.5 kcal/g, and a very high carbon
content, approximately 78%. Thus, differences in fetal fat concentration
lead to large differences in calculated caloric accretion rates and carbon
requirements of the fetal tissues for growth. The caloric concentration of
nonfat dry weight is fairly consistent at different developmental stages,
indicating that the ratio of protein to nonprotein substrates in the tissues
is relatively constant (Table 26-5). Thus, caloric accretion rate of any
fetus can be estimated from the growth curve of the fetus and the
changing fat and water concentrations (38).
Caloric Accretion Deficiency in Small-forGestational-Age Infants

Growth of fat and nonfat (protein plus other) tissues is metabolically
linked through energy supply that is used for protein synthesis and the
production of anabolic hormones (39). These promote positive protein,
fat, and carbohydrate growth. Thus, restriction of nutrient supply is

likely to produce growth deficits of all tissues, not just fat. Indeed,
growth restriction involves limitation of muscle growth and fat and
glycogen (33). For example, chronic selective caloric (glucose)
restriction in the experimental fetal sheep model leads to increased
protein breakdown and lower rates of fetal growth and lipid content (40).
Some growth curves, such as those shown in Figs. 26-8 and 26-11 from
human
P.501
infants born prematurely at different times over the last third of
gestation, show a bias toward thinner SGA infants with less fat relative
to nonfat weight and nitrogen content (32). These infants, however, were
stillborn and may have suffered extensive wasting and IUGR. Other
studies have shown that human SGA infants can be markedly deficient in
muscle mass, even more than for fat (33,41).
Figure 26-11 Dry weight (A) and fat content (B) plotted against
gestational age in the same newborn human infants shown in Fig. 26-8
for LGA (,), AGA (O, ), and SGA ([black down-pointing triangle],
) infants. Reproduced with permission from Sparks JW.
Intrauterine growth and nutrition. In: Polin RA, Fox WW, (eds.), Fetal
and neonatal physiology. Philadelphia: W. B. Saunders Co., 1992:184,
with permission.
REGULATION OF FETAL GROWTH

Fetal growth is regulated by maternal, placental, and fetal factors,
representing a mix of genetic mechanisms and environmental influences
through which genetic growth potential is expressed and modulated.
Epidemiologic Considerations
The major maternal risk factors for IUGR that vary among populations
and among individuals within populations include small maternal size
(height and prepregnancy weight) and low maternal weight gain during
pregnancy. Low maternal body mass index (the degree of thinness or
fatness, defined as [weight (kg)]/[height (cm)]2) is a major predicator
of IUGR. This characteristic interacts with other risk factors, such as
diet, smoking, illnesses, and so forth, to affect fetal growth, especially in
thin women. For example, smoking has only half the impact on fetal
growth in obese vs. thin women, and in black vs. white women (42). Low
blood pressure has a detrimental impact on fetal growth, mostly in thin
women (43). Moderate obesity,
P.502
therefore, protects against most growth-inhibiting risk factors except for
black race and female gender. This pattern also holds for certain
therapies. For example, zinc supplementation has a major impact on
fetal growth in black women who have relatively low plasma zinc levels
early in pregnancy, with all of the impact occurring in relatively thin
women (44). Also, low-dose maternal aspirin treatment has been shown
to improve fetal growth primarily in thin women (45).
TABLE 26-4 ENERGY VALUE OF HUMAN TISSUE COMPONENTSb
Tissue Component
Energy Value (kcal/g) a
H2O
Fat
9.45
Nonfat dry weight
Pig
4.0-4.6
Lamb
4.4-4.6
Guinea pig
4.6
Carbohydrate
4.15 (3.7-4.2)
Protein
5.65
In vivo catabolism
4.35
1 kcal = 4,190 J.
From Battaglia FC, Meschia G. An introduction to fetal physiology.
Orlando: Academic Press, 1986, with permission.
Genetic Factors
Many genes contribute to fetal growth. Table 26-6 lists estimates of the
quantitative contribution of fetal and parental factors to fetal growth and
birth weight at term. Maternal genotype is more important than fetal
genotype in the overall regulation of fetal growth. However, the paternal
genotype is essential for trophoblast development, which secondarily
regulates fetal growth by the provision of nutrients. More specific gene
targeting studies have shown the importance of genomic imprinting on
fetal growth. For example, normal fetal and placental growth in mice
require that the IGF-II gene be paternal and the IGF-II receptor gene be
maternal, whereas maternal disomy producing IGF-II under- expression
results in fetal dwarfism (39).
TABLE 26-5 CALCULATION OF THE CALORIC DISTRIBUTION IN THE
TERM HUMAN INFANT a
Wet
Weight
Fat
Nonfat
Nonfat
Wet
Dry
Weight
Weight
Weight (g)
3,450
386
3,064
511
Total calories
5,950
3,650
2,300
2,300
1.73
9.45
0.75
4.5
(kcal)
Caloric
concentration
(kcal/g)
Data from Ziegler EE, O'Donnell AM, Nelson SE, et al. Body composition of
the reference fetus. Growth 1976;40:329, with permission.
TABLE 26-6 FACTORS DETERMINING VARIANCE IN BIRTH WEIGH a , b
Percent of Total Variance

Fetal
Genotype
Sex
Total
18
Maternal
Genotype
20
Maternal environment
24
Maternal age
Parity
Total
52
Unknown
30
From Penrose LS. Proceedings of the Ninth International Congress of
Genetics, Part 1, 520, 1954, with permission.
From Milner RDG, Gluckman PD. Regulation of intrauterine growth. In:
Gluckman PD, Heymann MA, eds. Pediatrics & perinatology: the scientific
basis, 2nd ed. London: Arnold, 1993:284, with permission.
Nongenetic Maternal Factors

Under usual conditions, fetal growth follows its genetic potential, unless
the mother is unusually small and limits fetal growth by a variety of
factors considered collectively as maternal constraint. Maternal
constraint represents a relatively limited uterine size, including placental
implantation surface area and uterine circulation, and thus the capacity
to support placental growth and nutrient supply to the fetus. A clear
example of maternal constraint is the reduced rate of fetal growth of
multiple fetuses in a specieshumanthat optimally supports only one
fetus (Fig. 26-12). Obviously, small fetuses of small parents do not
reflect fetal growth restriction; in fact, their rates of growth are normal
for their genome and for the size of the mother. Unless maternal
constraint is particularly prominent, such fetuses would not grow faster
or to a larger size if more nutrients were provided, although they might
grow somewhat larger if the maternal uterine endometrial surface area,
and thus placental implantation and growth area, were increased. The
nongenetic, maternal nature of this effect has been demonstrated by
embryo transfer and cross-breeding experiments (29). For example, a
small-breed embryo transplanted into a large-breed uterus will grow
larger than a small-breed embryo remaining in a small-breed uterus.
Furthermore, partial reduction in fetal number in a polytocous species,
such as the rat, produces greater than normal birth weights in the
remaining offspring. Conversely, embryo transfer of a large-breed
embryo into a small-breed uterus will result in a newborn that is smaller
than in its natural large-breed environment. Such
P.503
evidence supports the concept that fetal growth is normally constrained,
and this constraint comes from the maternal environment, i.e., the size
of the uterus.
Figure 26-12 Mean birth weight of single and multiple human fetuses related to
duration of gestation. Adapted from McKeown T, Record RG. Observation on foetal
growth in multiple pregnancy in man. J Endocrinol 1952;8:386, with permission.
Maternal Nutrition
The single most important environmental influence that affects fetal
growth is the nutrition of the fetus. Normal variations in maternal
nutrition, however, have relatively little impact on fetal growth and the
severity of IUGR. This is because changes in maternal nutrition, unless
extreme and prolonged, do not markedly alter maternal plasma
concentrations of nutrient substrates or the rate of uterine blood flow,
the principal determinants of nutrient substrate delivery and transport to
the fetus by the placenta (46). Human epidemiologic data from
conditions of prolonged starvation, and nutritional deprivation in
experimental animals, indicate that severe limitations in maternal
nutrition limit fetal growth only by 10% to 20%. Epidemiologic data from
the Dutch during the Hunger Winter of 1944 showed an average
reduction in fetal weight at term of 300 g (47), whereas birth weight at
term was reduced by 500 g in women who suffered a more severe and
prolonged famine in wartime Leningrad (48). Interestingly, secondgeneration daughters of women who suffered extreme nutritional deficit
during gestation in turn tend to produce SGA infants who are 200 to 300
g less than normal at term with their first pregnancies (49). However,
attempts to limit weight gain in pregnancy with a 1,200-kcal diet (50%
of what is now recommended to prevent preeclampsia) increased the
incidence of fetal growth restriction up to tenfold (50). Restrictions of
calorie and protein intakes to less than 50% of normal for a considerable
portion of gestation are needed before marked reductions in fetal growth
are observed. Such severe conditions often result in fetal loss before the
impact of fetal growth rate in late gestation and fetal size at birth are
manifested.
Attempts to increase fetal weight gain with maternal nutritional
supplements have produced mixed results. Higher caloric feeding usually
increases fetal adiposity, not growth of muscle mass or gain in length or
head circumference. In contrast, high protein supplements tend to
produce delayed fetal growth (51). Mechanisms responsible for this
phenomenon are not known.
Maternal Chronic Diseases

Chronic hypertension, pregnancy-induced hypertension, and
preeclampsia, like other vascular disorders including severe and longstanding diabetes mellitus and serious autoimmune disease associated
with the lupus anticoagulant, have a common effect of limiting
trophoblast invasion, placental growth and development, uteroplacental
blood flow, and fetal oxygen and nutrient deficiency (52). Maternal
cyanotic congenital heart disease can limit fetal oxygen supply, which
can limit fetal growth (53). High-altitude hypoxia also can limit fetal
growth (54), but usually this is only clinically significant for
nonindigenous women who move to altitudes above 10,000 feet. Severe
sickle cell crises can damage uterine vasculature, leading to placental
growth and transport capacities (55).
Maternal Drugs
Specific effects of drugs on fetal growth (Table 26-7) are often difficult
to sort out clinically, as many women who abuse drugs do so with many
drugs taken intermittently, at different doses, and at different periods of
fetal vulnerability. These women also frequently suffer from other

disorders
P.504
that could lead to poor fetal growth, such as poor nutrition, recurrent
acute illnesses, and chronic diseases (56). Fetal growth restriction does
appear to be a major part of the fetal alcohol syndrome. It is not clear
when during gestation the specific effects of alcohol on fetal growth rate
occur. Alcohol may exert its nonteratogenic effects by limiting placentalto-fetal amino acid transport (57). Cocaine probably exerts its primary
effects on producing fetal growth restriction by causing uterine and
perhaps umbilical vasoconstriction and reduced placental perfusion (58).
The most consistent drug reducing fetal growth is cigarette smoking
(41). Deficits of at least 300 g (about 10% of normal term weight) are
not uncommon. A likely common mechanisms is the effect of nicotine,
and of catecholamines released in response, to constrict the uterine and
perhaps the umbilical vasculature, reducing placental perfusion. Carbon
monoxide, cyanide, and other cellular toxins may limit oxygen transport
to fetal tissues and cellular respiration.
TABLE 26-7 DRUGS ASSOCIATED WITH INTRAUTERINE GROWTH RESTRICTION
Amphetamines
Antimetabolites (e.g., aminopterin, busulfan, methotrexate)
Bromides
Cocaine
Ethanol
Heroin and other narcotics, such as morphine and methadone
Hydantoin
Isotretinoin
Metals such as mercury and lead
Phencyclidine
Polychlorinated biphenyls (PCBs)
Propranalol
Steroids
Tobacco (carbon monoxide, nicotine, thiocyante)
Toluene
Trimethadione
Warfarin
Placenta
The size of the placenta and its directly related nutrient transport
functions are the principal regulators of nutrient supply to the fetus and
thus the rate of fetal growth (36). Nearly all cases of IUGR are
associated with a smaller-than-normal placenta. Figure 26-13 shows a
direct relationship between fetal weight and placental weight in humans,
demonstrating that LGA, AGA, and SGA infants are directly associated
with LGA, AGA, and SGA placentas (59). Placental growth normally
precedes fetal growth, and failure of placental growth is directly
associated with decreased fetal growth. Variable limitations in placental
nutrient transfer capacity modulate this primary effect of placental size
on fetal growth. In some cases of experimentally reduced placental size,
for example, fetal weight is not reduced proportionately (60). This
indicates that either the capacity of the smaller placenta to transport
nutrients to the fetus increases adaptively or the fetus develops
increased capacity to grow. More characteristically, though, fetal growth
fails first, or in direct relation to decreased nutrient supply. With primary
fetal growth failure, placental growth can increase disproportionately,
resulting in a larger than normal placental-to-fetal weight ratio for
gestational age. This is characteristically seen under chronic hypoxic
conditions of high altitude exposure or maternal anemia and has been
seen in certain experimental situations of maternal undernutrition in
early gestation (61). A variety of placental pathologic conditions are
associated with IUGR (Table 26-8). In most of these cases, the placenta
is simply smaller than normal. In many, there also is abnormal
trophoblast development, including abnormal vascular growth in the
trophoblast villi, frequently associated with limited uterine vascular

perfusion of the intervillous spaces.
Figure 26-13 Mean placental weights for large-for-gestational-age (O),

average-for-gestational-age (), and small-for-gestational-age ()
human infants at each gestational age. SEM given for AGA infants
alone. From Molteni RA, Stys SJ, Battaglia FC. Relationship of fetal and
placental weight in human beings: fetal/placental weight ratios at
various gestational ages and birth weight distributions. J Reprod Med
Placental and fetal growth both depend on an adequate supply of
maternal blood to the placenta. IUGR is associated with inadequate
development of the uteroplacental circulation, and radioisotope studies
have demonstrated more than a twofold blood flow reduction in
P.505
comparison with normal pregnancies (62). IUGR in the second half of
gestation is due primarily to a failure of the normal villous vascular tree,
mainly in the phase of nonbranching angiogenesis, because terminal villi
are critical for oxygen and nutrient transport to the fetus (63). This
angiogenesis in turn depends on cytotrophoblast invasion of the uterus
and its arterioles. Cytotrophoblast invasion is actually a differentiation
process whereby the cells lose the ability to proliferate and modulate
their expression of state-specific antigens. These antigens include
members of the integrin family of cell-extracellular matrix receptors that
are required for migration and invasion of the endometrium and decidua
of the uterus (64). Preeclampsia, which is associated with IUGR, is
characterized by shallow cytotrophoblast invasion (65). Abnormal
cytotrophoblast differentiation also occurs, evidenced by the cells'
inability to switch on their integrin repertoire (66). The same
observations have been made on cultured normal cytotrophoblast cells in
a hypoxic environment (67). These in vitro results indicate that whatever
leads to hypoxia of the invading cytotrophoblast cells increases
cytotrophoblast proliferation over differentiation and invasion, thus
setting the stage for deficient placental development that can result in
deficient nutrient and growth factor supply to the fetus, producing fetal
growth restriction.
TABLE 26-8 PLACENTAL GROWTH DISORDERS THAT LEAD TO OR

ARE ASSOCIATED WITH INTRAUTERINE GROWTH RESTRICTION
Abnormal umbilical vascular insertions (circumvallate,

velamentous)
Abruption (chronic, partial)
Avascular villi
Decidual arteritis
Fibrinosis, atheromatous changes, cytotrophoblast hyperplasia,
basement membrane thickening
Infectious villitis (as with TORCH infections)
Ischemic villous necrosis and umbilical vascular thromboses
Multiple gestation (limited endometrial surface area, vascular
anastomoses)
Multiple infarcts
Partial molar pregnancy
Placenta previa
Single umbilical artery
Spiral artery vasculitis, failed or limited erosion into intervillous
space
Syncytial knots
Tumors, including chorioangioma and hemangiomas
At more advanced stages of placental development, placental production

of growth factors and growth regulating hormones develops, leading to
significant autocrine regulation of placental growth and placental
regulation of fetal growth processes. Human placental lactogen is
synthesized and secreted by the syncytiotrophoblast cells of the placenta
(68). Fetal growth-promoting actions of placental lactogen are mediated
by stimulation of IGF production in the fetus and by increasing the
availability of nutrients to fetal tissues (69). Obviously, placental growth
failure and/or nutrient deficit to the placenta can result in decreased
placental production of growth factors that then would lead to fetal
growth failure.
FETAL NUTRIENT UPTAKE AND METABOLISM

AND REGULATION OF FETAL GROWTH
In general, decreased rates of fetal growth represent an adaptation to
inadequate nutrient supply. IUGR that results from decreased nutrient
supply can be interpreted, therefore, as a successful, if not perfect,
adaptation to maintain fetal survival. Fetal undernutrition also appears
to affect rapidly growing fetuses more than slowly growing fetuses who
may have been programmed to grow more slowly for genetic or
embryonic and early fetal pathologic reasons (70). Reintroduction of
nutrients can return fetal growth to normal in those fetuses whose rapid
growth rate was decreased by undernutrition, but too rapid an
introduction of nutrients has often caused fetal pathology, including

hyperlactatemia (70) and occasionally even acidosis and hypoxemia.
Glucose Uptake, Metabolism, and Regulation

of Fetal Growth
Nearly all IUGR fetuses, whether studied experimentally in animal models
or in women by cordocentesis (direct umbilical blood sampling), have
relatively lower plasma glucose concentrations compared with normally
grown fetuses (71,72). Fetal hypoglycemia has several consequences
important to fetal adaptation and survival when maternal glucose supply
is limited. First, relative fetal hypoglycemia is an important and natural
compensatory mechanism that helps to maintain the maternal-to-fetal
glucose concentration gradient and thus the transport of glucose across
the placenta to the fetus (73). Despite this compensation, fetal
hypoglycemia limits tissue glucose uptake directly by diminished mass
action and indirectly by limiting fetal insulin secretion and thus the effect
of insulin to promote tissue glucose uptake by skeletal muscle, heart,
adipose tissue, and liver. Insulin also normally suppresses hepatic
glucose production and release, and it acts as an anabolic hormone that
increases net protein balance by inhibiting protein breakdown. Thus, a
decrease in fetal plasma insulin concentration initially may allow fetal
glucose production to take place, thereby providing glucose for both fetal
and placental needs, but subsequently, combined with hypoglycemia,
results in increased protein breakdown and decreased protein accretion
(39,74,75).
Circulating concentrations and tissue-specific expression of growth
factors such as IGF-I and IGF-II (see Fetal Amino Acid Metabolism) also
are decreased during fetal hypoglycemia (76), which may contribute to
increased fetal protein breakdown and decreased rates of fetal growth.
Thus, fetal hypoglycemia in response to a decrease in maternal glucose
supply acts to maintain fetal glucose supply, but it also leads to lower
anabolic hormone concentrations, which limit the rate of fetal growth,
thereby decreasing fetal nutrient needs.
Fetal Amino Acid Metabolism

The placenta contains a large variety of amino acid transporters, most of
which use energy to actively concentrate amino acids in the trophoblast,
which then diffuse into the fetal plasma, producing higher concentrations
than in the maternal plasma (77). With small placentas, fetal amino acid
supply is reduced, as are fetal amino acid concentrations, fetal protein
synthesis, fetal protein and nitrogen balance, and, ultimately, fetal
growth rate. Reduced energy supply to the placenta also reduces amino
acid transport to the fetus. This is especially the case for oxygen deficit,
either from primary hypoxemia or from reduced uteroplacental blood
flow, and glucose deficit from chronic maternal and fetal hypoglycemia
(78,79). Of course, hypoxemia and hypoglycemia could reduce fetal
growth independently of reduced amino
P.506
acid transport, for example, by limiting anabolic hormone and growth
factor production or by decreasing energy supply, both of which are
necessary to produce protein synthesis and to limit protein breakdown in
fetal tissues.
The reason why amino acid and energy supplies are so important for
fetal protein and nitrogen balance and for fetal growth is illustrated in
Fig. 26-14. This figure shows results of experiments in fetal sheep over
the second half of gestation, comparing fractional protein synthesis rates
derived from tracer amino acid data and fractional body growth rates
derived from carcass analysis data. The fractional rate of protein
turnover per unit wet weight of fetus is several fold higher at 50% to
60% of term gestation (equivalent to about 20 to 24 weeks of human
gestation). Such high rates of protein turnover require a much greater
rate of amino acid supply and energy than at term, when fetal protein
turnover rate is much lower. Indeed, in midgestation fetal sheep, glucose
utilization rates per whole fetal weight and oxygen consumption rates
per dry fetal weight are much higher in the early fetus than at term (80).
These conditions result in a 50% higher rate of net protein accretion and
fractional rate of fetal growth at midgestation than at term. Clearly,
amino acid and energy deficits will affect the growth rate of the fetus at
earlier stages of gestation, when fetal growth normally is very rapid
much more than at term, when fetal growth rate is slower.
In the normally growing fetus, net protein synthesis exceeds net protein
breakdown, resulting in net protein accretion. The mechanisms
underlying the reduction in protein synthesis rate over gestation appear
to be intrinsic to the fetus, and not to limitation of nutrient supply by the
placenta. These mechanisms include changing proportions of the organs
as fractions of body mass (Table 26-9).
Figure 26-14 Fractional rate of protein synthesis (K s ) over gestation in

fetal sheep studied with leucine () and lysine (O) radioactive tracers
compared with the fractional rate of growth (K G ) in the lower portion of
the figure (). From ref. 77. Adapted from refs. 46 and 80; and from
Meier PR, Peterson RG, Bonds DR, Meschia G, Battaglia FC. Rates of
protein synthesis and turnover in fetal life. Am J Physiol 1981;240:E320.
TABLE 26-9 PERCENT CONTRIBUTION OF ORGANS AND TISSUES TO

THE BODY WEIGHT OF HUMAN FETUSES AND NEWBORN INFANTS a
Fetus (20-24 wk)
Full-term Infant
Skeletal muscle
25
25
Skin
13
15
Skeleton
22
18
Heart
0.6
0.5
Liver
Kidneys
0.7
Brain
13
13
From Widdowson EM. Growth and composition of the fetus and
newborn. In: Assali NS, ed. Gestation. New York: Academic Press,
1968:27, with permission.
Fetal Endocrine and Autocrine/ParacrineActing Growth Factor Effects on Fetal Growth

Table 26-10 lists those hormones whose deficiency results in a reduction
in fetal growth supported by experimental evidence (81). These fetal
hormones promote growth (and development) in utero by altering both
the metabolism and gene expression of fetal tissues. These hormonal

actions ensure that fetal growth rate is commensurate with nutrient
supply.
Insulin
Insulin has direct mitogenic effects on cellular development and thus can
regulate cell number. It also enhances glucose consumption by many
cells, particularly in muscle, and limits protein breakdown (75). The
latter effects are associated with reduced fetal growth when insulin
concentration is low. This has been produced directly by experimental
surgical (82) and chemical (83) ablation of the pancreas and/or the
function of the pancreatic beta cells to secrete insulin, and has been
observed clinically in infants who suffer pancreatic agenesis (84). Figure
26-15 shows reduced rates of growth in fetal sheep that underwent
surgical pancreatectomy and a return to normal rates of growth with
insulin replacement. Much of the growth reduction with hypoinsulinemia
from pancreatectomy is caused by a release of insulin's normal inhibitory
role on glucose production, resulting in fetal hyperglycemia, a secondary
decrease in the maternal-fetal glucose concentration gradient, and thus a
decrease in glucose transport to the fetus. Without this glucose, fetal
growth decreases, as has been shown by a direct decrease in glucose
uptake by the fetus following the production of chronic maternal
hypoglycemia (85). Fetal amino acid uptake decreases under the same
circumstances. Thus, insulin deficiency, directly and indirectly, results in
a decrease in fetal nutrient supply. Initially, fetal protein breakdown
results in fetal
P.507
amino acid release for energy (via direct amino acid oxidation in the
tricarboxylic [citric] acid cycle) and glucose production. Later on, the
reduced rate of fetal growth during conditions of low insulin, glucose,
and amino acid concentrations is sustained by increased protein
breakdown (40); however, amino acids are used to maintain protein
turnover rate and not for protein accretion, oxidation, or glucose
production.
TABLE 26-10 EFFECTS OF SPECIFIC ENDOCRINE DEFICIENCIES ON BODY WEIGHT AND CROWNRUMP
LENGTH, AND INDIVIDUAL TISSUES ADVERSELY AFFECTED BY TREATMENT IN SHEEP FETUSES
DELIVERED NEAR TERM (>95% GESTATION) a
Gestational
Endocrine
Deficiency
Insulin
Procedure
Streptozotocin
Pancreatectomy
Thyroid
Thyroidectomy
Crown-
Tissues with Specific
Age at
Body
rump
Developmental
Onset (d)
Weight
Length
Abnormalities
70-85
50%
20%
None
115-120
30%
15%
None
80-96
30%
10%
Skeleton, skin, lungs,
hormones
nervous system
105-115
20%
10%
Skeleton, nervous
Adrenal
Adrenalectomy
110-120
10-15%
hormones
Pituitary
Hypophysectomy
70-79
30%
No
Liver, lungs, gut,
change
pituitary
8%
Bones, liver, lungs,
hormones
placenta
105-110
20%
10%
Bones, liver, lungs,

placenta, adrenal,
gonads
110-125
No
No
Bones, gonads,
change to
change
adrenal, liver
No
Adrenals, other
change
tissues?
5%
Pituitary stalk
section
108-112
15%
From Fowden AL. Endocrine regulation of fetal growth. In: Harding R, Genkin G, Grant A, eds. Progress
Figure 26-15 Relationship between growth and plasma insulin

concentrations in the sheep fetus. A: Mean daily crown-rump length
(CRL) increment with respect to plasma insulin concentration in
individual sheep fetuses [sham operated (O); pancreatectomized ();
pancreatectomized and given insulin treatment (^)]. B: Daily increment
in CRL with respect to time before and after abdominal surgery in
fetuses that were sham-operated (O), pancreatectomized (), and
pancreatectomized and given insulin treatment to restore normal insulin
concentrations (^). From Fowden AL. The role of insulin in prenatal
growth. J Dev Physiol 1989;12:173; Fowden AL. Endocrine regulation of
fetal growth. In: Harding R, Genkin G, Grant A, (eds.). Progress in
perinatal physiology. Reprod Fertil Dev 1995;7:50, with permission.
P.508
Insulin-Like Growth Factor-I

IGF-I is positively regulated by glucose supply in the fetus. Infusion of
IGF-I into fetal sheep decreases protein breakdown, especially when
protein breakdown is increased by fasting-induced hypoglycemia.
Metabolic effects of a decreased plasma concentration of IGF-I have not
been studied, although, as for insulin, such effects might be difficult to
separate from simultaneous changes in nutrient substrate supply and

concentration. Thus, IGF-I probably can regulate metabolic processes
that affect fetal protein balance and growth, but these have been difficult
to measure.
Plasma IGF-I concentrations are positively related to fetal size at birth. A
causative role for IGF-I has been shown by the production of transgenic
models of mice that have shown that decreased expression of IGF-I
results in markedly reduced rates of fetal growth. Other transgenic
models with increased expression of IGF-I have been associated with
increased brain growth (86,87), and this growth can be inhibited by
overexpression of IGF binding protein-1 (IGFBP-1) (88). IGF-I stimulates
an increase in oligodendrocytes and neuronal number, and neuronal
outgrowth with increased dendritic arborization and axon terminal fields
(89). Because IGF-I is decreased directly by reduced nutrient supply,
particularly glucose, and IGFBP-1 is decreased under these
circumstances, the smaller, more densely packed neuronal structure of
the undernourished brain that is seen in some infants with severe IUGR
may have been mediated by nutrient regulation of IGF-I and IGFBP-1
expression. Such developmental limitations might underlie the poorer
neurodevelopmental outcome of severely SGA infants who have relative
microcephaly.
Insulin-Like Growth Factor-II, Insulin-Like

Growth Factor Binding Protein-2, Insulin-Like
Growth Factor Binding Protein-3, and
Vasoactive Intestinal Polypeptide
Although serum concentrations of IGF-II do not correlate with fetal size
at birth in human infants, it has been shown conclusively that targeted
mutation of the IGF-II gene reduces fetal size in mice (90,91).
Furthermore, IGF-II is the predominant IGF expressed in the tissues of
embryos and fetuses of all species. As with IGF-I and IGFBP-1,
transgenic overexpression of IGF-II and IGFBP-2 shows that cellular
growth is dependent on the balance between the binding protein and the
IGF molecule itself. IGFBP-3 is the predominant IGF binding protein in
several mammalian species including humans (92,93) and is reduced in

cord blood of infants with IUGR (94). Vasoactive intestinal peptide (VIP)
is another growth factor in the fetus that affects neuronal and whole
body growth (95,96,97). Antagonists to VIP in pregnant mice produce
smaller fetuses that are particularly microcephalic (98), with central
nervous system neurons that show reduced mitosis and migration. These
effects of VIP occur in the first half of gestation, coincident with
transiently high VIP concentrations in the maternal plasma (99).
Thyroid Hormones
In all species, fetal thyroid hormone deficiency produces developmental
abnormalities in certain tissues. When maternal thyroid hormones cannot
compensate, as in the sheep, which does not transport maternal thyroid
hormones to the fetus, fetal growth restriction develops, primarily
reflecting deficient carcass growth (skin, bone, and muscle) (81). This
growth restriction results from both hypoplasia (in muscle) and
hypotrophy (in lung). More generally, fetal hypothyroidism decreases
oxygen consumption and oxidation of glucose, thereby potentially
decreasing fetal energy supply for growth. Hypothyroidism also can
decrease circulating and tissue concentrations of IGF-I.
Glucocorticoids
Glucocorticoids do not have strong effects on fetal growth rate, but they
are important in the maturation of many fetal enzymatic pathways (39).
These include glycogen deposition, gluconeogenesis, fatty acid oxidation,
induction of surfactant production and release, structural maturation of
alveoli, structural maturation of the gastrointestinal tract, increased
expression of digestive enzymes, increased adrenal function, switch from
fetal to adult hemoglobin synthesis, and others. Many IUGR fetuses have
increased cortisol concentrations that appear to result from intermittent
hypoxic stress. This may account for much of the apparent increased
maturation of IUGR fetuses, even when born preterm.
Growth Hormone
Growth hormone, which is the major hormonal regulator of postnatal
growth, has no demonstrable effect on fetal growth (39,81).
ANTENATAL CARE OF THE INTRAUTERINE

GROWTH-RESTRICTED FETUS
Diagnosis of Intrauterine Growth Restriction
Antenatal diagnosis of IUGR is difficult and often inaccurate. Despite
careful attention to gestational dating by maternal history and early and
serial fetal ultrasound evaluation, frequent maternal physical
examinations, and repeated assessment of risks for IUGR, more than
one-half of infants with IUGR may not be identified before birth (100).
Serial ultrasound evaluation of fetal growth rate and fetal body
proportions and Doppler velocimetry of the uterine, placental, and fetal
circulations are now the standard diagnostic approaches to determining
the severity of IUGR. In particular, Doppler velocimetry has provided
increasingly accurate prognostic evidence of worsening fetal condition
and impending fetal death (101,102,103,104,105,106). Chronic fetal
distress resulting from placental insufficiency, hypoxia, and ischemia
(with or without acidosis) is
P.509
associated with increased Doppler arterial waveform amplitudes that
indicate increased vascular resistance and reduced systemic flow in the
descending aorta and umbilical artery. Various ratios of systolic-todiastolic flow velocity (amplitude) waveforms have been used, including
the systolic-to-diastolic ratio, systolic-diastolic/systolic ratio (resistance
index), and systolic-diastolic/mean ratio (pulsatility index). Ratios or
indexes greater than two standard deviations from the mean are
associated with IUGR, whereas reversed or absent diastolic waveforms
represent severe fetal hypoxia and increased risk of fetal death. The
most severely affected IUGR fetuses with the greatest risk of death
demonstrate absent or reversed diastolic flow in systemic fetal arteries
(104), along with increased umbilical venous pulsation and reversed flow
in the abdominal vena cava. Interestingly, these same fetuses often have
decreased cerebral (internal carotid artery) pulsatility index, indicating
increased cerebral blood flow (105). This flow pattern has been
interpreted as one way in which brain growth is spared, as body growth

rate slows as a result of placental ischemia and/or placental growth
failure. Doppler waveform abnormalities usually precede less specific
signs of fetal distress, such as abnormal changes in fetal heart rate,
spontaneously or in response to oxytocin challenge testing.
The fetus also should be examined by ultrasound for anatomic
abnormalities that indicate congenital malformations, genetic syndromes,
and deformations. Amniotic fluid index also is useful to identify
oligohydramnios. The latter is a risk factor for congenital anomalies,
severe IUGR with reduced urine production, pulmonary hypoplasia,
variable decelerations from cord compression, and intrauterine fetal
death in as many a 5% to 10% of affected fetuses.
Future Diagnosis and Treatment of

Altered fetal growth rate, and pathophysiology associated with IUGR,
usually develop insidiously, such that once these conditions are clinically
obvious, injury has already taken place. It is important, therefore, to
develop and apply diagnostic techniques to the fetus that would establish
accurately even minimal changes in growth rate and physiologic function.
Currently, Doppler ultrasound measurements of fetal cardiac output,
systemic blood flow, and organ blood supply are close to achieving this
goal, particularly with respect to the placental circulation (107). Other
recent research trials have focused on quantifying placental transfer
functions in pregnant women who are carrying a severely IUGR fetus. For
example, women carrying an IUGR fetus can be given stable isotopes of
nutrients normally transported across the placenta, such as glucose and
selected amino acids, followed by timed cordocentesis to measure
placental transfer characteristics of these substrate isotopes (108). This
information can be compared with that from normal pregnancies with
normal placental function and fetal growth rates.
It is also important that diagnostic techniques are developed to assess
what damage is being done in the more extreme cases of IUGR. Current
techniques include magnetic resonance imaging, Doppler measurements
of blood flow to specific organs (107), cordocentesis (109,110,111,112),
and neurologic and neuromuscular response to vibroacoustic stimulation

(113). Based on such advances in fetal diagnosis, it soon may be
possible to assess whether detected changes in fetal growth rate and
measured fetal pathophysiology associated with IUGR are, in fact, as
serious and indicative of future handicap as current postnatal follow-up
studies have indicated.
Considerably more research is necessary to determine when and how
damage to the fetus can be reversed or ameliorated. Some efforts have
been made in animal models to improve maternal and fetal nutrition
(114) and to enhance development by organ-specific hormone targeting
(115). Examples of the latter include glucocorticoids that have been
administered (such as betamethasone) to both the mother and the fetus
to increase lung surfactant maturity (116), with additional potential
benefits to the maturation of other fetal organs including the gut, heart,
adrenals, and kidneys (115). Also, thyroid-releasing hormone has been
administered along with corticosteroids to increase lung maturity,
although currently there are mixed reports of its effectiveness (117).
Continued work is necessary to assess brain development with such
treatments, and long-term growth and development of all affected
organs. More philosophically, it is important to consider whether it is
wise to change normal developmental relationships among organs by
using organ-specific hormone therapy; more information is needed to
determine how altered patterns of organ development during fetal life
are beneficial and how they may promote, or hinder, normal fetal
development (118).
Antenatal Management
There are few, if any, proven treatments of IUGR. Bed rest and
treatment of acute and chronic illnesses appear beneficial. Having the
mother breathe supplemental oxygen improves fetal oxygenation, and in
a few studies of severe IUGR fetuses with signs of chronic distress this
has been associated with improved rates of fetal growth and reduced
fetal aortic blood flow velocity (increased flow) (119). Trials of low-dose
aspirin therapy, aimed primarily at treating preeclampsia, have not
consistently improved fetal growth (120). Correction of maternal
nutritional deficiencies also is useful, particularly when the mother is

markedly undernourished. Maternal dietary zinc supplements have
improved fetal growth when zinc deficiency was prominent. High protein
intakes have not helped and, in fact, have been associated with worse
IUGR and perinatal morbidity and mortality.
Fetal surveillance techniques should be instituted to determine whether
the fetal condition is beginning to fail and if delivery would be more
likely to result in a successful pregnancy outcome. Traditional fetal
surveillance techniques have included fetal activity recordings, the
oxytocin challenge test, which measures fetal heart rate changes after
oxytocin-induced uterine contractions, and the nonstress test, which
measures the acceleration and beat-to-beat
P.510
variability of the fetal heart rate after spontaneous fetal movement.
These tests, although still done, have been replaced by Doppler
velocimetry and the biophysical profile, which combine analyses of fetal
breathing movements, gross body movements, fetal heart rate, fetal
heart rate reactivity to movement, and estimated amniotic fluid volume.
The combined use of Doppler velocimetry and the biophysical profile has
improved the antenatal management of IUGR (121,122). A low
biophysical profile correlates with fetal hypoxia determined by absent or
reversed diastolic flow in the umbilical artery and fetal blood gas and
acid-base measurements obtained by cordocentesis, and with impending
fetal demise.
Most obstetricians avoid labor when combined fetal surveillance
techniques show severe fetal growth restriction and evidence of severe
chronic distress (123,124), including absent or reversed diastolic flow in
the fetal aorta, increased pulsations and/or reversed flow in the umbilical
veins, and a low biophysical profile score. Fetuses with these conditions
also usually have a nonreactive nonstress test result and a flat baseline
fetal heart rate variability pattern. Such fetuses tolerate labor poorly and
readily develop signs of acute distress. Saline amnioinfusion may be
beneficial in the presence of oligohydramnios and an amniotic fluid index
of less than 5 cm (125). Amnioinfusion to an index greater than 8 cm
may decrease the incidence of meconium-stained fluid, variable

decelerations in the fetal heart rate, end-stage bradycardia, and acute
fetal acidosis. In all such severe cases the delivery should be coordinated
with the neonatology service to provide prompt postnatal evaluation and
care and to prepare for resuscitation of a depressed or asphyxiated
neonate.
In the absence of repeated observations of severe or progressively
worsening IUGR and signs of fetal distress, the moderately affected IUGR
fetus should be left in utero, although providing good nutrition, perhaps
bed rest, and optimal health care to the mother, and continuing fetal
surveillance. Decisions to deliver these fetuses prematurely to prevent
fetal death should be tempered by the difficulties of accurately
diagnosing the worsening of fetal condition and successfully managing all
potential neonatal problems of a preterm infant. Although lung maturity
may be present, the many other problems associated with preterm
delivery should add caution to a decision for early delivery, especially
before 31 to 32 weeks of gestation.
CLINICAL EVALUATION AND TREATMENT OF

THE SMALL-FOR-GESTATIONAL AGE INFANT
General Evaluation in the Delivery Room
SGA infants often present a variety of clinical problems immediately after
birth in the delivery room. Because of their large surface area relative to
body weight, SGA infants lose heat rapidly. To prevent hypothermia, they
should be dried quickly and completely, placed under a radiant warmer,
and protected from drafts with warmed blankets. Severely SGA infants
who suffered marked oxygen and substrate deprivation in utero may
have cardiopulmonary difficulties at birth. Closer to term they may pass
meconium and present with meconium aspiration syndrome, signs of
asphyxia, including hypoxemia, hypotension, mixed metabolic and
respiratory acidosis, and persistent pulmonary hypertension.
Immediately after birth, these infants need prompt and careful attention
to their airway, breathing, and oxygen needs.
Brief Physical Examination in the Delivery

Room
SGA infants have several characteristic features, even when those
infants with obvious anomalies and syndromes and those born to mothers
with severe illness or malnutrition are excluded (126,127,128). Severely
SGA infants who had marked IUGR have relatively large heads for their
undergrown trunks and extremities. The abdomen often appears
shrunken or scaphoid and must be distinguished from infants with
diaphragmatic hernias. The extremities appear scrawny with thin
skinfolds, with evidence of decreased subcutaneous fat and skeletal
muscle. The skin is loose and often rough, dry, and peeling. In term and
postterm severely SGA infants, the fingernails may be long, and the
hands and feet tend to look large for the size of the body. The face
appears shrunken or wizened. Cranial sutures may be widened or
overriding, and the anterior fontanelle may be larger than expected,
representing diminished membranous bone formation. The umbilical cord
often is thinner than usual. When meconium has been passed in utero,
the cord is yellow-green stained, as are the nails and skin.
Gestational Age Assessment of the Small-forGestational-Age Infant

Gestational age assessment based on physical criteria often is erroneous.
Vernix caseosa frequently is reduced or absent as a result of diminished
skin perfusion during periods of fetal distress or because of depressed
synthesis of estriol, which normally enhances vernix production. In the
absence of this protective covering, the skin is continuously exposed to
amniotic fluid and will begin to desquamate. Sole creases appear more
mature as a result of increased wrinkling from increased exposure to
amniotic fluid. Breast tissue formation also depends on peripheral blood
flow and estriol levels and will be reduced in SGA infants. The female
external genitalia will appear less mature because of the absence of the
perineal adipose tissue covering the labia. Ear cartilage also may be
diminished. Specific organ maturity often continues at normal
developmental rates despite diminished somatic growth in most IUGR
infants. Cerebral cortical convolutions, renal glomeruli, and alveolar

maturation all relate to gestational age and are not delayed with IUGR.
P.511
Neurologic Examination of the Small-forGestational-Age Infant

Neurologic examination for gestational age assessment may be little
affected by IUGR. These infants often appear to have advanced
neurologic maturity, although this observation is derived mostly from
comparisons with infants of similar birth weight, not similar gestational
age. Peripheral nerve conduction velocity and visual- or auditory-evoked
responses correlate well with gestational age and are not impaired as a
result of IUGR. These aspects of neurologic maturity are not sensitive to
nutritional deprivation. Active and passive tone and posture are usually
normal in SGA infants and are reliable guides to gestational age,
assuming that infants with significant central nervous system and
metabolic disorders are excluded.
Behavioral Observations
SGA infants demonstrate specific abnormal behaviors. They often have a
hyperalert appearance and generally look starved and hungry, and
they often are described as being jittery and hypertonic, even without
simultaneous hypoglycemia. They may be hyperexcitable, showing
aberrations in tone from hypertonia to hypotonia and, in many cases,
apathy. The Moro response is increased, with exaggerated extension and
abduction of the arms, windmill motions, and prolongation of the tonic
neck posture (129,130). When IUGR is particularly severe, SGA infants
tend to show abnormal sleep cycles and a more consistent picture of
diminished muscle tone, deep tendon and facial tactile reflexes, general
physical activity, and excitability. These more severe changes indicate
that functional central nervous system maturity is impaired, despite the
presence of electrical neurologic maturity (131,132,133). Such severely
SGA infants often appear floppy and develop exhaustion more easily
after handling (134). The behavioral disorders occur even in the absence
of significant central nervous system disease. The hypoexcitability

indicates an adverse effect on polysynaptic reflex propagation and
implies that central nervous system functional maturity does not
necessarily proceed independently of the intrauterine events that result
in IUGR.
Deferred Physical Examination to be Done in

the Neonatal Intensive Care Unit
Careful evaluation is important as there is an increased incidence of
severe malformation, chromosomal abnormality, and congenital infection
among SGA infants (135,136). Dysmorphic features, funny-looking
facies, abnormal hands and feet, and the presence of palmar creases, in
addition to gross anomalies, suggest congenital malformation syndromes,
chromosomal defects, or teratogens. Ocular disorders, such as
chorioretinitis, cataracts, glaucoma, and cloudy cornea, in addition to
hepatosplenomegaly, jaundice, and a blueberry-muffin rash, suggest a
congenital infection. Toxoplasmosis, other (syphilis, hepatitis, Zoster),
rubella, cytomegalovirus, and herpes simplex (maternal infections)
(TORCH) infections resulting in IUGR are unusual in the absence of other
clinical signs of chronic congenital infection; however, screening cord
blood for antibodies and antigens specific to certain infections (which can
be augmented by polymerase chain reaction techniques), and a urine
culture for cytomegalovirus may be indicated. Radiographic examination
of the long bones, looking for possible anomalies and for the quality of
mineralization, may be useful. Examination of the head with
ultrasonography to establish the presence of congenital anatomic
abnormalities or evidence of congenital infection also may be helpful in
making a diagnosis.
CLINICAL PROBLEMS OF THE SMALL-FORGESTATIONAL-AGE NEONATE (TABLE 26-11)

Mortality
The consequences of small size for gestational age depend on the
etiology, severity, and duration of growth restriction. There continues to
be much debate on this subject. Previous studies have included

heterogeneous groups of infants with respect to the degree and cause of
IUGR, the degree of prematurity, and the severity of clinical problems in
the early neonatal period. By now, many studies have been conducted
over extended periods of changing perinatal management and increasing
survival rates of smaller, more preterm infants, many of whom have
been classified differently at different times and among different studies
regarding their degree of IUGR. Some studies have indicated that the
fetus responds to the stress of growth restriction with an acceleration
of maturity, which ultimately is protective for the infant. Others have
found no evidence of improved survival after perinatal stress, and SGA
status has been shown to be an independent predictor of increased fetal,
perinatal, and neonatal death (137). On balance, there is little evidence
to support the concept of improved survival after perinatal stress in SGA
infants, and the perinatal mortality rate for SGA infants with relatively
severe IUGR is 5 to 20 times that of AGA infants of the same gestational
age (136,137). Particularly when adjusted for maternal neonatal risk
factors of IUGR, including birth weight percentile, gestational age at
birth, maternal height, prepregnancy weight, gestational weight gain,
race, and parity, a subgroup of SGA infants is defined that has
consistently and markedly higher perinatal mortality and morbidity rates
than normally grown infants (138).
Intrauterine Growth Restriction/Small-forGestational-Age Status vs. Preterm Birth, and

Effects on Mortality and Morbidity
Preterm and growth-restricted infants have independent and overlapping
problems. Constitutionally small infants are not likely to have increased
risks of mortality or morbidity. As gestational age decreases, the
problems of
P.512
prematurity have a larger role in the outcome of both SGA and AGA
infants. In contrast, the more mature preterm or term infant may suffer
more from the impact of growth restriction. When IUGR has been severe
and prolonged, these infants have a higher perinatal mortality rate than
their AGA peers. Intrauterine fetal death from chronic fetal hypoxia,
immediate birth asphyxia, the multisystem disorders associated with
asphyxia (hypoxic-ischemic encephalopathy, persistent fetal circulation,
cardiomyopathy, meconium aspiration), and lethal congenital anomalies
are the main contributing factors to the high mortality rate for IUGR
fetuses and neonates. Most intrauterine fetal deaths occur between 38
and 42 weeks of gestation. Improved survival depends on achieving an
optimal balance between the consequences of elective preterm delivery
and the risks of continued IUGR.
TABLE 26-11 CLINICAL PROBLEMS OF THE SMALL-FORGESTATIONAL-AGE NEONATE
Problem
Intrauterine death
Pathogenesis/Pathop
Prevention/Trea
hysiology
tment
Chronic hypoxia
Antenatal
Placental insufficiency
surveillance
Growth failure
Malformation
Infection
Infarction/abruption
Preeclampsia
Fetal growth
by ultrasound
Biophysical
profile
Doppler
velocimetry
Maternal
treatment: ? bed
rest, ?O 2
Delivery for
severe/worsenin
g fetal distress
Asphyxia
Acute
Antepartum/intr
hypoxia/abruption
apartum
Chronic hypoxia
monitoring
Placental
Adequate
insufficiency/preeclam
neonatal
psia
resuscitation
Acidosis
Glycogen depletion
Meconium
Hypoxia
Resuscitation
aspiration definite,
including
severe
tracheal
suctioning for
aspiration
Hypothermia
Cold stress
Protect against
Hypoxia
increased heat
Hypoglycemia
loss
Decreased fat stores
Dry infant
Decreased
Radiant
subcutaneous
Persistent
warmer
insulation
Hat
Increased surface
Thermoneutral
area
environment
Catecholamine
Nutritional
depletion
support
Chronic hypoxia
Cardiovascular
pulmonary
support
hypertension
Mechanical
ventilation,
Nitric oxide
Hypoglycemia
Decreased
Frequent
hepatic/muscle
measurement of
glycogen
blood glucose
Decreased alternative
Early
energy sources
intravenous
Heat loss
glucose support
Hypoxia
Decreased
gluconeogenesis
Decreased
counterregulatory
hormones
Increased insulin
sensitivity
Hyperglycemia
Low insulin secretion
Glucose
rate
monitoring
Excessive glucose
Glucose infusion
delivery
<10 mg/min/kg
Increased
Insulin
catecholamine and
administration
glucagon effects
Polycythemia/hyper
Chronic hypoxia
Glucose, oxygen
viscosity
Maternal-fetal
Partial volume
transfusion
exchange
Increased
transfusion
erythropoiesis
Gastrointestinal
Focal ischemia
Cautious enteral
perforation
Hypoperistalsis
feeding
Acute renal failure
Hypoxia/ischemia
Cardiovascular
support
Immunodeficiency
Malnutrition
Early, optimal
Congenital infection
nutrition
Specific
antibiotic and
immune therapy
Asphyxia
Perinatal asphyxia is an uncommon event in SGA infants, but it does
occur at increased frequency in SGA infants and can complicate the
immediate neonatal course of severe IUGR infants. SGA infants
frequently do not tolerate labor and vaginal delivery, and signs of fetal
distress are common. In such cases, the already stressed, chronically
hypoxic infant is exposed to the acute stress of diminished blood flow
during uterine contractions. Cord blood lactate concentrations are often
increased despite overall normal cord blood pH. Preterm SGA infants are
delivered by cesarean section twice as often as preterm AGA infants
(136,139). SGA infants have an increased incidence of low Apgar scores
at all gestational ages (136), and these infants frequently need
resuscitation.
P.513
The acute fetal hypoxia, acidosis, and cerebral depression may result in
fetal death or neonatal asphyxia. Severe IUGR results in a large
proportion of stillborn infants. Myocardial infarction, amniotic fluid and

meconium aspiration, and cerebral edema and necrosis often are noted
at autopsy in these severely IUGR stillborn infants. An inadequate
resuscitation at birth adds double jeopardy to the in utero insults.
Sequelae of perinatal asphyxia include multiple organ system dysfunction
that, in the more severe cases, includes hypoxic-ischemic
encephalopathy, heart failure from hypoxia-ischemia and glycogen
depletion, meconium aspiration syndrome, persistent pulmonary
hypertension, gastrointestinal hypoperistalsis and ischemia- induced
necrosis leading to focal perforation, and acute renal tubular necrosis
and renal failure. However, surfactant-deficiency respiratory distress
syndrome is not increased in more mature SGA infants closer to term,
despite the increased incidence of other forms of respiratory distress.
Hypocalcemia may occur following asphyxia in SGA infants, although this
is uncommon unless there is excessive phosphate released from damaged
cells by acidosis followed by correction with sodium bicarbonate.
Neonatal Metabolism
Hypoglycemia
Hypoglycemia is extremely common in SGA infants, increasing with the
severity of IUGR (Fig. 26-16) (140,141,142,143,144,145,146,147,148).
The risk of hypoglycemia is greatest during the first 3 days of life, but
fasting hypoglycemia, with or without ketonemia, can occur repeatedly
up to weeks after birth. Early hypoglycemia usually is as a result of
diminished hepatic and skeletal muscle glycogen contents (145,147).
Early hypoglycemia is aggravated by diminished alternative energy
substrates, including reduced concentrations of fatty acids from the
scant adipose tissue and decreased concentrations of lactate from the
hypoglycemia. Hyper-insulinemia, increased sensitivity to insulin, or both
may contribute to a greater incidence of hypoglycemia, although there
are very few, if any, accurate measures of insulin sensitivity at different
times and among different conditions in SGA infants to support such
assumptions (142). SGA infants also demonstrate decreased
gluconeogenesis (149), and resolution of persistent hypoglycemia is
coincident with improved capacity for, and rates of, gluconeogenesis.
Deficient counterregulatory hormones also contribute to the pathogenesis

of hypoglycemia in SGA infants (150). Catecholamine release is deficient
in these neonates during periods of hypoglycemia. Although basal
glucagon levels may be elevated, exogenous administration of glucagon
fails to enhance glycemia because of the decreased hepatic glycogen
stores.
Fasting hypoglycemia is now less common in SGA infants as conventional
nutritional treatment now involves earlier enteral and intravenous
feeding and a more liberal use of intravenous glucose (139). All SGA
infants should have early and frequent measurements of blood or plasma
glucose concentrations. Blood glucose concentrations should be kept at
greater than 50 mg/dL. Early enteral feeding usually can prevent
hypoglycemia. In less mature infants or those who have other clinical
problems, intravenous glucose should be started at 4 to 8 mg/min/kg as
soon after birth as possiblepreferably by 30 minutes of age. This initial
infusion rate should be adjusted in response to measurements of blood
glucose concentration every 30 to 60 minutes until these values are
consistently greater than 50 mg/dL. Less frequently repeated
measurements should be continued until the infant is tolerating
reasonably full enteral feedings. Infants with severe hypoglycemia (less
than 20 mg/dL) that is persistent for more than 10 to 30 minutes, with
coma and seizures, should be treated immediately with an intravenous
minibolus of 10% dextrose in water at 200 mg/kg (2 mL/kg), followed
by a glucose infusion of 10% dextrose in water at 4 to 8 mg/min/kg.
Glucose concentrations should be measured at least every 30 minutes
until concentrations are consistently above 45 mg/dL for blood or above
50 mg/dL plasma (or serum). Infants at greatest risk of having severe
hypoglycemia are those who have been asphyxiated and those who are
the thinnest according to the Ponderal index, representing those infants
with the least amount of body glycogen content. Similarly, breast-fed
twins who are not supplemented with a carbohydrate source are at risk,
particularly the smaller twin, and they should be monitored carefully.
Figure 26-16 Incidence of hypoglycemia prior to first feeding (blood

glucose less than 30 mg/dL) among large-for-gestational-age, averagefor-gestational-age, and small-for-gestational-age (SGA) infants,
demonstrating the much greater incidence of early hypoglycemia among
SGA infants at all gestational ages. From Lubchenco LO, Bard H.
Incidence of hypoglycemia in newborn infants classified by birth weight
and gestational age. Pediatrics 1971;47:831, with permission.
P.514
Hyperglycemia
Very preterm SGA infants have developmentally low insulin secretion
rates and plasma insulin concentrations, which may underlie the
relatively common problem of hyperglycemia in ELBW SGA infants (151).
Unnecessarily high rates of glucose infusion (greater than 14 mg/min/kg)
also contribute to this hyperglycemia (152). Higher concentrations of
counterregulatory hormones, such as epinephrine, glucagon, and cortisol,
also may contribute, although there is only limited evidence to support

this commonly held assumption. In contrast, administration of insulin to
even preterm SGA infants usually produces prompt decreases in glucose
concentration, indicating at least normal and probably greater than
normal insulin sensitivity (153).
Lipid Metabolism
SGA infants also have lower plasma free fatty acid levels than normally
grown infants. Fasting blood glucose levels in SGA infants directly
correlate with plasma free fatty acid and ketone body levels.
Additionally, once fed, SGA infants have a deficient utilization of
intravenous triglycerides. After the intravenous administration of
triglyceride emulsion, SGA infants have high free fatty acid and
triglyceride levels, but ketone body formation is attenuated (154,155).
These observations indicate that the utilization and oxidation of free
fatty acids and triglycerides are diminished in SGA neonates. Free fatty
acid oxidation is important because it spares peripheral tissue use of
glucose, whereas the hepatic oxidation of free fatty acids may contribute
the reducing equivalents and energy required for hepatic
gluconeogenesis. Deficient provision or oxidation of fatty acids may be
partly responsible for the development of fasting hypoglycemia in these
infants.
Energy Metabolism
When nursed in a neutral thermal environment, SGA infants demonstrate
the usual decline of the respiratory quotient after birth, representing a
shift toward free fatty acid oxidation. During the first 12 hours after
birth, basal oxygen consumption may be diminished in SGA neonates.
Similar observations have been recorded in utero among spontaneously
SGA fetal lambs, indicating a deficiency of potentially oxidizable
substrates in both situations. Supporting this hypothesis is the marked
increment of oxygen consumption that occurs in well-fed SGA infants
(156), similar to the increase in energy production after nutritional
rehabilitation of infants with marasmic kwashiorkor. The increment of
oxygen consumption after fetal or infantile malnutrition represents the
energy cost of growth. Partly because of enhanced caloric intake, and
because metabolic rate and oxygen consumption are related more to

gestational age than birth weight, SGA infants have a higher oxygen
consumption rate and a higher rate of total energy expenditure
(primarily as a result of increased resting energy expenditure) than less
mature neonates (155). This also reflects an increase in cell number
relative to total mass and greater heat production in response to
increased heat loss. Although some nutritional balance studies of
preterm SGA infants have demonstrated an increase of fecal fat and
protein loss, more recent studies indicate adequate digestion of nutrients
and percent nutrient retention of metabolizable nutrient intake. Thus,
these infants can achieve normal, and occasionally faster, rates of
growth compared with preterm AGA infants.
Those SGA infants who, because of chromosomal or infectious insults
early in gestation, cannot increase their growth rate do not demonstrate
increased rates of oxygen consumption after appropriate caloric intake.
Amino Acid and Protein Metabolism

SGA infants are particularly deficient in muscle mass. Improving
nutrition of skeletal muscle, and total body protein, is a priority in these
infants. There is conflicting information from a limited number of studies,
however, about how well SGA infants tolerate aggressive amino acid and
protein nutrition. ELBW and VLBW SGA infants have higher rates of
protein loss in stools, and of lipids (141), with 11% to 14% lower rates
of absorption. This may be partly compensated for by higher intakes,
which can normalize metabolizable protein intakes. Further, although
growth rate may be increased in SGA infants by increased intake of
protein and nonprotein calories, specific evidence for this comes largely
from preterm infants, some of whom were AGA and some SGA (157).
Also, animal studies show more limited pancreatic development and
intestinal size in SGA offspring (158,159), which may limit feeding
tolerance, protein digestion, and the production of insulin.
Some studies have shown that amino acid turnover rates are higher in
SGA LBW infants (160), but other studies show no difference (161,162).
SGA infants may be more energy efficient in protein synthesis (162).
This could explain faster rates of growth in SGA vs. AGA infants of the
same gestational age, who are fed the same diet (156,157). Thus, SGA
infants possibly may tolerate higher protein intake, but the benefit of
increased intake is not clear (157).
Nutritional Problems and Management

In fetal sheep that are acutely glucose deficient, amino acids are used
for oxidation and for glucose production (163). With chronic glucose
deficiency, protein breakdown remains increased and IUGR develops, but
amino acids are not used for glucose production or oxidation more than
normally (164). It is not clear if human infants born following similar
patterns of IUGR and nutrient deficiency will have similar patterns of
metabolism after birth, nor is it known what types and amounts of
nutrients are best fed to such infants to restore normal metabolism and
to reestablish normal rates of growth as quickly as possible. A rapid rate
of glucose supply can lead to marked hyperglycemia,
P.515
especially in the ELBW preterm SGA infant. On the other hand, amino
acid intolerance is not exaggerated in SGA infants, despite some earlier
evidence that amino acids are not used as readily for gluconeogenesis. If
insulin and IGF-I are deficient in these infants, one would also expect
lower anabolic rates until glucose and amino acid supplies and
concentrations, and production rates of these growth factors, are
restored. Similar issues may apply to lipid tolerance. Such considerations
have prompted some reluctance to feed the SGA infant as aggressively
as their deprived nutritional state would indicate, but large-scale,
rigorous trials of different rates and amounts of nutrition to such infants
have not been conducted. Such trials are needed to determine whether
these infants will tolerate more aggressive feeding and whether this will
result, safely, in improved nutritional rehabilitation, growth, and
perhaps, neurodevelopmental outcome.
Temperature Regulation
Impaired placental function leading to ineffective heat elimination from
the SGA fetus may result in a higher than normal temperature in the
infant at birth (150,153). A normal increase in nonshivering
thermogenesis is seen in these infants because brown fat is available

(165,166). However, depletion of brown fat may occur more readily if
exposure to cold is prolonged (167). In utero stress that depletes
catecholamine stores can contribute to a failure of brown fat to produce
heat. Compared with term infants, SGA infants have a narrow
thermoneutral range. Heat production cannot match the rate of heat loss
with continued cold stress. The rapid heat loss as a result of the large
head-to-body ratio and increased surface area seen in all infants is
exaggerated in the SGA infant. Heat is also lost more quickly through a
thin layer of subcutaneous fat insulation (165,167). Because they are
gestationally more mature than their preterm peers, SGA infants do have
a more generous thermoneutral range for weight and are better able to
maintain the increased metabolic rate necessary to increase heat
production (168). Also, SGA infants of more than 30 weeks of gestation
may have increased skin maturity and less evaporative heat loss than
AGA infants of comparable weight (169), indicating that thermal neutral
environments should be based on gestational age and not weight alone
(170). Heat production may be impaired by concurrent conditions of
hypoglycemia and hypoxia seen commonly in these patients. The normal
response to cold involves increased muscular activity and catecholamine
(norepinephrine) release. Central nervous system depression may
prevent this normal response to cold (165).
During the first few hours of life, oxygen consumption and heat
production may be less than anticipated as a result of decreased
available substrate. Fewer fatty acids are available for oxidation. Later,
as nutritional support is provided, the infant may have higher than
expected oxygen consumption. Brain oxygen requirements are high, and
in the SGA infant brain tissue represents a large proportion of body
weight. Limited availability of glucose in utero limits metabolic rate.
After delivery, as glucose substrate is provided, the brain increases its
metabolic rate and oxygen consumption. Based on brain size, increased
rates of oxygen consumption are appropriate in the SGA infant (171). It
is critical, therefore, that the SGA infant be resuscitated and nursed in a
thermoneutral environment. The newborn should be placed immediately
under a radiant warmer and dried well. A prewarmed hat will minimize
excessive heat loss from the head. A flexed posture decreases exposed
surface area and may slow heat dissipation.
Current and future studies of the value of selective brain cooling of
infants suffering perinatal hypoxic-ischemic encephalopathy may indicate
an advantage for this unique approach in SGA infants as well.
Polycythemia-Hyperviscosity Syndrome
SGA infants manifest an increased incidence of polycythemia (172).
Increased red blood cell volume is likely related to chronic in utero
hypoxia leading to increased erythropoiesis (173,174). Maternal-fetal
transfusion may occur chronically with fetal hypoxia or more acutely with
episodes of fetal distress. Even when not polycythemic (venous
hematocrit greater than 60), SGA infants have higher than normal
hematocrit (174). Approximately half of all term SGA infants have a
central hematocrit above 60% and about 17% of term SGA infants have a
central hematocrit above 65% in contrast to only about 5% in AGA term
infants (175,176). The plasma volume of SGA infants immediately after
birth averages 52 mL/kg, as opposed to 43 mL/kg in AGA infants. Once
equilibrated at 12 hours of life, the plasma volume becomes equivalent in
the two groups. In addition to an enhanced plasma space, the circulating
red blood cell mass is expanded.
Viscosity is directly related to venous hematocrit, and increased viscosity
interferes with normal tissue perfusion. Although the incidence of
hyperviscosity is about 5% in the general population, it is seen much
more frequently (18%) in SGA infants (175). In these cases,
polycythemia is the most likely etiology of hyperviscosity. Most
polycythemic infants remain asymptomatic, but SGA infants are at
greater risk of symptoms and clinical consequence (176). Interestingly,
male SGA infants are at highest risk. Polycythemia contributes to
hypoglycemia and hypoxia. Altered viscosity interferes with neonatal
hemodynamics and results in abnormal postnatal cardiopulmonary and
metabolic adaptation. There is also an increased risk of necrotizing
enterocolitis. In addition to correcting hypoxia and hypoglycemia in
these infants, partial volume exchange transfusion should be considered
to lower hematocrit and minimize the risks of polycythemia and

hyperviscosity.
Immune Function and Infectious Disease Risk

Immunologic function of SGA infants may be depressed at birth and may
persist into childhood, as in older infants with postnatal onset of
malnutrition (177). Deficiencies
P.516
have been demonstrated in lymphocyte number and function, which
include decreased spontaneous mitogenesis and reduced response to
phytohemagglutinin. Similarly, these infants tend to have lower
immunoglobulin levels during infancy and demonstrate an attenuated
antibody response to oral polio vaccine.
Miscellaneous Problems
At birth, cord prealbumin and bone mineral content are low in term SGA
infants (178). Calcium and iron stores may be low as a result of chronic
decreased placental blood flow and insufficient nutrient supply.
Significant hypocalcemia can occur after stressful birth complicated by
acidosis. Thrombocytopenia, neutropenia, prolonged thrombin and partial
thromboplastin times, and elevated fibrin degradation products are also
problems among SGA infants (179,180,181). Sudden infant death
syndrome may be more common after IUGR. Inguinal hernias also
disproportionally follow preterm IUGR.
OUTCOMES AND LONG-TERM CONSEQUENCES

OF SMALL-FOR-GESTATIONAL-AGE INFANTS
Hospitalization
At term, SGA infants are admitted more frequently to the intensive care
unit and have longer hospital stays than their AGA counterparts. Even
among term infants well enough to avoid neonatal intensive care unit
admission as newborns, the incidence of readmission to hospital during
the first year is significantly increased for SGA infants. Neurologic
disorders and other morbidities requiring follow-up and hospitalization
are more frequent among IUGR infants, occurring 5 to 10 times more

often than among AGA infants. SGA infants also are more likely to be
hospitalized for serious respiratory infections, especially if the mother
smoked cigarettes at the time of conception and at other times.
Growth and Developmental Outcome

Most studies of normal and restricted fetal growth and development
support the concept that critical windows of time are present in human
development during which normal growth of certain tissues (e.g., fat,
muscle, bone) or organs (pancreas, brain) must occur. Insults at such
times limit growth and can program persistent, even lifelong failures in
growth and development (182). In rats, for example, undernutrition at a
vulnerable period of brain development permanently decreases brain
size, brain cell number, normal behavioral development, learning, and
memory (183,184). Permanent deficits may result if growth failure
occurs during these critical periods (185).
SGA infants are a heterogeneous group of babies with the potential for a
variety of outcomes. Some are small from genetic or familial causes and
therefore may be expected to achieve their full growth potential and
have normal neurodevelopment (186). Others have specific chromosomal
errors or injury from infections, which are likely to result in severe and
unrecoverable failure of growth and development. Most have a less
defined reason for abnormal in utero growth. The infant with symmetric
growth restriction may have little chance for postnatal catch-up growth
after an early, global disruption of growth. However, a neonate who had
normal growth in early gestation, but developed growth restriction from
limited nutrient availability in later gestation later, has a reasonable
potential for catch-up growth and normal development.
Studies of growth-restricted infants have been plagued by methodologic
problems. Many early studies included all small infants without adequate
distinction between those born at different gestational ages or those with
limited familial or genetic growth potential. Infants with obvious
chromosomal abnormalities and evidence of congenital infection also
were included. Only relatively recently have studies of outcome of IUGR
and SGA infants included in the study design the recognition that the
etiology of small size at birth carries great prognostic value. More recent
studies of this subject also have been limited by uncontrolled
confounding factors. An infant's perinatal morbidity, including inborn or
outborn (requiring transport) status, presence of abnormal umbilical
artery waveforms (121), and a variety of neonatal complications, such as
asphyxia, hypoglycemia, polycythemia, and cold stress, can impact on
ultimate outcome (185,186,187). Multiple gestation and even birth order
can influence future growth potential.
Socioeconomic status and environment are among the most important,
but difficult to control for, variables affecting the growth and
development of SGA infants. Several studies have attempted to
differentiate between the influences of biologic and environmental
variables (187, 188). There are strong associations between
socioeconomic factors and the cognitive development and school
performance of growth-restricted children.
Postnatal Physical Growth of Small-forGestational-Age Infants

Although measurements of weight, length (or height), and head
circumference are standardized and reproducible, many authors have
given more attention to one measurement over another or have been
more concerned with a specific interrelationship of measurements, such
as the Ponderal index (189). In general, SGA infants continue to be
smaller and relatively underweight for age as they grow older, even
through adolescence and early adulthood (Fig. 26-17). These infants
more commonly have short stature as teenagers and young adults,
indicating lifelong growth deficit.
Differences in patterns of early growth have been observed in SGA
infants. Normal infants experience a period of rapid growth during the
first 3 years of life. Adult size correlates with the individual growth curve
after this time. Moderately affected SGA infants who had primarily a
reduction in weight in the third trimester of gestation
P.517
follow the same pattern of normal neonatal and infant growth, but tend
to have an accelerated velocity of growth during the first 6 months
(189). This catch-up growth occurs primarily from birth to 6 months of
age, with some infants continuing an accelerated rate of growth for the
first year. A few of these infants will achieve a normal growth percentile
and thereafter have a growth rate similar to appropriately grown children
(189,190). Head circumference parallels growth in length during catchup and sustained growth periods. After the first year, no difference in
the rate of growth has been noted (189,190).
Figure 26-17 Distribution of height at age 17 years in 30 boys (A) and

34 girls (B) born small for gestational age (solid bars) and their peers
who were born appropriate for gestational age (shaded bars). From Paz
I, Seidman DS, Danon YL et al. Are children born small for gestational
age at increased risk of short stature? AM J Dis Child 1993;147:337.
Ultimate weight and height are less in SGA children when compared to
their normal siblings (189,190). In one study of 4- to 6-year-old children
(189), 45% of siblings were at or above the 50th percentile for weight
and height, whereas only 12% of their SGA siblings achieved the 50th
percentile. Interestingly, a subgroup of severely growth-restricted
infants (less than 40% of expected birth weight) compared to less
affected SGA peers showed no difference in weight or height at 6 months
of age, adding concern for the growth outcome of even modest degrees
of IUGR. Former SGA infants have been shown to have no delay in bone
age, puberty, or sexual maturation at adolescence, although they were
shorter, lighter, and had smaller heads. Muscle mass between the two
groups was similar, but adipose tissue development was less in the SGA
group (191).
Because head size correlates with brain size, volume, weight, and
cellularity, head growth at the time of birth and the degree of catch-up
growth thereafter are prognostic of future neurodevelopment. Deficient
fetal head growth recognized by relative microcephaly at birth, whether
at term or preterm, is felt to be a poor prognostic indicator, as it reflects
the severity and duration of in utero growth failure. A lack of head
sparing and small occipital-frontal circumference (OFC) is associated with
poor neurologic and psychological outcome (192). Head size, if catch-up
head growth has not occurred by 8 months of age, is a predictor of lower
intelligence test scores at 3 years of age (193). This correlation seems to
be independent of environmental or other risks. Decreased head size
when compared to siblings carries significant risks of deficient mental
and motor function (188).
Postnatal Neurodevelopmental Outcome of

Full-Term, Small-for-Gestational-Age Infants
Neurologic disorders and other morbidities are more frequent in SGA
infants as a group, occurring 5 to 10 times more often than in AGA
infants. However, in full-term, mild-to-moderately SGA infants who have
normal brain growth, no hypoxic-ischemic injury, and good
environmental support, IUGR may have little impact on behavior or
mental ability in adolescence or adulthood (185,186, 192). There is no

increased incidence of major handicap (189,191,193,194) and no
increased risk of severe neurologic morbidity (192) in term infants born
small for gestational age. Cerebral palsy is infrequent in this group
(194). Routine neurologic examination is usually normal (191).
Although the absence of gross neurologic outcome in the term SGA infant
is reassuring, evidence of minimal brain dysfunction among these
children continues to be of concern. Many studies have revealed signs of
minor brain damage, including hyperactivity, short attention span,
learning problems, poor fine motor coordination, and hyperreflexia. An
increased number of diffuse abnormalities are seen on
electroencephalogram (194). In one study, 30% of term SGA children at
5 years of age had speech problems, which included a delayed onset of
speech, immature vocabulary with persistent infantile articulation, and
poor receptive and expressive abilities for age (194). In contrast, only
1.5% of the general population and about 5% of their siblings had
speech difficulties.
Former term SGA infants more frequently have substandard school
performance and display subtle neurologic and
P.518
behavioral problems despite a normal IQ. Sensori-motor abilities are
frequently affected. Overall there is no correlation between the severity
of growth restriction and learning deficit, although those with poor early
brain growth in infancy have more problems (193). Measures of cognition
at 4 to 6 years of age correlate well with testing at adolescence. This
suggests that cognitive potential is reached early, with little change
later. It seems likely that environmental and socioeconomic factors play
a significant role in the learning deficiencies seen in these children
(188).
TABLE 26-12 FETAL, NEONATAL/INFANCY, AND ADULT DISORDERS

THAT MIGHT RESULT FROM FETAL PROGRAMMING AS A
CONSEQUENCE OF FETAL UNDERNUTRITION AT DIFFERENT STATES
OF GESTATION a , b
Trimester of Pregnancy
First
Second
Third
Consequences
Low
growth
trajectory
Disturbed
fetalplacental
Brain growth
sustained, but not
relationships body
Fetal
adaptation
Down
regulation
of fetal
growth
Insulin
resistance
Growth factor(s)
resistance/
deficiency
Mixed
Asymmetric
Anthropometry
Symmetric
Infant
growth
Reduced
infant
growth
Reduced
infant
growth
Catch-up growth
possible
Adult life
Increased
BP
Increased
BP,
NIDDM,
Increased BP, NIDDM,

ischemic heart disease
hypercholesterolemia,
ischemic heart disease
BP, blood pressure; NIDDM, noninsulin-dependent diabetes

mellitus.
From Barker D. Mothers, babies, and diseases later in life.

London: BMJ Books, 1994, with permission.
At adolescence, trends toward lower test scores, especially in

mathematics, and an increased incidence of learning disabilities have
been noted (191,194). There also is a trend toward slower alternating
movements on physical examination. Most believe, however, that the
cognitive and academic differences in these children are small and do not
significantly affect school performance or ultimate intellectual ability.
Postnatal Neurodevelopmental Outcome in

Preterm, Small-for-Gestational-Age Infants
The prognosis for preterm SGA infants is less clear and is easily
confounded by other problems of preterm birth. In general, subnormal
intellectual outcomes are more common among preterm SGA infants than
term SGA infants. Some authors have shown that those infants suffering
the dual insults of preterm birth and growth restriction are at higher risk
of neurodevelopmental deficit (195,196). Among extremely preterm
infants, gestational age, and not growth status, has been the most
significant predictor of intellectual outcome (197). By 3 years of age,
social and environmental factors play an increasing role and may impact
on development in either a positive or negative way (197).
Socioeconomic status is independently associated with learning
disabilities in these children (187).
When compared with their appropriately grown same-weight
counterparts, SGA preterm infants have a lower incidence of major
developmental handicaps and cerebral palsy. Specific motor deficits are
more common in AGA infants (198,199). Minor neurologic abnormalities
are more frequently observed in SGA infants (198). Diffuse cerebral
damage as a result of hypoxia and decreased intrauterine blood flow,
especially to the brain, probably accounts for the differences in
expression of brain damage seen in preterm AGA and SGA children (200).
Restricted growth for gestational age is associated with more cognitive
disability in infants who were born preterm (199,201). The need for
special education is higher and becomes apparent at an earlier age in
these preterm SGA infants when they reach school age (200).
Possible Adult Disorders Resulting from

Recent epidemiologic evidence indicates that obesity, insulin resistance,
diabetes, and cardiovascular disease are more common among adults
who were smaller than normal at birth and very likely SGA secondary to
IUGR, particularly those who had a high placental-to-fetal-weight ratio
(202). A variety of animal studies support this concept, including the
greater incidence of obesity, glucose intolerance, plasma lipid
abnormalities, and hypertension in offspring whose mothers were fed a
low-protein diet during pregnancy (203,204,205,206). These examples
indicate that certain adult pathologies may be unavoidable consequences
of environmentally imposed conditions, such as severe and prolonged
fetal undernutrition, which lead to fetal growth restriction to ensure fetal
survival. These conditions may represent an example of programming,
in which an insult, when applied at a critical or sensitive stage in
development, may result in a lasting, even lifelong, effect on the
structure
P.519
or function of the organism (183). IUGR, therefore, is increasingly seen
as an adaptive physiologic process, even though it can produce adverse
fetal, neonatal, and potentially adult consequences (Table 26-12) (207).
Mechanisms responsible for these later-life morbidities in adults, who
were growth restricted in utero, are not yet established. There is some
evidence of diminished pancreatic growth and development, which might
becomes manifest in later life as pancreatic insufficiency when the adult
starts and then continues eating a diet rich in simple carbohydrates and
lipids. Peripheral insulin resistance may develop in the same way, and
hypertension in adulthood may be the result of altered adrenal
development in response to IUGR.
REFERENCES
1. Lubchenco LO, Searls DT, Brazie JV. Neonatal mortality rate:
relationship to birth weight and gestational age. J Pediatr 1972; 81:814.
2. Lubchenco LO. The high risk infant. Philadelphia: WB Saunders, 1976.
3. Chard T, Costeloe K, Leaf A. Evidence of growth retardation in
neonates of apparently normal weight. Eur J Obstet Gynecol Reprod
Endocrinol 1992;45:59.
4. Chard T, Yoong A, Macintosh M. The myth of fetal growth retardation
at term. Br J Obstet Gynaecol 1993;100:1076.
5. Wark L, Malcolm LA. Growth and development of the Lumi child in the
Sepik district of New Guinea. Med J Aust 1969;2:129.
6. Ashcroft MT, Buchanan IC, Lovell HG, et al. Growth of infants and
preschool children in St. Christopher-Nevis-Anguilla West Indies. Am J
Clin Nutr 1966;19:37.
7. Evans MI, Mukherjee AB, Schulman JD. Animal models of intrauterine
growth retardation. Obstet Gynecol Surv 1983;38:183.
8. Simmons RA, Gounis AS, Bangalore SA, et al. Intrauterine growth
retardation: fetal glucose transport is diminished in lung but spared in
brain. Pediatr Res 1992;32:59.
9. Sabbagha RE. Intrauterine growth retardation. In: Sabbagha RE, ed.
Ultrasound applied to obstetrics and gynecology. Philadelphia: JB
Lippincott, 1987:112.
10. McGiven J, Pastor Angladi A. Regulatory and molecular aspects of
mammalian amino acids transport. Biochem J 1993;299:321.
11. Hill RDG. Insulin as a growth factor. Pediatr Res 1985;19:879.
12. Sparks JW, Ross JR, Cetin I. Intrauterine growth and nutrition. In:
Polin RA, Fox WW, eds. Fetal and neonatal physiology, 2nd ed.
Philadelphia: WB Saunders, 1998:267.
13. Lubchenco LO, Hansman C, Boyd E. Intrauterine growth in length and
head circumference as estimated from live births at gestational ages
from 26 to 42 weeks. Pediatrics 1966;37:403.
14. Kramer MS. Intrauterine growth and gestational duration
determinants. Pediatrics 1987;80:502.
15. Virgi SK, Cottington E. Risk factors associated with preterm

deliveries among racial groups in a national sample of married mothers.
J Perinatol 1991;8:347.
16. Klebanoff MA, Schulsinger C, Mednick BR, et al. Preterm and smallfor-gestational-age birth across generations. Am J Obstet Gynecol
1997;176:521.
17. Abrams B, Newman V. Small-for-gestational-age birth: maternal
predictors and comparison with risk factors of spontaneous preterm
delivery in the same cohort. Am J Obstet Gynecol 1991;164:785.
18. Freemark M, Comer M. The role of placental lactogen in the
regulation of fetal metabolism and growth. J Pediatr Gastroenterol Nutr
1989;8:281.
19. Lacroix MC, Devinov E, Servely JL, et al. Expression of the growth
hormone gene in ovine placenta: detection and cellular localization of the
protein. Endocrinology 1996;137:4886.
20. Gluckman PD. The endocrine regulation of fetal growth in late
gestationthe role of insulin like growth factors. J Clin Endocrinol Metab
1995;80:1047.
21. Gibbs RS, Eschenbach DA. Use of antibiotics to prevent preterm
birth. Am J Obstet Gynecol 1997;177:375.
22. Salas SP, Altermatt F, Campos M, et al. Effects of long-term nitric
oxide synthesis inhibition on plasma volume expansion and fetal growth
in the pregnant rat. Hypertension 1995;26:1019.
23. Krebs C, Macara LM, Leiser R, et al. Intrauterine growth restriction
with absent end-diastolic flow velocity in the umbilical artery is
associated with maldevelopment of the placental terminal villous tree.
24. Nicolaides KH, Economides DL, Soothill PW. Blood gases, pH, and
lactate in appropriate- and small-for-gestational-age fetuses. Am J
25. Scholl TO, Hediger ML, Schall JO, et al. Maternal growth during
pregnancy and the competition for nutrients. Am J Clin Nutr
1994;60:183.
26. Wallace JM, Aitken RP, Cheyne MA. Nutrient partitioning and fetal
growth in rapidly growing adolescent ewes. J Reprod Fertil
1996;107:183.
27. Lucas A, Gore SM, Cole TJ, et al. A multicentre trial on feeding low
birthweight infants: effects of diet on early growth. Arch Dis Child
1984;59:722.
28. Davis JA, Dobbing J. Scientific foundations of paediatrics.
29. Ounsted M, Ounsted C. On fetal growth rate. Clinics in developmental
medicine no. 46. Philadelphia: JB Lippincott, 1973.
reference fetus. Growth 1976;40:329.
31. Nimrod CA. The biology of normal and deviant fetal growth. In:
Reece EA, Hobbins JC, Mahoney MJ, et al, eds. Medicine of the fetus &
mother. Philadelphia: JB Lippincott, 1992:285.
32. Widdowson EM. Changes in body proportions and composition during
growth. In: Davis JA, Dobbing J, eds. Scientific foundations of
paediatrics. Philadelphia: WB Saunders, 1974:155.
33. Lapillonne A, Brailon P, Claris O, et al. Body composition in
appropriate and in small for gestational age infants. Acta Paediatr
1997;86:196.
34. Shelley HJ. Glycogen reserves and their changes at birth. Br Med Bull
1961;17:137.
35. Philipps AF. Carbohydrate metabolism of the fetus. In: Polin RA, Fox
WW, eds. Fetal and neonatal physiology, 2nd ed. Philadelphia: WB
Saunders, 1998:560.
36. Hay WW Jr. Glucose metabolism in the fetal-placental unit. In:
Cowett RM, ed. Principles of perinatal-neonatal metabolism, 2nd ed. New
York: Springer-Verlag, 1998:337.
37. Hay WW Jr. Nutrition and development of the fetus: carbohydrate
and lipid metabolism. In: Walker WA, Watkins JB, eds. Nutrition in
pediatrics, 2nd ed. Hamilton: BC Decker, 1996:364.
38. Sparks JW, Girard J, Battaglia FC. An estimate of the caloric
requirements of the human fetus. Biol Neonate 1980;38:113.
39. Milner RDG, Gluckman PD. Regulation of intrauterine growth. In:

Gluckman PD, Heymann MA, eds. Pediatrics & perinatology: the scientific
basis, 2nd ed. London: Arnold, 1993:284.
40. Carver TD, Quick AN Jr, Teng CC, et al. Leucine metabolism in
chronically hypoglycemia, hypoinsulinemic growth restricted fetal sheep.
Am J Physiol 1997;272:E107.
41. Yau K-IT, Chang M-H. Growth and body composition of preterm,
small for gestational age infants at a postmenstrual age of 37-40 weeks.
Early Hum Dev 1993;33:117.
42. Cliver SP, Goldenberg RL, Cutter GR, et al. The effect of cigarette
smoking on neonatal anthropometric measurements. Obstet Gynecol
1995;85:635.
43. Goldenberg RL, Cliver SP, Cutter GR, et al. Blood pressure, growth
retardation, and preterm delivery. Int J Technol Assess Health Care
1992;8:82.
44. Goldenberg RL, Tamura T, Neggers Y, et al. The effect of zinc
supplementation on pregnancy outcome. JAMA 1995;274:463.
45. Goldenberg RL, Hauth JC, DuBard MD, et al. Fetal growth in women
using low-dose aspirin for the prevention of preeclampsia: effect of
maternal size. J Matern Fetal Med 1995;4:218.
46. Battaglia FC, Meschia G. An introduction to fetal physiology. Orlando:
Academic Press, 1986.
47. Lumey LH. Decreased birthweights in infants after maternal in utero
exposure to the Dutch famine of 1944-1945. Paediatr Perinat Epidemiol
1992;6:240.
48. Antonov AM. Children born during the siege of Leningrad in 1942. J
Pediatr 1947;30:250.
49. Stein Z, Susser M, Rush D. Prenatal nutrition and birth weight:
experiments and quasi-experiments in the past decade. J Reprod Med
1978;21:287.
P.520
50. Hickey CA, Cliver SP, Goldenberg RL, et al. Prenatal weight gain,
term birth weight, and fetal growth retardation among high-risk
multiparous black and white women. Obstet Gynecol 1993;81: 529.
51. Rush D, Stein Z, Susser M. A randomized controlled trial of prenatal
nutritional supplementation in New York City. Pediatrics 1980;68:683.
52. Sibai B, Anderson GD. Pregnancy outcome of intensive therapy in
severe hypertension in first trimester. Obstet Gynecol 1986;67: 517.
53. Novy MJ, Peterson EN, Metcalfe J. Respiratory characteristics of
maternal and fetal blood in cyanotic congenital heart disease. Am J
54. Lichty JA, Ting RY, Bruns PD, et al. Studies of babies born at high
altitude. Am J Dis Child 1957;93:666.
55. Brown AK, Sleeper LA, Pegelow CH, et al. The influence of infant and
maternal sickle cell disease on birth outcome and neonatal course. Arch
Pediatr Adolesc Med 1994;148:1156.
56. Goldenberg RL, Davis RO, Nelson KG. Intrauterine growth
retardation. In: Merkatz IR, Thompson JE, Mullen PD, et al, eds. New
perspectives on prenatal care. New York: Elsevier, 1990:461.
57. Abel EL. Consumption of alcohol during pregnancy: a review of
effects on growth and development of offspring. Hum Biol 1982;54:421.
58. Woods JR, Plessinger MA, Clark KE. Effect of cocaine on uterine blood
flow and fetal oxygenation. JAMA 1986;257:957.
59. Molteni RA, Stys SJ, Battaglia FC. Relationship of fetal and placental
weight in human beings: fetal/placental weight ratios at various
gestational ages and birth weight distributions. J Reprod Med
1978;21:327.
60. Owens JA, Falconer J, Robinson JS. Effect of restriction of placental
growth on fetal and utero-placental metabolism. J Dev Physiol
1987;9:225.
61. Beischer NA, Sivasamboo R, Vohra S, et al. Placental hypertrophy in
severe pregnancy anaemia. J Obstet Gynaecol Br Commonw
1970;77:398.
62. Nylund L, Lunell NO, Lewander R, et al. Uteroplacental blood flow
index in intrauterine growth retardation of fetal or maternal origin. Br J
Obstet Gyneacol 1983;90:16.
63. Macara L, Kingdom JC, Kaufman P, et al. Structural analysis of

placental terminal villi from growth-restricted pregnancies with abnormal
umbilical artery Doppler waveforms. Placenta 1996; 17:37.
64. Damsky CH, Fitzgerald ML, Fisher SJ. Distribution of extracellular
matrix components and adhesion receptors are intricately modulated
during first trimester cytotrophoblast differentiation along the invasive
pathway, in vivo. J Clin Invest 1992;89:210.
65. Damsky CH, Librach C, Lim K-H, et al. Integrin switching regulates
normal trophoblast invasion. Development 1994;120:3057.
66. Zhou Y, Damsky CH, Chiu K, et al. Preeclampsia is associated with
abnormal expression of adhesion molecules by invasive cytotrophoblasts.
J Clin Invest 1993;91:950.
67. Genbacev O, Joslin RJ, Damsky CH, et al. Hypoxia alters early
gestation human cytotrophoblast differentiation/invasion in vitro and
models the placental defects that occur in preeclampsia. J Clin Invest
1996;97:540.
68. Handwerger S. The physiology of placental lactogen in human
pregnancy. Endocr Rev 1992;12:329.
69. Freemark M, Handwerger S. The role of placental lactogen in the
regulation of fetal metabolism. J Pediatr Gastroenterol Nutr 1989;8:281.
70. Harding JE, Johnston BM. Nutrition and fetal growth. Reprod Fertil
Dev 1995;7:539.
71. Thureen PJ, Trembler KA, Meschia G, et al. Placental glucose
transport in heat induced fetal growth retardation. Am J Physiol
1992;263:R578.
72. Marconi AM, Cetin I, Davoli E, et al. An evaluation of fetal
gluconeogenesis in intrauterine growth retarded pregnancies. Metabolism
1993;42:860.
73. Molina RD, Meschia G, Battaglia FC, et al. Gestational maturation of
placental glucose transfer capacity in sheep. Am J Physiol 1991;61:R697.
74. Ross JC, Fennessey PV, Wilkening RB, et al. Placental transport and
fetal utilization of leucine in a model of fetal growth retardation. Am J
Physiol 1996;270:E491.
75. Milley JR. Effects of insulin on ovine fetal leucine kinetics and protein
metabolism. J Clin Invest 1994;93:1616.
76. Townsend SF, Briggs KK, Carver TD, et al. Altered fetal liver and
kidney insulin-like growth factor II mRNA in the sheep after chronic
maternal glucose or nutrient deprivation. Clin Res 1992;40:91A.
77. Hay WW Jr. Fetal requirements and placental transfer of nitrogenous
compounds. In: Polin RA, Fox WW, eds. Fetal and neonatal physiology,
2nd ed. Philadelphia: WB Saunders, 1998:619.
78. Milley JR. Ovine fetal leucine kinetics and protein metabolism during
decreased oxygen availability. Am J Physiol 1998;274:E618.
79. Milley JR. Ovine fetal protein metabolism during decreased glucose
delivery. Am J Physiol 1993;265:E525.
80. Bell AW, Kennaugh JM, Battaglia FC, et al. Metabolic and circulatory
studies of the fetal lamb at mid gestation. Am J Physiol 1986;250:E538.
81. Fowden A. Endocrine regulation of fetal growth. Reprod Fertil Dev
1995;7:469.
82. Fowden AL, Hay WW Jr. The effects of pancreatectomy on the rates
of glucose utilization, oxidation and production in the sheep fetus. Q J
Exp Physiol 1988;73:973.
83. Hay WW Jr, Meznarich HK, Fowden AL. The effects of streptozotocin
on rates of glucose utilization, oxidation and production in the sheep
fetus. Metabolism 1988;38:30.
84. Sherwood WG, Chance GW, Hill DE. A new syndrome of pancreatic
agenesis. The role of insulin and glucagon in cell and cell growth. Pediatr
Res 1974;8:360.
85. Carver TD, Anderson SM, Aldoretta PW, et al. Glucose suppression of
insulin secretion in chronically hyperglycemic fetal sheep. Pediatr Res
1995;38:754.
86. Mathews LS, Hammer RE, Behringer RR, et al. Growth enhancement
of transgenic mice expressing human insulin-like growth factor I.
Endocrinology 1988;123:2827.
87. Behringer RR, Lewin TM, Quaife CJ, et al. Expression of Insulin-like
growth factor I stimulates normal somatic growth in growth hormonedeficient transgenic mice. Endocrinology 1990;127:1033.
88. D'Ercole AJ, Dai Z, Xing Y, et al. Brain growth retardation due to the
expression of human insulin like growth factor binding protein 1 (IGFBP-
1) in transgenic mice: an in vivo model for the analysis of IGF function in

the brain. Dev Brain Res 1994;82:213.
89. Ye P, Carson J, D'Ercole AJ. In vivo actions of insulin-like growth
factor-I (IGF-I) on brain myelination: studies of IGF-I and IGF binding
protein-1 (IGFBP-1) transgenic mice. J Neurosci 1995; 15:7344.
90. Delhanty PJD, Han VKM. The expression of insulin-like growth factor
(IGF)-binding protein-2 and IGF-II genes in the tissues of the developing
ovine fetus. Endocrinology 1993;132:41.
91. Wood TL, Rogler L, Streck RD, et al. Targeted disruption of IGFBP-2
gene. Growth Regul 1993;3:3.
92. Pintar JE, Wood TL, Streck RD, et al. Expression of IGF-II, the IGFII/mannose-6-phosphate receptor and IGFBP-2 during rat
embryogenesis. Adv Exp Med Biol 1991;293:325.
93. Wood TL, Streck RD, Pintar JE. Expression of the IGFBP-2 gene in
post-implantation rat embryos. Development 1992;114:59.
94. Crystal RA, Giduice LC. Insulin-like growth factor binding protein
(IGFBP) profiles in human fetal cord sera: ontogeny during gestation and
differences in newborns with intrauterine growth retardation (IUGR) and
large for gestational age (LGA) newborns. In: Spencer EM, ed. Modern
concepts of insulin-like growth factors. New York: Elsevier, 1991:395.
95. Brenneman DE, Eiden LE. Vasoactive intestinal peptide and electrical
activity influence neuronal survival. Proc Natl Acad Sci U S A
1986;83:1159.
96. Hill JM, Agoston DV, Gressens P, et al. Distribution of VIP mRNA and
two distinct VIP binding sites in the developing brain; relation to
ontogenic evens. J Comp Neurol 1994;342:186.
97. Gressens P, Hill JM, Gozes I, et al. Growth factor function of
vasoactive intestinal peptide in whole cultured mouse embryos. Nature
1993;362:155.
98. Gressens P, Hill JM, Paindaveine B, et al. Severe microcephaly
induced by blockade of vasoactive intestinal peptide function in the
primitive neuroepithelium of the mouse. J Clin Invest 1994;94:2020.
99. Hill JM, McCune SK, Alvero RJ, et al. Maternal vasoactive intestinal
peptide and the regulation of embryonic growth in the rodent. J Clin
Invest 1996;97:202.
100. Kliegman RM. Intrauterine growth retardation. In: Fanaroff A,

Martin R, eds. Neonatal-perinatal medicine, 6th ed. St. Louis: MosbyYear Book, 1997:203.
P.521
101. Arduini D, Rizzo G, Romanini C. The development of abnormal heart

rate patterns after absent end-diastolic velocity in umbilical artery:
analysis of risk factors. Am J Obstet Gyencol 1993; 168:43.
102. Beattie RB, Whittle MJ. Doppler and fetal growth retardation. Arch
Dis Child 1993;69:271.
103. Hitschold T, Weiss E, Beck T. Low target birthweight or growth
retardation? Umbilical Doppler flow velocity waveforms and histometric
analysis of fetoplacental vascular tree. Am J Obstet Gynecol
1993;168:1260.
104. Karsdorp VHM, vanVugt JMG, vanGeijn HP, et al. Clinical
significance of absent or reversed end-diastolic velocity waveforms in
umbilical artery. Lancet 1994;344:1664.
105. Noordam MJ, Heydanus R, Hop WC, et al. Doppler colour flow
imaging of fetal intracerebral arteries and umbilical artery in the small
for gestational age fetus. Br J Obstet Gynaecol 1994;101:504.
106. Valcamonico A, Danti L, Frusca T, et al. Absent end-diastolic
velocity in umbilical artery: risk of neonatal morbidity and brain damage.
107. Marconi AM, Ferrazzi E, Cetin I, et al. Umbilical velocimetry
differentiates growth retarded fetuses at risk of acidemia. Third Congress
of the International Perinatal Doppler Society, Los Angeles, CA, 1990.
108. Cetin I, Marconi AM, Baggiani AM, et al. In vivo transplacental
transport of 1-13C-glycine and 1-13C-leucine inhuman pregnancies. Ital
J Gastroenterol 1992;24:8.
109. Marconi AM, Ferrazzi E, Cetin I, et al. Lactate metabolism in normal
and growth retarded human fetuses. Pediatr Res 1990;28:652.
110. Pardi G, Cetin I, Marconi AM, et al. Diagnostic value of blood
sampling in fetuses with growth retardation. N Engl J Med 1993;328:692.
111. Pardi G, Cetin I, Marconi AM, et al. Venous drainage of the human
uterus: respiratory gas studies in normal and fetal growth-retarded
pregnancies. Am J Obstet Gynecol 1992;166:699.
112. Pardi G, Cetin I, Marconi AM, et al. The role of fetal blood sampling
in relation to fetal heart rate and Doppler velocimetry in growth retarded
fetuses. J Soc Gynecol Invest 1993:220.
113. Richards DS. The fetal vibroacoustic stimulation test: an update.
Semin Perinatol 1990;14:305.
114. Charlton V, Johengen M. Effects of intrauterine nutritional
supplementation on fetal growth retardation. Biol Neonate 1985;48: 125.
115. Padbury JF, Ervin MG, Polk DH. Extrapulmonary effects of
antenatally administered steroids. J Pediatr 1996;128:167.
116. Liggins GC, Howie RN. A controlled trial of antepartum
glucocorticoid treatment for the prevention of the respiratory distress
syndrome in premature infants. Pediatrics 1972;50:515.
117. Ikegami M, Polk D, Tabor B, et al. Corticosteroid and thyrotropinreleasing hormone effects on preterm sheep lung function. J Appl Physiol
1991;70:2268.
118. Hay WW Jr, Catz CS, Grave GD, et al. Workshop summary: fetal
growth: its regulation and disorders. Pediatrics 1997;99:585.
119. Battaglia FC, Battaglia C, Artini PG, et al. Maternal
hyperoxygenation in the treatment of intrauterine growth retardation.
120. McFarland P, Pearce JM, Chamberlain GVP. Doppler ultrasound and
aspirin in recognition and prevention of pregnancy-induced hypertension.
Lancet 1990;335:1552.
121. McDonnell M, Serra-Serra V, Gaffney G, et al. Neonatal outcome
after pregnancy complicated by abnormal velocity waveforms in the
umbilical artery. Arch Dis Child 1994;70:F84.
122. Hobbins J. Morphometry of fetal growth. Acta Paediatr Suppl
1997;423:165.
123. Gaziano EP, Knox L, Ferrera B, et al. Is it time to reassess the risk
for the growth-retarded fetus with normal Doppler velocimetry of the
umbilical artery? Am J Obstet Gynecol 1994;170:1734.
124. Gazzolo D, Scopesi FA, Bruschettini PL, et al. Predictors of perinatal

outcome in intrauterine growth retardation: a long-term study. J Perinat
Med 1994;22:71.
125. Strong TH Jr, Hetzler G, Sarno AP, et al. Prophylactic intrapartum
amnioinfusion: a randomized clinical trial. Am J Obstet Gynecol
1990;162:1370.
126. Morrison J, Olsen J. Weight-specific stillbirths and associated causes
of death: an analysis of 765 stillbirths. Am J Obstet Gynecol
1985;152:975.
127. Yogman MW, Kraemer HC, Kindon D, et al. Identification of
intrauterine growth retardation among low birth weight preterm infants.
J Pediatr 1989;115:799.
128. Villar J, deOnis M, Kestler E, et al. The differential neonatal
morbidity of the intrauterine growth retardation syndrome. Am J Obstet
Gynecol 1990;163:151.
129. Frederickson WT, Brown JV. Gripping and moro responses:
differences between small-for-gestational age and normal weight
newborn. Early Hum Dev 1980;4:69.
130. Michaelis R, Schulte FS, Nolte R. Motor behavior of small for
gestation age newborn infants. J Pediatr 1970;76:208.
131. Cruz Martinez A, Perez Conde MC, Ferrer MT. Motor conduction
velocity and H-reflex in infancy and childhood. I: Study in newborns,
twins and small-for-dates. Electromyogr Clin Neurophysiol 1977;17:493.
132. Moosa A, Dubowitz V. Assessment of gestational age in newborn
infants: nerve conduction velocity vs maturity score. Dev Med Child
Neurol 1972;14:290.
133. Schulte FJ, Michaelis R, Linke I, et al. Motor nerve conduction
velocity in term, preterm and small-for-dates newborn infants. Pediatrics
1968;42:17.
134. Als H, Tronick E, Adamson L, et al. The behavior of the full-term but
underweight newborn infant. Dev Med Child Neurol 1976;18:590.
135. Chiswick ML. Intrauterine growth retardation. BMJ 1985;291: 845.
136. Wennergren M, Wennergren G, Vilbergasson G. Obstetric
characteristics and neonatal performance in a four-year small for
gestational age population. Obstet Gynecol 1988;72:615.
137. Piper JM, Xenakis EM-J, McFarland M, et al. Do growth-retarded

premature infants have different rates of perinatal morbidity and
mortality than appropriately grown premature infants? Obstet Gynecol
1996;87:169.
138. Sciscione AC, Gorman R, Callan NA. Adjustment of birth weight
standards for maternal and infant characteristics improves the prediction
of outcome in the small-for-gestational-age infant. Am J Obstet Gynecol
1996;175:544.
139. Hawdon JM, Platt MPW. Metabolic adaptation in small for gestational
age infants. Arch Dis Child 1993;68:262.
140. Antunes JD, Geffner ME, Lippe BM, et al. Childhood hypoglycemia:
differentiating hyperinsulinemic from nonhyperinsulinemic causes. J
Pediatr 1990;116:105.
141. Chessex P, Reichman B, Verellen G, et al. Metabolic consequences
of intrauterine growth retardation in very low birthweight infants. Pediatr
Res 1984;18:709.
142. Collins J, Leonard JV. Hyperinsulinism in asphyxiated and small for
dates infants with hypoglycemia. Lancet 1984;2:311.
143. Frazer T, Karl IE, Hillman LS, et al. Direct measurement of
gluconeogenesis from [2,313C2] alanine in the human neonate. Am J
Physiol 1981;240:615.
144. Haymond MW, Karl IE, Pagliana AS. Increased gluconeogenic
substrates in the small for gestational age infant. N Engl J Med
1974;291:322.
145. Kliegman RM. Alterations of fasting glucose and fat metabolism in
intrauterine growth-retarded newborn dogs. Am J Physiol
1989;256:E380.
146. LeDune M. Response to glucagon in small for dates hypoglycemic
and nonhypoglycemic newborn infants. Arch Dis Child 1972;47:754.
147. Shelly HJ, Neligan GA. Neonatal hypoglycemia. Br Med Bull
1966;22:34.
148. Holtrop PC. The frequency of hypoglycemia in full-term large and
small of gestational age newborns. Am J Perinatol 1993; 10:150.
149. Williams PR, Fiser RH Jr, Sperling MA, et al. Effects of oral alanine
feeding on blood glucose, plasma glucagon, and insulin concentrations in
small for gestational age infants. N Engl J Med 1975;292:612.
150. Hawdon JM, Weddell A, Aynsley-Green A, et al. Hormonal and
metabolic response to hypoglycemia in small for gestational age infants.
Arch Dis Child 1993;68:269.
151. King RA, Smith RM, Dahlenberg GW. Long term postnatal
development of insulin secretion in early premature neonates. Early Hum
Dev 1986;13:285.
152. Cowett RM, Oh W, Pollak A, et al. Glucose disposal of low birth
weight infants: steady state hyperglycemia produced by constant
intravenous glucose infusion. Pediatrics 1979;63:389.
153. Hay WW Jr. Fetal and neonatal glucose homeostasis and their
relation to the small for gestational age infant. Semin Perinatol
1984;8:101.
P.522
154. Bougneres PF, Castano L, Rocchiccioli F, et al. Medium-chain fatty

acids increase glucose production in normal and low birth weight
newborns. Am J Physiol 1989;256:E692.
155. Sabel K, Olegard R, Mellander M, et al. Interrelation between fatty
acid oxidation and control of gluconeogenic substrates in small for
gestational age (SGA) infants with hypoglycemic and with
normoglycemia. Acta Paediatr Scand 1982;71:53.
156. Wahlig TM, Georgieff MK. The effect of illness on neonatal
metabolism and nutritional management. Clin Perinatol 1995; 22:77.
157. Hay WW Jr. Nutritional requirements of the extremel-low-birthweight infant. In: Hay WW Jr, ed. Neonatal nutrition and metabolism. St.
Louis: Mosby-Year Book, 1991:361.
158. De Prins FA, Van Assche FA. Intrauterine growth retardation and
development of endocrine pancreas in the experimental rat. Biol Neonate
1981;1:16.
159. Lebenthal E, Nitzan M, Lee PC, et al. Effect of intrauterine growth

retardation on the activities of fetal intestinal enzymes in rats. Biol
Neonate 1981;39:14.
160. Pencharz PB, Masson M, Desgranges F, et al. Total-body protein
turnover in human premature neonates: effects of birth weight,
intrauterine nutritional status and diet. Clin Sci 1981; 61:207.
161. Cauderay M, Schutz Y, Micheli JL, et al. Energy-nitrogen balances
and protein turnover in small and appropriate for gestational age low
birthweight infants. Eur J Clin Nutr 1988;42:125.
162. FAO/WHO/UNU. Energy and protein requirements. World Health
Organization Technical Report Series. Report of a Joint Expert
Consultation. 1985;724:1.
163. Van Veen LCP, Ten C, Hay WW Jr, et al. Leucine disposal and
oxidation rates in the fetal lamb. Metabolism 1987;36:48.
164. Carver TD, Hay WW Jr. Uteroplacental glucose metabolism and
oxygen consumption after long term hypoglycemic in pregnant sheep. Am
J Physiol 1995;269:E299.
165. Sinclair J. Heat production and thermoregulation in the small for
date infant. Pediatr Clin North Am 1970;17:147.
166. Bhakoo ON, Scopes JW. Minimal rates of oxygen consumption in
small for dates babies during the first week of life. Arch Dis Child
1974;49:583.
167. Aherne W, Hull D. Brown adipose tissue and heat production in the
newborn infant. J Path Bact 1966;91:223.
168. Silverman WA, Sinclair JC, Agate FJ Jr. Oxygen cost of minor
variations in heat balance of small newborn infants. Acta Paediatr Scand
1966;55:294.
169. Hanmerlund K, Sedis G. Transepidermal water loss in newborn
infants. IV. Small for gestational age infants. Acta Paediatr Scand
1980;69:377.
170. Klaus MH, Fanaroff AA, eds. Care of the high-risk neonate, 2nd ed.
171. Sinclair JC, Silverman WA. Intrauterine growth in active tissue mass
of the human fetus, with particular reference to the undergrown baby.
172. Hakanson DO, Oh W. Hyperviscosity in the small-for-gestational age

infant. Biol Neonate 1980;37:109.
173. Cassady G. Body composition in intrauterine growth retardation.
Pediatr Clin North Am 1970;17:79.
174. Snijders RJM, Abbas A, Melby O, et al. Fetal plasma erythropoietin
concentration in severe growth retardation. Am J Obstet Gynecol
1993;168:615.
175. Wirth FH, Goldberg KE, Lubchenco LO. Neonatal hyperviscosity.
Incidence and effect of partial plasma exchange transfusion. Pediatr Res
1975;19:372(abst).
176. Humbert JR, Abelson H, Hathaway WE, et al. Polycythemia in small
for gestational age infants. J Pediatr 1969;75:812.
177. Ferguson S. Prolonged impairment of cellular immunity in children
with intrauterine growth retardation. J Pediatr 1978;93:52.
178. Minton S, Steichen JJ, Tsang RC. Decreased bone mineral content in
small for gestational age infants compared with appropriate for
gestational age infants: normal serum 25-hydroxyvitamin D and
decreasing parathyroid hormone. Pediatrics 1983;71:383.
179. Meberg A. Hermatologic syndrome of growth-retarded infants. Am J
Dis Child 1989;143:1260.
180. Mehta P, Vasa R, Neumann L, et al. Thrombocytopenia in the highrisk infant. J Pediatr 1980;97:791.
181. Perlman M, Dvilansky A. Blood coagulation status of small for dates
and postmature infants. Arch Dis Child 1975;50:424.
182. Lucas A. Programming by early nutrition in man. In: Block GR,
Whelan J, eds. The childhood environment and adult disease (CIBA
Foundation Symposium 156). Chichester: Wiley, 1991:38.
183. Smart J. Undernutrition, learning and memory: review of
experimental studies. In: Taylor TG, Jenkins NK, eds. Proceedings of XII
International Congress of Nutrition. London: John Libbey, 1986:74.
184. Dobbing J. Nutritional growth restriction and the nervous system.
In: Davison AN, Thompson RHS, eds. The molecular bases of
neuropathology. London: Edward Arnold, 1981:221.
185. Hack M. Effects of intrauterine growth retardation on metal

performance and behavior outcomes during adolescence and adulthood.
Eur J Clin Nutr 1998;52:S65.
186. Spinillo A, Stronati M, Ometto A, et al. Infant neurodevelopmental
outcome in pregnancies complicated by gestational hypertension and
intra-uterine growth retardation. J Perinat Med 1993;21:195.
187. Low JA, Handley-Derry MH, Burke SO, et al. Association of
intrauterine fetal growth retardation and learning deficits at age 9 to 11
years. Am J Obstet Gynecol 1992;167:1499.
188. Strauss RS, Dietz WH. Growth and development of term children
born with low birth weight: effects of genetic and environmental factors.
J Pediatr 1998;133:67.
189. Fitzhardinge PM, Steven EM. The small-for-date infant. I. Later
growth patterns. Pediatrics 1972;49:671.
190. Behrman RE. Handicap in the preterm small-for-gestational age
infant. J Pediatr 1979;94:779.
191. Westwood M, Kramer MS, Munz D, et al. Growth and development of
full-term nonasphyxiated small-for-gestational-age newborns: follow-up
through adolescence. Pediatrics 1983;71:376.
192. Berg AT. Indices of fetal growth retardation, perinatal hypoxiarelated factors and childhood neurological morbidity. Early Hum Dev
1989;19:271.
193. Hack M, Breslau N, Weissman B, et al. Effect of very low birth
weight and subnormal head size on cognitive abilities at school age. N
Engl J Med 1991;325:231.
194. Fitzharding PM, Steven EM. The small-for-date infant. II. Neurologic
and intellectual sequelae. Pediatrics 1972;50:50.
195. Pena IC, Teberg AJ, Finello KM. The premature small-forgestational-age infant during the first year of life: comparison by birth
weight and gestational age. J Pediatr 1988;113:1066.
196. Allen MC. Developmental outcome and follow up of the small for
gestational age infant. Semin Perinatol 1984;8:123.
197. Sung I, Vohr B, Oh W. Growth and neurodevelopmental outcome of
very low birth weight infants with intrauterine growth retardation:
comparison with control subjects matched by birth weight and

gestational age. J Pediatr 1993;123:618.
198. Veelken N, Stollhoff K, Claussen M. Development and perinatal risk
factors of very low-birth-weight infants. Small versus appropriate for
gestational age. Neuropediatrics 1992;23:102.
199. Hutton JL, Pharoah POD, Cooke RWI, et al. Differential effects of
preterm birth and small gestational age on cognitive and motor
development. Arch Dis Child 1997;76:F75.
200. Kok JH, den Ouden AL, Verloove-Vanhorick SP, et al. Outcome of
very preterm small for gestational age infants: the first nine years of
life. Br J Obstet Gynaecol 1998;105:162.
201. McCarton CM, Wallace IF, Divon M, et al. Cognitive and neurologic
development of the premature, small for gestational age infant through
age 6: comparison by birth weight and gestational age. Pediatrics
1996;98:1167.
202. Barker DJP. Fetal and infant origins of adult disease. BMJ
1993;301:1111.
203. Snoeck A, Remacle C, Reusens B, et al. Effect of a low protein diet
during pregnancy on the fetal rat endocrine pancreas. Biol Neonate
1990;57:107.
204. Dahri S, Snoeck A, Reusesn B, et al. Islet function in offspring of
mothers on a low protein diet during pregnancy. Diabetes 1991;40:115.
205. Dahri S, Cherif H, Reusens B, et al. Effect of an isolcaloric low
protein diet during gestation in rat on in vitro insulin secretion by islets
of the offspring. Diabetologia 1994;37:A80.
206. Rasschaert J, Reusens B, Dahri S, et al. Impaired activity of rat
pancreatic islet mitochondrial glycerophosphate dehydrogenase in protein
malnutrition. Endocrinology 1995;136:2631.
207. Hay WW Jr, Catz CS, Grave GD, et al. Workshop summary: fetal
growth: its regulation and disorders. Pediatrics 1997;99:585.
Chapter 27
Multiple Gestations
Mary E. Revenis
Lauren A. Johnson-Robbins
The incidence of twins, triplets, and higher-order multiple gestations now accounts
for approximately 3% of all pregnancies in the United States. The products of
multiple gestations comprise a disproportionate number of admissions to neonatal
intensive care units and suffer greater morbidity than do singletons. In 2002 16%
of all preterm deliveries in the United States were due to multifetal gestations (1).
In addition to prematurity, the products of multiple gestations are susceptible to
unique problems which may intensify as the numbers increase. A review of the
major problems can help the clinician to anticipate the medical needs and prepare
the parents for what lies ahead. Most of the issues about twins discussed in this
chapter apply to all multiple gestations.
EPIDEMIOLOGY
The incidence of multiple gestations in the United States has increased
dramatically during the past three decades as a result of the shift in maternal age
distribution to older ages, as well as the increased use of fertility enhancement
therapy. The number of births from twin deliveries and higher-order multiples rose
respectively to 31.1 and 1.84 per 1,000 live births in 2002 (1). The actual rate of
twin conceptions is much higher because early fetal loss with a vanishing twin is
far more common than clinically recognized (2). In 1,000 pregnancies studied
early with ultrasonography, Landy and associates (2) found a twin conception rate
of 3.29%, with subsequent reduction to a single fetus in 21.2% of those
pregnancies.
The incidence of naturally conceived higher-multiple births is mathematically
described by the Hellin-Zeleny law, which states that if twins occur at a frequency
of 1/N, triplets occur at a frequency of (1/N)2, quadruplets at (1/N)3, and so on.
Because most epidemiologic studies exclude data on twins with no live-born
member, they grossly underestimate the incidence of multiple gestations.
Natural monozygotic twinning occurs at a fairly constant rate of 3.5 per 1,000 live
births, with limited variation among populations. The occurrence of monozygosity
is not affected by environment, race, physical characteristics, or fertility. The
relatively new reproductive technology of zona manipulation enhances artificial
reproductive technology success, but is also associated with a remarkable increase
in monoamniotic twinning. Following zona manipulation, 17.3% of the resulting
multiple gestations are monoamniotic. The zona may act as a container for the
dividing cell mass, disruption of which may allow for monozygotic twinning (3).
In contrast, rates for dizygotic twins vary greatly among populations, from 1.3 per
1,000 live births in Japan to 49 per 1,000 live births in Nigeria (4). Other factors
that influence the incidence of dizygotic twinning include a maternally transmitted
familial tendency, race, nutrition, parity, advanced maternal age, coital frequency,
and seasonality. Twins are found most often in black populations and least often
in Asians. Taller, heavier women bear twins at a rate 25% to 30% higher than
short, undernourished women (4). Parity is an independent risk factor, with
multiparous women having a greater likelihood for multiple gestations (5).
Advanced maternal age predisposes to dizygotic twinning, with peak incidence at
37 years of age (5). Coital frequency has a positive affect, with a high rate of twin
conceptions within the first 3 months of marriage (6). Another factor is the effect
of the climatic seasons. In the Northern hemisphere, most dizygotic births are
autumnal, reflecting more multiple ovulations during the winter and spring
months. The seasonality of multiple births does not coincide with the peak months
of singleton births (7).
High circulating levels of follicle-stimulating hormone (FSH) and luteinizing
hormone (LH) lead to the release of more than one ovum per menstrual cycle,
making multizygotic conceptions more likely. Conception stimulants such as
clomiphene citrate (Clomid, Serophene), which act by stimulating endogenous
secretion of gonadotropins, raise the incidence of multiple gestations by 6.8% to
17%; exogenous gonadotropins such as menotropins (FSH and LH; Pergonal) or
human chorionic gonadotropins (APL., Follutein, Pregnyl, Profasi HP) may increase
the incidence as much as 18% to 53.5% (8). The women of the Nigerian Yoruba
tribe, who have naturally elevated levels of FSH and LH, have a remarkably high
rate of spontaneous, dizygotic
P.524
twinning (1 in 20) (9). Martin and associates (10) examined another population
and found that women with dizygotic twins have higher levels of FSH and estradiol
than women bearing singletons. A phenomenon likely due to increased pituitary
gonadotropin release is the twofold higher incidence of twin conceptions in the 2
months after the cessation of oral contraceptives (11). High FSH and LH levels
probably account for the seasonal variation in twinning observed in many
countries (12,13).
IMPACT OF REPRODUCTIVE TECHNOLOGY

The increasing use of ovulation-inducing drugs and assisted reproductive
technology (ART; in vitro fertilization, gamete intrafallopian transfer, and zygote
intrafallopian transfer) has contributed to the recent 65% increase in multiplegestation births in the past two decades. In 2000 there were 99,629 ART
procedures in the United States. Of the 35,025 infants born after these
procedures, 53% were products of multifetal pregnancies (44% twins and 9%
triplets and higher-order multiples) (14). It is estimated that the number of
nonART fertility treatments (ovulation induction and intrauterine insemination) are
comparable in number to the ART procedures and associated with even more
triplet and higher-order multiple gestations. In 2000 only 18% of triplets were
naturally conceived, with 40% a result of ART and estimated 40% due to
ovulation induction (15). In addition to the expected increase in dizygotic twins
following assisted reproduction, monozygotic twinning with its higher incidence of
complications is also more frequent in multiple gestations following assisted
reproduction with an incidence of 3.2%, eight times higher than for spontaneously
conceived pregnancies (16).
The multiple births resulting from ART are the major factor responsible for the
increase in preterm delivery to 12.1%, up 29% from 1981. The use of ART
accounts for 3.5% of low-birth-weight (LBW) infants and 4.3 % of very-low-birthweight infants in the United States because of absolute increases in multiple
gestations and also because of higher rates of LBW among singleton infants
conceived with this technology (17).
The number of embryos transferred during ART procedures is directly related to
the risk of multiple gestations. The multiple rate increases from 33.9% with
transfer of two embryos to 41.4%, 43.2%, and 46.5% with transfer of three, four,
or five or more embryos, respectively. The rate of triplets increases from 0.8%
with transfer of two embryos to 7.4%, 8.4%, and 10.7% with transfer of three,
four, or five or more embryos (14).
Each cycle of ART is expensive and not often covered by insurance. An attempt to
improve pregnancy success has encouraged transfer of multiple embryos during
each procedure, especially for older women. As increasing the number of embryos
transferred with ART increases the risk of multiple gestations and thus the risk of
complications, the live birth rate is not always improved. For women less than age
35, the live birth rate for each embryo transferred is higher when only two
embryos are transferred (42%) than when three (39.7%), four (35.4%), or five or
more (33%) embryos are transferred (14,18).
The increase in multiple gestations due to ART and nonART fertility procedures are
associated with significant expenses due to increased need for perinatal
surveillance and intervention, increased neonatal intensive care utilization, and
long-term costs of care for chronic disabilities such as cerebral palsy. To reduce
health care expenses, several European countries have passed regulations or
guidelines addressing the number of embryos permitted to be transferred during
ART procedures (19) or have agreed to pay the expense of ART cycles if a reduced
number of embryos are transferred. In the United States, the progressively rising
higher-order multiple birth rate finally decreased 9% from 193.5 per 100,000 live
births in 1998 to 180.5 in 2000 (1), possibly indicating some moderation of the
number of embryos transferred during ART procedures. There is ongoing debate

as to whether regulations in the United States would be effective in reducing the
numbers of multiple gestations following ART procedures and improving outcome
at a lower cost.
ZYGOSITY
Zygosity is determined by the number of ova fertilized. Higher-order pregnancies
may be monozygotic, dizygotic, or multizygotic. In 1955 Corner (20) postulated
that monozygotic twins develop by splitting of the conceptus at any time from day
2 after conception through days 15 to 17. The timing of division determines
whether monozygotic twins are dichorionic, monochorionic, or conjoined. Dizygotic
or multizygotic gestations result if more than one ovum has undergone
fertilization at the same coitus or even at different times or with different mates.
At birth, zygosity can be determined by gender differences or by direct placental
examination. Other techniques include blood typing, dermatoglyphics, and
chromosome banding (21,22). The most precise technique is DNA-variant
restriction fragment length polymorphisms (23). Because monozygotic twins carry
significantly higher risks of morbidity and mortality prenatally and postnatally,
establishing the zygosity of all multiple gestations is clinically important. More
effort is going into determining zygosity prenatally using ultrasonography or
genetic identification techniques.
PLACENTATION
The placenta from a twin gestation can be monochorionic or dichorionic; if
dichorionic, it can be fused or separated, making four types of placentation
possible:
Diamniotic, dichorionic separate
Diamniotic, dichorionic fused

P.525
Diamniotic, monochorionic
Monoamniotic, monochorionic
TABLE 27-1 ZYGOSITY DETERMINATION

Percentage of Total
Clinical Finding
Deliveries
Zygosity
Different genders
35
Dizygotic
Monochorionic placenta
Same gender and
20
45
Monozygotic
8% of monozygotic and
dichorionic placenta
37% of placenta
dizygotic infants
Further differentiation can be obtained by genotyping. From Cameron AH.
The Birmingham twin survey. Proc R Soc Med 1968;61:229234, with

permission.
All dizygotic twins have a diamniotic, dichorionic placenta; all monochorionic twins
are monozygotic. Zygosity should be determined in the case of twins of the same
gender if the placenta is not monochorionic, because these siblings may be
monozygotic or dizygotic. Fusion of the placenta does not differentiate zygosity.
Table 27-1 lists zygosity determination based on placental examination.
Benirschke (24) described how to determine chorionicity of a fused placenta based
on examination of the dividing membranes. The amnion contains no blood vessels
and is more transparent than the chorion, which contains fetal vessels and
remnants of villous tissue. A monochorionic placenta is one in which the septum is
composed of a thin, translucent amnion that can be easily separated and lifted
from the chorionic plate. In a dichorionic placenta, the septum is thicker and more
opaque. It does not separate as easily from the chorionic plate. Ultrasonography
of the dividing membranes early in gestation is useful in some cases to determine
the chorionicity, but it is not always technically feasible (25).
A monochorionic, monoamniotic placenta is formed by division of the embryonal
disc at 7 to 13 days, which is after differentiation of the amnion. Only 1% to 2%
of monozygotic twins are monoamniotic; the fetal mortality rate is as high as
50%, primarily due to twisting, knotting, or entanglement of the umbilical cords
(26). Conjoined twins with their necessarily monoamniotic placenta result from
the latest and incomplete splitting of the embryonic disc at days 13 to 15 of
gestation. The monochorionic, diamniotic placenta with a dividing membrane
consisting of two layers of amnion without an intervening chorion is formed at
approximately 5 days of gestation. Dichorionic, diamniotic placentas are formed
the earliest, within the first 3 days after conception.
ANTEPARTUM COMPLICATIONS
Many complications of pregnancy occur more frequently in multiple gestations.
Preterm labor is the most frequent complication, occurring in 20% to 50% of

multiple gestations, most likely due to uterine overdistention. Pregnancy- induced
hypertension, placenta previa, antenatal and intrapartum hemorrhage,
hyperemesis gravidarum, and premature rupture of membranes all occur at a
higher rate (27,28). Polyhydramnios, an almost expected complication of multiple
gestations, is transient in pregnancies in which there are no other complications.
If persistent, the polyhydramnios suggests abnormal fetal conditions, such as twinto-twin transfusion syndrome (TTTS) or congenital anomalies (29).
ANTENATAL MANAGEMENT
Recommendations for managing multiple gestations are controversial. The only
unquestioned aspect of management is the benefit of early diagnosis, facilitating
referral to an appropriate facility for high-risk infants. Antenatal management
includes the following components:
Early diagnosis
Nutritional intervention
Cervical cerclage
Prophylactic tocolysis
Steroid stimulation of fetal lung maturity
Therapeutic amniocentesis
Multifetal reduction
Bed rest
Bed rest beginning before 28 weeks is commonly advised to decrease perinatal

mortality (30). The National Institutes of Health (NIH) Collaborative Study showed
no significant effect of antenatal betamethasone therapy in inducing fetal lung
maturity in twins, but it is important to note that a relatively small number of
twins were enrolled in the study (31). The 1994 NIH Consensus Statement on
antenatal corticosteroids recommends betamethasone for all fetuses between 24
and 34 weeks of gestation, including multiple gestations (32).

There are several approaches to limiting the complications seen in higher-order
multiple gestation. The first is to limit the number of embryos transferred during
in vitro fertilization to two. This results in a reduction in pre- and postnatal
complications in the mother and infants without affecting the pregnancy or takehome baby rates (18). An alternative is the use of multifetal reduction. Reduction,
most often to twins, is usually performed at 9 to 12 weeks of gestation via either
the transvaginal or transabdominal route (33). Reduction of quadruplets to twins
improves overall outcome, but studies on the effect of reduction of triplets to
twins have had conflicting results (33). Infants who are the products of reduced
twin gestations have an increased incidence of impaired fetal growth and a lower
gestational age at delivery when compared with nonreduced twins (17). Selective
termination is used during the second trimester in pregnancies in which one twin
is discordant for a major genetic disease or anomaly (34).
P.526
LABOR AND DELIVERY

The total duration of labor in a twin gestation is similar to a singleton gestation,
with some differences in the lengths of each stage. Friedman and Sachtelben (35)
observed a shorter latent phase during twin labor but a longer active phase and
second stage, probably due to dysfunctional labor in an overdistended uterus.
There are many potential complications associated with delivery of multiple
gestations, including malpresentation, cord prolapse, cord entanglement, vasa
previa, locked twins, and fetal distress. Locked twins occur most often if the chins
interlock to prevent expulsion or extraction of the first twin. Locking occurs at a
rate of 1 per 817 twin gestations, and uterine hypertonicity, monoamniotic
twinning, fetal demise, and decreased amniotic fluid are all contributing factors
(36).
The best method of delivery depends on the number of fetuses, presentation of
the first fetus, and gestational age. Table 27-2 details the frequencies of each
variation of presentation. If both twins are vertex, there is no evidence that
cesarean section improves outcome (37). In vertex-nonvertex twin gestations
longer than 32 weeks, vaginal delivery is recommended (38). Delivery of the
nonvertex second twin can be by total breech extraction or external cephalic
version under ultrasound guidance and epidural anesthesia (39). If the first twin is
nonvertex, delivery is usually by cesarean section.
Mode of delivery of the preterm multiple gestation depends on many factors, only
one of which is the fetal presentation. If premature twins present in a vertexvertex pattern without other complications, vaginal delivery is attempted, whereas
a cesarean section is recommended for all other combinations of presentation if
the gestational age is less than 34 weeks (40). When these recommendations are
followed, there is no effect of mode of delivery or birth order on the incidence of
intracranial hemorrhage in very-low-birth-weight twins (41,42). Extremely lowbirth-weight twins (less than 1,000 grams) have been shown to benefit from
cesarean section, regardless of their positioning, with a reduction in postnatal
mortality (43).
Delayed-interval delivery of multiple gestations is reported (44,45). Management
typically involves the placement of a cervical cerclage, tocolysis, and antibiotic
therapy following the delivery of the first fetus to delay the delivery of the
remaining fetuses as long as fetal distress is not present. The duration of
pregnancy extension is highly variable, with one series achieving a mean of
prolongation of 49 days (45). Delayed delivery allows for fetal maturation through
either antenatal steroid administration or increasing gestational age.
TABLE 27-2 TWIN PRESENTATION
a
Delivery (AB)
Percentage of Total Deliveries
Vertex-vertex
Vertex-nonvertex
42.5
38.4
Nonvertex
19.1
A, first-born twin; B, second-born twin. (From Young BK, Suidan, J,

Antoine C, et al. Differences in twins: the importance of birth order. Am J
Obstet Gynecol 1985;151:915921, with permission).
MORTALITY
The frequency of single fetal demise in multiple gestation is reported as 0.5% to
6.8%, although early ultrasonography suggests a much higher rate of early loss
(2,46,47,48). The causes of antepartum death include cord accidents, vascular
anastomoses with overwhelming blood volume shifts, and velamentous insertion
of the umbilical cord. Velamentous insertion, which makes the cord more
vulnerable to trauma from twisting and compression, is six to nine times more
common with twin gestation and increases the risks for fetal distress and for vasa
previa with fetal hemorrhage (49). Most intrauterine deaths in twins are
associated with monochorionic placentation (46,47,48).
After the demise of a fetal twin, the surviving fetus is at increased risk for
distress, abnormal presentation, or dystocia, and the mother is at risk for
toxemia, chorioamnionitis, or disseminated intravascular coagulation. In
dichorionic twins, if the cause of death is intrinsic only to that fetus, complications
to the surviving co-twin are rare, except from spontaneous premature labor (48).
When one twin dies after at least 15 weeks of gestation in diamniotic pregnancies,
a fetus papyraceous develops. The fetus loses all water content, becomes
compressed, and because of oligohydramnios, may be mistakenly identified on
sonography as a stuck twin. A retained twin may be large enough to hinder labor
mechanically, necessitating cesarean section (50). Before 15 weeks, the fetus is
resorbed; this is called the vanishing twin phenomenon.
For most monochorionic twins, the death of one twin has little adverse effect on
the surviving fetus (51). However, if vascular connections are present, the
surviving twin is at risk for complications related to interfetal blood exchange,
including disseminated intravascular coagulation. After the death of one twin,
partial abruptio placentae, which separates further during labor, may cause
asphyxia or demise of the other twin (47). Fetal transfusion syndrome may be
related to many of the antepartum deaths complicating twin pregnancy
(46,47,48,52).
Fetal mortality for twins 20 or more weeks gestation declined from 31.2 to 20.7
twin fetal deaths per 1,000 live twin births from 1981 to 1997 (53,54). This
improvement is associated with an increased preterm birth rate and reductions of
complications of the placenta, cord, and membranes, intrauterine hypoxia, and
birth asphyxia, suggesting intensified obstetric management of prenatal care and
earlier obstetric intervention. The mortality of multiple gestations higher than two
is greater than for twins because of smaller fetal size and placental or cord
compromise from competition for space (55).
P.527
Multiple gestations account for 10% to 12% of perinatal deaths (1). Increased
frequencies of prematurity, pre-eclampsia, hydramnios, placenta previa, abruptio
placentae, and cord prolapse contribute to increased mortality. The perinatal
mortality rate for triplet pregnancies is reported as 7% to 23% and is strongly
related to gestational age at delivery (56).
Despite recent increased rates of preterm delivery of twins, infant mortality of
twins has improved in the United States and Canada (53,54), with a decrease
from 54 per 1,000 live births in 1983-1984, to 30 per 1,000 live births in 1996
(53) compared to a general infant mortality rate of 6.9 per 1,000 live births in
2000 (1).
TWIN-TO-TWIN TRANSFUSION SYNDROME

Interfetal blood exchange occurs almost exclusively in monochorionic twins with
circulations shared through vascular anastomoses that are present in most
monochorionic placentas. Only 5% to 18% of these communications are
sufficiently imbalanced to produce TTTS, but the actual rate would be higher if all
cases of early fetal death of one twin were identified (27,46,57). Placentas from
twin pregnancies complicated by TTTS have been shown to have significantly
fewer vascular anastomoses, which are more commonly deep than superficial in
location, compared to monochorionic placentas in pregnancies not complicated by
TTTS (58). Vascular anastomoses and TTTS are rare in fused dichorionic placentas
of dizygotic or monozygotic twins (46,49,51,57).
Acute and chronic forms of TTTS have been described (59,60). The onset of
symptoms depends on what type of vessels are in communication, with an
unbalanced, arteriovenous anastomosis and unidirectional shunt leading to the
earliest and most profound symptoms. If anastomoses are balanced (i.e., artery
to artery, vein to vein), the onset and severity of symptoms depend on changes in
perfusion pressures that may be temporary and vary throughout gestation or
become problematic only after delivery or demise of one twin.
Chronic, unidirectional TTTS manifests at any time after 16 weeks and can occur
when an arteriovenous anastomosis joins a high-pressure system with a lowpressure system. The donor twin becomes progressively anemic, hypovolemic,
and growth retarded, with oligohydramnios, and is at risk for tissue hypoxia and
acidosis from reduced perfusion (52,61). The recipient twin becomes polycythemic
and hypervolemic, with polyhydramnios developing from increased urine
production to relieve the circulatory volume overload. Disparities in the weight of
the heart and other viscera and in the size of glomeruli and pulmonary and
systemic arterioles have been reported (52). Both twins are at risk for ischemia,
thromboembolism, disseminated intravascular coagulation, and death. In the
donor twin, there is hypotension and poor tissue perfusion; in the recipient, there
also is poor tissue perfusion from hyperviscosity and polycythemia. Although the
net transfusion is in the direction of the recipient, thrombi can exchange freely in
either direction through vascular anastomoses, resulting in infarcts or the death of
either twin.
Manifestations of TTTS range in severity from mild differences in blood hematocrit
to the extremes of anemia and polycythemia affecting the pair (62,63). In the
most severe cases, the growth-retarded donor twin may die of chronic hypoxia;
the recipient develops congestive heart failure and hydrops and may die.
Premature rupture of membranes, preterm labor, and delivery of compromised,
premature infants are the usual sequelae. The perinatal mortality is 70% or more
(64,65). Prognosis is better if symptoms, diagnosis, and delivery occur at a later
gestational age or if hydrops does not develop.
In rare cases, after death of one twin with TTTS, the polyhydramnios resolves and
a healthy survivor is born at a later time. However, compromising volumes of
blood may be lost from the survivor into the dead twin. Other morbidity results
from the release of thrombogenic material from degenerating fetal tissues,
resulting in disseminated intravascular coagulation, multiple infarcts, and tissue
necrosis in the live twin (46). Possibly related to thromboembolic arterial

occlusion, severe defects, such as porencephaly, multicystic encephalomalacia,
renal cortical necrosis, infarcts of the spleen, cutis aplasia, small bowel atresia,
colonic and appendiceal atresia with horseshoe kidney, hemifacial microsomia,
and necrotic limb, have been observed in the survivor of monochorionic twins
after one fetal demise (46,48,66,67,68,69). An increased incidence of these
defects is not reported in dichorionic twin survivors after the death of a cotwin.
One suggested criterion for diagnosing chronic TTTS is a hemoglobin difference
between twins of 5 g/dL or more. In itself, this is not sufficient to establish TTTS
because large differences in hemoglobin concentration also can occur in separate,
dichorionic placentas (70). Tan and associates (58) diagnosed TTTS if a birth
weight difference of 20% and a hemoglobin difference of 5 g/dL or more was
found. However, a birth weight difference of more than 20% occurs with similar
frequency in monochorionic and dichorionic pregnancies. The smaller twin may
have polycythemia, secondary to intrauterine growth retardation (70). Fetal
transfusion studies using adult cells as markers indicate significant interfetal blood
exchange sufficient to cause discordant growth, and amniotic fluid volumes occur
far more often than differences in hemoglobin concentrations suggest (71).
Because TTTS in all degrees is limited to monochorionic placentations,
differentiating the type of placenta and detecting vascular anastomoses are
important. Measurement of the difference in pulsatility index between fetuses by
Doppler examination of umbilic arterial blood flow has been shown to be helpful in
the diagnosis of TTTS even before the development of fetal hydrops (72).
Acardiac twinning (i.e., reversed arterial perfusion syndrome) is a rare but
interesting variation of TTTS, occurring in 1% of monozygotic twins (73). The
nonviable acardiac
P.528
twin's survival depends on the existence of artery-to-artery and vein-to-vein
anastomoses to the other twin (74). The structurally normal, pump twin provides
the circulation for itself and for its abnormal, acardiac twin, permitting slow
growth of the abnormal twin. This can be detected by pulsed and color Doppler
flow studies demonstrating arterial blood flow perfusing the acardiac twin. The
reversed direction of flow in the umbilical arteries of the acardiac twin (75) leads
to unusual congenital anomalies. Often assuming an amorphous shape, the
cephalic pole is affected most severely, because it is the region most distal to the
retrograde perfusion. The closer and better perfused lower part of the body is
relatively spared (74). The diagnosis of acardiac twinning can be suspected
prenatally by absence or marked undergrowth of the heart, head, and trunk, and
by increased body soft tissue (75,76) of the acardiac twin. Frequent complications
include congestive heart failure of the pump twin developing between 22 and 30
weeks of gestation, with cardiomegaly, hepatomegaly, hydrops, intrauterine
growth failure, maternal hydramnios, preterm delivery, malpresentation, and fetal
distress (74,77). Increased right atrial enlargement, increased reverse flow in the
inferior vena cava, reversed flow in the ductus venosus, and pulsatile flow in the
umbilical vein are early indications of hemodynamic decompensation in the pump
twin (34,64), which can guide timing of intervention. The mortality rate of the
pump twin is 50% to 55%, primarily due to prematurity (74,77).
An acute form of TTTS occurs with rapid transfer of blood through large superficial
artery-to-artery or vein-to-vein anastomoses during labor and delivery, resulting
in a hypovolemic donor and a hypervolemic recipient with similar birth weights
(59,60). The transfusion is from the first to the second twin during the delivery of
the first twin. However, if the first cord clamping is delayed, blood from the
undelivered twin can be transfused into the first infant. The potential for acute
volume changes during labor and delivery of monochorionic twins contributes to
their vulnerability, need for resuscitation, and volume management.
Antenatal management of the TTTS previously was limited to close observation
and bed rest. Acute polyhydramnios, which often complicates TTTS, is managed
with serial amniocenteses of enough amniotic fluid to lessen fetal symptoms (78).
Digoxin has been used successfully to treat cardiac failure in a recipient twin (79).
Endoscopic laser coagulation of connecting vessels is used in treating severe TTTS
(64,65). When the death of both twins is anticipated, selective feticide or
fetectomy of the donor twin may permit survival of the recipient twin (80). In
some instances, decreasing the polyhydramnios seems to stop or ameliorate the
interfetal transfusion dramatically. Serial amniocentesis of the recipient twin with
polyhydramnios has a 50% to 60% survival rate with 25% neurologic
complications, while endoscopic laser photocoagulation of the communicating
vessels has a 52% to 70% survival with only 4% neurologic complications
(34,64,65).
STUCK TWIN
The stuck twin phenomenon occurs in a diamniotic pregnancy if there is a
relatively acute onset of severe disparity in amniotic volumes, with one growthretarded twin in an oligohydramniotic sac compressed against the uterine wall. If
the oligohydramnios is severe enough, this twin may suffer all the complications
of prolonged compression, including pulmonary hypoplasia, abnormal facies, and
orthopedic deformation. The other twin is in a distended, polyhydramniotic sac,
adding to compression of the smaller twin (81).
The stuck twin phenomenon occurs to some degree in as many as 35% of
monochorionic diamniotic twin pregnancies, and it can develop in dichorionic
pregnancies (82). In monochorionic twins, the phenomenon may be related to
TTTS. Other causes, regardless of placentation, include uteroplacental
dysfunction, congenital infection, discordant aneuploidy, and structural
malformations. Both twins are structurally normal in 95% of cases. Disparity in
volumes of amniotic fluid can occur if one twin has structural anomalies that lead
to polyhydramnios (e.g., neural tube defect, upper gastrointestinal obstruction,
congenital heart disease) or oligohydramnios (e.g., ruptured amnion, urinary tract
anomalies, growth retardation) (83). The onset is usually between 18 and 30
weeks of gestation (82). Premature labor, possibly related to uterine distention
from polyhydramnios and to preterm rupture of membranes, develops in most
cases. Without intervention to reverse the fetal compression and uterine
overdistention, the chance of survival of both twins is less than 20% (84).
ASPHYXIA
Despite the clinical impression that the first-born twin does better than the secondborn twin, there is no demonstrable increase in neonatal death in the second-born
twin (37,85). Breech presentation is more frequent, and large placental
abruptions are more common in second twins (85). Differences in the 1-minute
Apgar score, umbilical venous pH, oxygen pressure, and partial carbon dioxide
pressure favor the first-born twin, regardless of route of delivery, placentation,
interval between twins, or presentation (86). The second-born twin has potentially
greater risk for hypoxia and trauma, regardless of the route of delivery,
suggesting physiologic changes after the birth of the first twin. Findings in the
venous blood gases suggest compromised intervillous placental blood flow after
delivery of the first twin as a major factor.
In triplet pregnancies, although preterm labor is the most frequent complication
and most important factor in perinatal morbidity and mortality, the mode of
delivery is also important. If delivery is by cesarean section, the third-born triplet
has a higher 5-minute Apgar score, and the second- and third-born triplets have
increased survival compared with triplets of vaginal delivery (87). If triplets are
delivered by cesarean section, the three triplets have a
P.529
similar acid-base status despite the finding of lower 1-minute Apgar scores for the
third-born triplet (88). The influence of birth order on acid-base status becomes
significant during vaginal births if there is a longer time in utero after delivery of
the first triplet. Triplets of more than 34 weeks of gestation and with birth weights
greater than 2,000 g for each fetus tolerate vaginal delivery more successfully
than smaller triplets (87).
GROWTH
Examination of fetuses between 8 and 21 weeks of gestation show similar weightto-length ratios for singleton and twin fetuses (89). Birth weights of live-born
twins up to 30 weeks of gestation are slightly smaller but similar to singletons of
the same gestational age, indicating that the growth rate is similar in twins and
singletons until 30 weeks of gestation (Table 27-3) (90,91,92). After 30 weeks,
the singleton fetus has accelerated, exponential growth, and twin fetuses have a
more linear rate of growth (93). Triplet growth previously was reported to decline
progressively after 27 weeks of gestation (64). More recent studies indicate that
growth of individual triplets and triplet sets remains linear throughout the third
trimester (94) (See Appendices E-1 and E-2).
Better growth in the third trimester for multiple gestations reflects the positive
impact of more aggressive maternal nutritional and obstetric care management.
In a prospective study of nutritional intervention, the incidence of preterm
delivery, LBW, and very low birth weight was lowered by 30%, 25%, and 50%,
respectively, compared with twin pregnancies without nutritional intervention, but
the rates of intrauterine growth retardation were not affected (95). Singletons are
more likely to have LBW if there had been more than one fetal heart on early
ultrasonography, and twins are more likely to have LBW if there had been more
than two fetal hearts (17), indicating a persistent effect of the prior multiple
condition.
Multiple gestations account for 17% of intrauterine growth retardation, with
higher mortality rates for affected infants, particularly for the growth-retarded
twin if only one is affected (90,96). Monochorionic twins show greater degrees of
intrapair variation in birth weight than dichorionic twins, and true intrauterine
growth retardation occurs more often in monochorionic twins. The individual
members of twin pairs frequently are discordant for the rate of growth due to
TTTS, placental insufficiency, intrauterine crowding, or an unequal impact of
maternal complications that impair growth, such as preeclampsia. Ultimately, the
underlying factor in most instances is a limitation of intrauterine nutrition, which
may be shared unequally by the fetuses.
TABLE 27-3 BIRTH STATISTICS FOR MULTIPLE GESTATIONS
Twins
Triplets
Quadruplets
Gestational Age (average in

weeks)
Birth Weight (average in

grams)
37.1
2,390
33.0
1,720
31.4
1,482
From Newton W, Keith L, Keith D. The Northwestern University

multihospital twin study. IV. Duration of gestation according to fetal sex.
Am J Obstet Gynecol 1984;149:655658, with permission.
b
From Sassoon DA, Castro LC, Davis JL, et al. Perinatal outcome in triplet
versus twin gestations. Obstet Gynecol 1990;75:817820, with permission.
From Collins MS, Bleyl JA. Seventy-one guadruplet pregnancies:management and outcome. Am J Obstet Gynecol 1990;162:13841391, with
permission.
The incidence of discordant fetal growth as measured by biparietal diameter

increases significantly as gestation advances. It is important to differentiate
discordant growth due to TTTS, in which both twins are at increased risk for
morbidity and mortality often before the last trimester, from a twin gestation in
which only one fetus shows growth retardation, which usually becomes evident
during the last trimester. With discordant growth not due to TTTS, the prognosis
for the growth-retarded fetus depends on the severity of the growth failure and its
cause, and the prognosis for the normally growing fetus may not be compromised.
During the postnatal period, the smaller of the discordant twins has an increased
incidence of hypoglycemia and is more likely to have retarded growth and
development during childhood (97,98).
Monozygotic twins have an increased frequency of congenital anomalies compared
with dizygotic twins or singletons (69). Monozygotic twins frequently are
discordant for malformations or for the severity of a given malformation. Some
structural defects are related to the monozygotic twinning process, such as
conjoined twins or some amorphous twins. Early embryonic malformations and
malformation complexes such as sirenomelia, holoprosencephaly, and
anencephaly are increased in monozygotic twins, suggesting a common cause for
monozygotic twinning and early malformation complexes. Structural defects that
result from the disruption of previously normal tissues are associated with the
exchange of circulation in monochorionic twins with vascular connections. Those
defects in which a vascular disruptive cause has been suggested include central
nervous system defects (e.g., microcephaly, porencephalic cysts,
hydranencephaly), gastrointestinal defects (e.g., intestinal atresia), renal cortical
necrosis, hemifacial microsomia, cutis aplasia, and terminal limb defects (67).
Deformations due to crowding and constraint molding of the normal fetus in utero
during late gestation are similar in type and frequency in dizygotic and
monozygotic twins and include foot-positioning deformations. The products of
multiple gestation resulting from the use of ART and nonART fertility enhancement
procedures are not at higher risk for major congenital malformations compared to
naturally conceived multiples. However, limited data show a
P.530
slight but significant increase in the rate of spontaneous sex-chromosome
anomalies (0.8%) (99) (aneuploidies and structural de novo autosomal
aberrations) after intracytoplasmic sperm injection (which is used in cases of male
infertility), compared to 0.2% in the general population. Most of the abnormalities
are paternally transmitted, consistent with the higher incidence of chromosomal
abnormalities in the sperm of men with infertility.
Conjoined twins represent a unique structural defect of monozygotic
monoamnionic twins. The nonseparated parts of the otherwise normal twins
remain fused throughout the remaining period of development (100). The

incidence of conjoined twins is between 1 in 80,000 to 1 in 25,000 births, and
70% to 80% of these cases are female twins (101). Approximately 40% are
joined at the chest (thoracopagus), 34% at the anterior abdominal wall
(xiphopagus or omphalopagus), 18% at the buttocks (pygopagus), 6% at the
ischium (ischiopagus), and 2% at the head (craniopagus). With ultrasonography,
the diagnosis of conjoined twins can be established as early as week 12 of
gestation (102). Forty percent of conjoined twins are stillborn, and an additional
35% survive only 1 day (103). Long-term survival with or without surgical
separation depends on the ana-tomic site of attachment and the extent of shared
organs (104).
NEONATAL DISORDERS
Prematurity
The preterm birth rate of twins in the United States increased from 40.9% of twin
gestations in 1981 to 55% in 1997. This increase was related to more aggressive
prenatal surveillance, an increase in labor induction, and an increase in first
cesarean delivery (53). Canadian data similarly shows an increase in the preterm
birth rate of twins, with a decline in stillbirth rates at and near term gestation
(54). Despite increased rates of preterm delivery of twins, infant mortality of
twins has improved in the United States and Canada (53,54). Ninety percent of
triplets are born prematurely. In 2002, 12 % of twins, 36% of triplets, and 60% of
quadruplets were born prior to 32 completed weeks of gestation (1).
Hyaline Membrane Disease

Twins are at increased risk of developing hyaline membrane disease due to the
increase in preterm delivery. Both twins are usually affected by hyaline membrane
disease. However, if only one twin is affected, it is usually the second-born twin,
who had a lower Apgar score at 1 minute and a higher birth weight than the firstborn twin (105). The greater risk to the second-born twin probably is related to
birth asphyxia (105). Monozygotic twins more often are born prematurely and at
an earlier gestational age than dizygotic twins, and they are more prone to
develop hyaline membrane disease.
Unique risk factors for the development of necrotizing enterocolitis have not been
identified for twins or higher multiple gestations, but as a group, they are at an
increased risk due to the greater likelihood of prematurity and LBW. Comparisons
of twins showed that the most significant factor in predicting the occurrence of
necrotizing enterocolitis and need for surgical intervention was a lower 1-minute
Apgar score for the affected twins, predominantly the second-born twin, compared
with unaffected cotwins (106). Samm and associates (107) found that, in all their
case pairs, it was the first-born twin who had developed necrotizing enterocolitis;
in no case did only the second twin have necrotizing enterocolitis. In that study,
the first-born infants were more stable, were fed sooner, and had feedings
advanced more rapidly than the second-born twins, implicating feeding practices
in the higher incidence of necrotizing enterocolitis for the first-born twin.
Infection
One early study reported an increased rate of early-onset group B streptococcal
disease in LBW twins compared to LBW singletons (108). Subsequent large
population-based studies failed to show an increased risk of early-onset group B
streptococcal disease in multiple gestations independent of prematurity
(109,110). If just one of a pair of twins is infected or colonized with group B
Streptococcus in utero, it is most likely the twin positioned adjacent to the cervix,
with the exposure due to ascending spread of group B Streptococcus through the
membranes. Spread of infection through the vascular connections between
monochorionic twins has not been documented, although it is theoretically
possible. However, spread of group B Streptococcus from the amniotic fluid of an
exposed twin to a co-twin may occur through intact dividing membranes (111).
The risk of neonatal listeriosis infection is increased with multiple gestations, to
2.8 and 21 times the risk for twin and triplet pregnancies, respectively, compared
to singleton births (112). The risk is especially increased when maternal age is
greater than 35 years. It is possible that the increased production of hormones or
other inhibitors due to larger placental mass with multiple gestations versus
singletons decreases immunity to Listeria. Discordance of infection is 66%, with
the first-born twin at greater risk.
One study showed that multiple-birth preterm infants with bronchopulmonary
dysplasia are at an increased risk of developing respiratory syncytial virus illness
and pneumonia than are singletons matched for gestational age, and that if one
member of a multiple gestation developed respiratory syncytial virus disease,
usually the other member(s) did also. Other risk factors that contributed to this
were the higher density of adults and children in the households of multiple
gestations (113).
P.531
Sudden Infant Death Syndrome

Monozygotic and dizygotic twins are at some increased risk of sudden infant death
syndrome (SIDS) compared with singletons, and this is especially true for LBW
pairs (114). If the birth weights of the twins differ significantly, it is usually the
smaller twin who dies of SIDS (114). For twins discordant for size, the risk of
SIDS for the smaller of twins is greater than for LBW and premature singletons or
other groups of infants at high risk for SIDS (114). It is unusual for the surviving
cotwin also to die of SIDS.
POSTNEONATAL CARE AND FOLLOW-UP

In addition to the long-term impact of some of the perinatal conditions previously
mentioned, twins and higher multiples continue to be at risk for medical,
developmental, and social problems beyond those experienced by singletons born
at similar gestational ages. A brief list of factors that should be considered in
following these patients is included to assist the clinician in anticipating the
problems, many of which can be lessened by preventive measures, such as the
following:
Parental stress of child rearing (115,116)
Child abuse and neglect (116,117)
Intratwin favoritism (118)
Developmental delay (e.g., performance below chronologic age, especially

in language and speech) (119)
Mental retardation (120,121)
Cerebral palsy (120,122), Growth delay (123,124)
Glucose intolerance (125)
REFERENCES
1. Martin JA, Hamilton BE, Sutton PD, et al. Births: final data for 2002. National
Vital Statistics Reports 52(10). Hyattsville, MD: Centers for Disease Control and
Prevention/National Center for Health Statistics, 2003.
2. Landy HJ, Weiner S, Corson SL, et al. The vanishing twin: ultrasonographic
assessment of fetal disappearance in the first trimester. Am J Obstet Gynecol
1986;155:14-19.
3. Slotnick RN, Ortega JE. Monoamniotic twinning and zona manipulation: a

survey of U.S. IVF centers correlating zona manipulation procedures and highrisk twinning frequency. J Assist Reprod Genet 1996;13:381-385.
4. Nylander PP. Biosocial aspects of multiple births. J Biosoc Sci Suppl

1971;3:29-38.
5. Bulmer MG. The effect of parental age, parity and duration of marriage on
the twinning rate. Ann Hum Genet 1959;23:454-458.
6. James WH. Dizygotic twinning, marital stage and status and coital rates. Ann
Hum Biol 1981;8:371-378.
7. Picard R, Fraser D, Hagay ZJ, et al. Twinning in southern Israel. Seasonal

variation and effects of ethnicity, maternal age and parity. J Reprod Med
1990;35:163-167.
8. Schenker JG, Yarkoni S, Granat M. Multiple pregnancies following induction of

ovulation. Fertil Steril 1981;35:105-123.
9. Nylander PP. Serum levels of gonadotrophins in relation to multiple

pregnancy in Nigeria. J Obstet Gynaecol Br Commonw 1973; 80:651-653.
10. Martin NG, Olsen ME, Theile H, et al. Pituitary-ovarian function in mothers
who have had two sets of dizygotic twins. Fertil Steril 1984;41:878-880.
11. Bracken MB. Oral contraception and twinning: an epidemiologic study. Am J

12. Timonen S, Carpen E. Multiple pregnancies and photoperiodicity. Ann Chir

Gynaecol Fenn 1968;57:135-138.
13. Elwood JM. Maternal and environmental factors affecting twin births in
Canadian cities. Br J Obstet Gynaecol 1978;85:351-358.
14. Wright VC, Schieve LA, Reynolds MA, et al. Assisted reproductive
technology surveillanceUnited States, 2000. MMWR Surveill Summ 2003;52:116.
15. Contribution of assisted reproduction technology and ovulation- inducing

drugs to triplet and higher-order multiple birthsUnited States, 1980-1997.
MMWR Morb Mortal Wkly Rep 2003; 49:535-538.
16. Wenstrom KD, Syrop CH, Hammitt DG, et al. Increased risk of
monochorionic twinning associated with assisted reproduction. Fertil Steril
1993;60:510-514.
17. Schieve LA, Meikle SF, Ferre C, et. al. Low and very low birth weight in
infants conceived with use of assisted reproductive technology. N Engl J Med
2002;346:731-737.
18. Templeton A, Morris JK. Reducing the risk of multiple births by transfer of
two embryos after in vitro fertilization. N Engl J Med 1998;339:573-577.
19. Katz P, Nachtigall R, Showstack J. The economic impact of the assisted

reproductive technologies. Nat Cell Biol 2002;4 Suppl:S29-S32.
20. Corner GW. The observed embryology of human single-ovum twins and
other multiple births. Am J Obstet Gynecol 1955;70: 933-951.
21. Robertson JG. Blood grouping in twin pregnancy. J Obstet Gynaecol Br

Commonw 1969;76:154-156.
22. McCracken AA, Daly PA, Zolnick MR, et al. Twins and Q-banded
chromosome polymorphisms. Hum Genet 1978;45:253-258.
23. Hill AV, Jeffreys AJ. Use of minisatellite DNA probes for determination of
twin zygosity at birth. Lancet 1985;2:1394-1395.
24. Benirschke K. Multiple pregnancy. In: Fox W, Polin R, eds. Fetal and
neonatal physiology. Philadelphia: WB Saunders, 1998:115-123.
25. Robinson JN, Abuhamad AZ. Determining chorionicity and amnionicity in

multiple pregnancies. Contemp Ob/Gyn 2002;6: 92-108.
26. Timmons JD, De Alvarez RR. Monoamniotic twin pregnancy. Am J Obstet

Gynecol 1963;86:875-881.
27. Newton ER. Antepartum care in multiple gestation. Semin Perinatol

1986;10:19-29.
28. Polin JI, Frangipane WL. Current concepts in management of obstetrics

problems for pediatricians. II. Modern concepts in the management of multiple
gestation. Pediatr Clin North Am 1986;33:649-661.
29. Hashimoto B, Callen PW, Filly RA, et al. Ultrasound evaluation of

polyhydramnios and twin pregnancy. Am J Obstet Gynecol 1986;154:10691072.
30. Gilstrap LC 3rd, Hauth JC, Hankins GD, et al. Twins: prophylactic
hospitalization and ward rest at an early gestational age. Obstet Gynecol
1987;69:578-581.
31. Collaborative Group on Antenatal Steroid Therapy. Effect of antenatal

dexamethasone administration on the prevention of respiratory distress
syndrome. Am J Obstet Gynecol 1981;141: 276-287.
32. Effect of corticosteroids for fetal maturation on perinatal outcomes. NIH

Consens Sci Statements 1994;12:1-24.
33. Bush M, Eddleman KA. Multifetal pregnancy reduction and selective

termination. Clin Perinatol 2003;30:623-641.
34. Moldenhauer JS, Gillert A, Johnson A. Vascular occlusion in the

management of complicated multifetal pregnancies. Clin Perinatol 2003;30:601621.
35. Friedman EA, Sachtelben MR. The effect of uterine overdistention on labor I.
Multiple pregnancy. Obstet Gynecol 1964;23: 164-172.
36. Cohen M, Kohl SG, Rosenthal AH. Fetal interlocking complicating twin
gestation. Am J Obstet Gynecol 1965;91:407-412.
37. McCarthy BJ, Sachs BP, Layde PM, et al. The epidemiology of neonatal
death in twins. Am J Obstet Gynecol 1981;141:252-256.
38. Adam C, Allen AC, Baskett TF. Twin delivery: influence of presentation and
method of delivery on the second twin. Am J Obstet Gynecol 1991;165:23-27.
39. Chervenak FA. The controversy of mode of delivery in twins: the

intrapartum management of twin gestation (part II). Semin Perinatol
1986;10:44-49.
P.532
40. Cetrulo CL. The controversy of mode of delivery in twins: the intrapartum
management of twin gestation (part I). Semin Perinatol 1986;10:39-43.
41. Morales WJ, O'Brien WF, Knuppel RA, et al. The effect of mode of delivery
on the risk of intraventricular hemorrhage in nondiscordant twin gestations
under 1500 g. Obstet Gynecol 1989;73:107- 110.
42. Pearlman SA, Batton DG. Effect of birth order on intraventricular

hemorrhage in very low birth weight twins. Obstet Gynecol 1988; 71:358-360.
43. Zhang J, Bowes WA Jr, Grey TW, et al. Twin delivery and neonatal and
infant mortality: a population-based study. Obstet Gynecol 1996;88:593-598.
44. Lavery JP, Austin RJ, Schaefer DS, et al. Asynchronous multiple birth. A
report of five cases. J Reprod Med 1994;39;55-60.
45. Arias F. Delayed delivery of multifetal pregnancies with premature rupture

of membranes in the second trimester. Am J Obstet Gynecol 1994;170:12331237.
46. Benirschke K. Twin placenta in perinatal mortality. NY State J Med

1961;61:1499-1508.
47. Litschgi M, Stucki D. [Course of twin pregnancies after foetal death in

utero.] Z Geburtschilfe Perinatol 1980;184:227-230.
48. D'Alton ME, Newton ER, Cetrulo CI. Intrauterine fetal demise in multiple
gestation. Acta Genet Med Gemellol 1984;34:43-49.
49. Benirschke K. Multiple gestation: incidence, etiology and inheritance. In:

Creasy RK, Resnik R, eds. Maternal-fetal medicine: principles and practice.
50. Leppert PC, Wartel L, Lowman R. Fetus papyraceus causing dystocia:

inability to detect blighted twin antenatally. Obstet Gynecol 1979;54:381-383.
51. Johnson SF, Driscoll SG. Twin placentation and its complications. Semin
Perinatol 1986;10:9-13.
52. Naeye R. Human intrauterine parabiotic syndrome and its complications. N

Engl J Med 1963;268:804-809.
53. Kogan MD, Alexander GR, Kotelchrch M, et. al. Trends in twin birth
outcomes and prenatal care utilization in the United States, 1981-1997. JAMA
2000;284:335-341.
54. Joseph KS, Marcoux S, Ohlsson A, et al. Changes in stillbirth and infant
mortality associated with increases in preterm birth among twins. Pediatrics
2001;108:1055-1061.
55. McKeown T, Record RG. Observations on foetal growth in multiple

pregnancy in man. J Endocrinol 1952;8:386-401.
56. Egwuata VE. Triplet pregnancy: a review of 27 cases. Int J Gynaecol Obstet
1980;18:460-464.
57. Robertson EG, Neer KJ. Placental injection studies in twin gestation. Am J
58. Tan KL, Tan R, Tan SH, et al. The twin transfusion syndrome. Clinical
observations on 35 affected pairs. Clin Pediatr (Phila) 1979;18:111-114.
59. Bajoria R, Wigglesworth J, Fisk NM. Angioarchitecture of monochorionic

placentas in relation to the twin-twin transfusion syndrome. Am J Obstet
Gynecol 1995;172:856-863.
60. Klebe JG, Ingomar CJ. The fetoplacental circulation during parturition
illustrated by the interfetal transfusion syndrome. Pediatrics 1972;49:112-116.
61. Dudley DK, D'Alton ME. Single fetal death in twin gestation. Semin Perinatol
1986;10:65-72.
62. Benirschke K, Driscoll SG. The pathology of the human placenta. New York:
Springer-Verlag, 1967:87-.
63. In: Fox H. Pathology of the placenta. Philadelphia: WB Saunders, 1978:81-.
64. Quintero RA, Morales W. Operative fetoscopy: a new frontier in fetal

medicine. Contemp Ob/Gyn 1999;44:45-68.
65. Quintero, RA. Twin-twin transfusion syndrome. Clin Perinatol 2003;30:591600.
66. Galea P, Scott JM, Goel KM. Feto-fetal transfusion syndrome. Arch Dis Child
1982;57:781-783.
67. Hoyme HE, Higginbottom MC, Jones KL. Vascular etiology of disruptive
structural defects in monozygotic twins. Pediatrics 1981;67:288-291.
68. Mannino FL, Jones KL, Benirschke D. Congenital skin defects and fetus
papyraceus. J Pediatr 1977;91:559-564.
69. Schinzel AA, Smith DW, Miller JR. Monozygotic twinning and structural
defects. J Pediatr 1979;95:921-930.
70. Danskin FH, Neilson JP. Twin-to-twin transfusion syndrome: what are
appropriate diagnostic criteria? Am J Obstet Gynecol 1989;161:365-369.
71. Fisk NM, Borrell A, Hubinont C, et al. Fetofetal transfusion syndrome: do the
neonatal criteria apply in utero? Arch Dis Child 1990;65:657-661.
72. Ohno Y, Ando H, Tanamura A, et al. The value of Doppler ultrasound in the
diagnosis and management of twin-to-twin transfusion syndrome. Arch Gynecol
Obstet 1994;255:37-42.
73. Napolitani FE, Schreiber I. The acardiac monster. A review of the world
literature and presentation of 2 cases. Am J Obstet Gynecol 1960;80:582-589.
74. Van Allen MI, Smith DW, Shepard TH. Twin reversed arterial perfusion
(TRAP) sequence: a study of 14 twin pregnancies with acardius. Semin Perinatol
1983;7:285-293.
75. Billah KL, Shah D, Odwin C. Ultrasonic diagnosis and management of

acardius acephalus twin pregnancy. J Med Ultrasound 1984;8:108-113.
76. Stiller RJ, Romero R, Pace S, et al. Prenatal identification of twin reversed
arterial perfusion syndrome in the first trimester. Am J Obstet Gynecol
1989;160:1194-1196.
77. Moore TR, Gale S, Benirschke K. Perinatal outcome of forty-nine

pregnancies complicated by acardiac twinning. Am J Obstet Gynecol
1990;163:907-912.
78. Radestad A, Thomassen PA. Acute polyhydramnios in twin pregnancy. A

retrospective study with special reference to therapeutic amniocentesis. Acta
Obstet Gynecol Scand 1990;69:297-300.
79. De Lia J, Emery MG, Sheafor SA, et al. Twin transfusion syndrome:
successful in utero treatment with digoxin. Int J Gynaecol Obstet 1985;23:197201.
80. Wittman BK, Farquarson DF, Thomas WD, et al. The role of feticide in the
management of severe twin transfusion syndrome. Am J Obstet Gynecol

1986;155:1023-1026.
81. Urig MA, Clewell WH, Elliott JP. Twin-twin transfusion syndrome. Am J
Obstet Gynecol 1990;163:1522-1526.
82. Chescheir NC, Seeds JW. Polyhydramnios and oligohydramnios in twin

gestations. Obstet Gynecol 1988;71:882-884.
83. Pretorius DH, Mahony BS. Twin gestations. In: Nyberg DA, Mahony BS,
Pretorius DH, eds. Diagnostic ultrasound of fetal anomalies. Chicago: Year Book
Medical, 1990:609-622.
84. Mahony BS, Petty CN, Nyberg DA, et al. The stuck twin phenomenon:
ultrasonographic findings, pregnancy outcome, and management with serial
amniocenteses. Am J Obstet Gynecol 1990;163:1513-1522.
85. Naeye RL, Tafari N, Judge D, et al. Twins: causes of perinatal death in 12
United States cities and one African city. Am J Obstet Gynecol 1978;131:267272.
86. Young BK, Suidan J, Antoine C, et al. Differences in twins: the importance
of birth order. Am J Obstet Gynecol 1985;151:915-921.
87. Deale CJ, Cronje HS. A review of 367 triplet pregnancies. S Afr Med J
1984;66:92-94.
88. Creinin M, MacGregor S, Socol M, et al. The Northwestern University triplet

study. IV. Biochemical parameters. Am J Obstet Gynecol 1988;159:1140-1143.
89. Iffy L, Lavenhar MA, Jakobovits A, et al. The rate of early intrauterine
growth in twin gestation. Am J Obstet Gynecol 1983; 146:970-972.
90. Hendricks CH. Twinning in relation to birth weight, mortality, and congenital
anomalies. Obstet Gynecol 1966;27:47-53.
91. Naeye RL, Benirschke K, Hagstrom JW, et al. Intrauterine growth of twins
as estimated from liveborn birth-weight data. Pediatrics 1966;37:409-416.
92. Wilson RS. Twins: measures of birth size at different gestational ages. Ann
Hum Biol 1974;1:57-64.
93. Arbuckle TE, Sherman GJ. An analysis of birth weight by gestational age in
Canada. CMAJ 1989;140:157-160.
94. Jones JS, Newman RB, Miller MC. Cross-sectional analysis of triplet birth
weight. Am J Obstet Gynecol 1991;164:135-140.
95. Dubois S, Dougherty C, Duquette MP, et al. Twin pregnancy: the impact of
the Higgins Nutrition Intervention Program on maternal and neonatal outcomes.
Am J Clin Nutr 1991;53:1397-1403.
96. Powers WF. Twin pregnancy, complications and treatment. Obstet Gynecol
1973;42:795-808.
97. Reisner SH, Forbes AE, Cornblath M. The smaller of twins and
hypoglycaemia. Lancet 1965;144:524-526.
98. Babson SG, Phillips DS. Growth and development of twins dissimilar in size
at birth. N Engl J Med 1973;289:937-940.
99. Van Steirteghem A, Bonduelle, M, Liebaers I, et al. Children born after

assisted reproductive technology. Am J Perinatol 2002;19: 59-65.
P.533
100. Benirschke K, Temple WW, Bloor CM. Conjoined twins: nosology and
congenital malformations. Birth Defects Orig Artic Ser 1978;16:179-192.
101. Rudolph AJ, Michaels JP, Nichols BL. Obstetric management of conjoined
twins. Birth Defects Orig Artic Ser 1967;3:28-.
102. Schmidt W, Heberling D, Kubli F. Antepartum ultrasonographic diagnosis of

conjoined twins in early pregnancy. Am J Obstet Gynecol 1981;139:961-963.
103. Edmonds LD, Layde PM. Conjoined twins in the United States, 1970-1977.
Teratology 1985;25:30-308.
104. Filler RM. Conjoined twins and their separation. Semin Perinatol
1986;10:82-91.
105. de la Torre Verduzco R, Rosario R, Rigatto H. Hyaline membrane disease in

twins. A 7 year review with a study on zygosity. Am J Obstet Gynecol
1976;125:668-671.
106. Powell RW, Dyess DL, Luterman A, et al. Necrotizing enterocolitis in

multiple-birth infants. J Pediatr Surg 1990;25:319-321.
107. Samm M, Curtis-Cohen M, Keller M, et al. Necrotizing enterocolitis in

infants of multiple gestation. Am J Dis Child 1986;140: 937-939.
108. Pass MA, Khare S, Dillon HC Jr. Twin pregnancies: incidence of group B
streptococcal colonization and disease. J Pediatr 1980; 97:635-637.
109. Schuchat A, Oxtoby M, Cochi S, et al. Population-based risk factors for

neonatal group B streptococcal disease: results of a cohort study in
metropolitan Atlanta. J Infect Dis 1990;162: 672-677.
110. Schuchat A, Deaver-Robinson K, Plikaytis BD, et al. Multistate case-control

study of maternal risk factors for neonatal group B streptococcal disease.
Pediatr Infect Dis J 1994;13:623-629.
111. Benirschke K. Routes and types of infection in the fetus and the newborn.
Am J Dis Child 1960;99:714-721.
112. Mascola L, Ewert DP, Eller A. Listeriosis: a previously unreported medical

complication in women with multiple gestations. Am J Obstet Gynecol
1994;170:1328-1332.
113. Simoes EA, King SJ, Lehr MV, et al. Preterm twins and triplets. A high-risk
group for severe respiratory syncytial virus infection. Am J Dis Child
1993;147:303-306.
114. Beal S. Sudden infant death syndrome in twins. Pediatrics 1989;84:10381044.
115. Goshen-Gottstein ER. The mothering of twins, triplets and quadruplets.

Psychiatry 1980;43:189-204.
116. Tanimura M, Matsui I, Kobayashi N. Child abuse of one of a pair of twins in

Japan. Lancet 1990;336:1298-1299.
117. Groothuis JR, Altemeier WA, Rubarge JP, et al. Increased child abuse in
families with twins. Pediatrics 1982;70:769-773.
118. Minde K, Corter C, Goldberg S, et al. Maternal preference between

premature twins up to age four. J Am Acad Child Adolesc Psychiatry
1990;29:367-374.
119. Record RG, McKeown T, Edwards JH. An investigation of the difference in

measured intelligence between twins and single births. Ann Hum Genet
1970;34:11-20.
120. Durkin MV, Kaveggia EG, Pendleton E, et al. Analysis of etiologic factors in
cerebral palsy with severe mental retardation. I. Analysis of gestational,
parturitional and neonatal data. Eur J Pediatr 1976;123:67-81.
121. Kragt H, Huisjes HJ, Touwen BC. Neurobiological morbidity in newborn

twins. Eur J Obstet Gynecol Reprod Biol 1985;19: 75-79.
122. Petterson B, Nelson KB, Watson L, et. al. Twins, triplets and cerebral palsy
in births in Western Australia in the 1980s. BMJ 1993;307:1239-1243.
123. Silva PA, Crosado B. The growth and development of twins compared with
singletons at ages 9 and 11. Aust Paediatr J 1985;21: 265-267.
124. Morley R, Cole TJ, Powell R, et al. Growth and development in premature
twins. Arch Dis Child 1989;64:1042-1045.
125. Jefferies CA, Hofman PL, Knoblauch H, et. al. Insulin resistance in healthy
prepubertal twins. J Pediatr 2004;144:608-613.
126. Cameron AH. The Birmingham twin survey. Proc R Soc Med 1968;61:229234.
127. Newton W, Keith L, Keith D. The Northwestern University multihospital

twin study. IV. Duration of gestation according to fetal sex. Am J Obstet
Gynecol 1984;149:655-658.
128. Sassoon DA, Castro LC, Davis JL, et al. Perinatal outcome in triplet versus
twin gestations. Obstet Gynecol 1990;75:817-820.
129. Collins MS, Bleyl JA. Seventy-one quadruplet pregnancies: management

and outcome. Am J Obstet Gynecol 1990;162:1384- 1391
Chapter 28
Control of Breathing: Development, Apnea of
Prematurity, Apparent Life-Threatening Events,
Sudden Infant Death Syndrome
Mark W. Thompson
Carl E. Hunt
The fetus makes breathing movements in utero and these movements change in character and
frequency throughout gestation (1). This maturation of breathing control is likely critical to
initiation of spontaneous, regular respiratory effort at birth and normal control of postnatal
breathing. Studies in animals and preterm infants have explored the complexities of this
maturational process, but an exact timeline has yet to be delineated. In this chapter, normal
control of breathing will first be reviewed, the end-state of this maturational process will be
described, genetic influences on control of breathing will be summarized, and then what is known
about fetal breathing, the effects of sleep on fetal breathing, and the factors involved with
establishing normal patterns of breathing at birth will be reviewed. Finally, we will review apnea
of prematurity (AOP), how this disorder may shed light on the normal maturational process of
breathing control, how to treat AOP, and the potential adverse clinical consequences.
NORMAL CONTROL OF BREATHING

The control and maintenance of normal breathing largely reside within the respiratory control
centers of the bulbopontine region of the brainstem. Neurons within this area of the brain
respond to multiple afferent inputs to modulate their own inherent rhythmicity and provide
efferent output to the respiratory control muscles. Multiple afferent inputs induce modulation of
the central respiratory center efferent outputs to the respiratory and airway muscles and lungs.
These afferent inputs are categorized by the respiratory control center; some inputs cause an
instantaneous response in the control center output, while others only act to shape the
respiratory response, resulting in small changes in muscular output, tidal volume, and airway
tone (2). Among these inputs are signals from central and peripheral chemoreceptors, pulmonary
stretch receptors, and cortical and reticuloactivating system neurons. These afferent inputs and
resultant efferent outputs of the central respiratory center are summarized in Fig. 28-1. Sleep
state can also have a profound effect on respiratory pattern.
In the adult, these multiple afferent inputs act upon the neurons within the respiratory control
center and provide a well-integrated response to perturbations in the system and characteristic
respiratory patterns. For example, increased respiratory rate and depth will characteristically
result from activation of central chemoreceptors in response to hypercarbia. In the newborn
infant and especially in the preterm infant, however, these responses are less well-organized and
apnea is a common result of a disorganized response to multiple nonintegrated afferent inputs.
Figure 28-1 Major factors influencing respiratory control. PCO2, carbon dioxide partial pressure;
PO2, oxygen pressure. (Reprinted from Martin RJ, Miller MJ, Carlo WA. Pathogenesis of apnea in
preterm infants. J Pediatr 1986;109:733-741, with permission.)
P.536
Genetic Influences
Sequencing the approximately 25,000 genes in the human genome has resulted in fundamental
changes in our understanding of the role of specific gene products in the pathophysiology of all
diseases. Partly related to studies of sudden infant death syndrome (SIDS) pathophysiology,
recent genetic studies of cardiorespiratory regulation linking specific genotypes to specific
components of brainstem autonomic regulation are also relevant to understanding control of
breathing and its developmental regulation in general and to understanding the pathophysiology
of AOP and apparent life-threatening events (ALTE) in particular (3).
Neural control of breathing and sleep are closely integrated, and abnormalities in regulation of
sleep and circadian rhythmicity can impair cardiorespiratory integration and arousal
responsiveness from sleep (4). Circadian rhythmicity has been extensively studied in animals,
and homologous counterparts of essential circadian clock genes isolated in Drosophila have been
identified in mammals (3). Since the sleep-wake cycle is under control of the circadian clock,
these circadian master genes, as well as other sleep-related genes, likely influence sleep
regulation.
Targeted gene inactivation studies in animals have identified several genes involved with prenatal
brainstem development of respiratory control including arousal responsiveness (3). During
embryogenesis, the survival of specific cellular populations composing the respiratory neuronal
network is regulated by neurotrophins, a multigene family of growth factors and receptors. Brainderived neurotrophic factor (BDNF) is required for development of normal breathing behavior in
mice, and newborn mice lacking functional BDNF exhibit ventilatory depression associated with
apparent loss of peripheral chemoafferent input (5). Ventilation is depressed and hypoxic
ventilatory drive is deficient or absent.
Krox-20, a homeobox gene important for mouse hindbrain morphogenesis, also appears to be
required for normal development of the respiratory central pattern generator (6). Krox-20-null
mutants exhibit an abnormally slow respiratory rhythm and increased incidence of respiratory
pauses, and this respiratory depression can be further modulated by endogenous enkephalins.
Inactivation of Krox-20 may result in the absence of a rhythm-promoting reticular neuron group
localized in the caudal pons and could thus be a cause of life-threatening apnea (3).
Brainstem muscarinic cholinergic pathways are important in ventilatory responsiveness to carbon
dioxide (CO2) (3). The muscarinic system develops from the neural crest, and the ret protooncogene is important for this development. Ret knockout mice have a depressed ventilatory
response to hypercarbia, implicating absence of the ret gene as a cause of impaired hypercarbic
responsiveness. Diminished ventilatory responsiveness to hypercarbia has also been
demonstrated in male newborn mice heterozygous for Mash-1 (7). There is a molecular link
between ret and Mash-1, and the latter is expressed in embryonic neurons in vagal neural crest
derivatives and in brainstem locus coeruleus neurons, an area involved with arousal
responsiveness.
Serotonergic receptors in the brainstem are critical components of respiratory drive, and
abnormalities have been implicated as causal in SIDS (8). Review of the genetics of serotonin
metabolism may be very pertinent to understanding important genetic risk factors for SIDS, and
serotonin metabolism may be a model for multiple other genetically derived neurotransmitter
systems important in maturation and development of control of breathing. Serotonin is a
widespread neurotransmitter affecting breathing, cardiovascular control, temperature, and mood
(3,9). Serotonin modulates activity of the circadian clock and appears to be the major
neurotransmitter of nonrapid eye movement (nonREM), or quiet, sleep. Many genes are involved
in the control of serotonin synthesis, storage, membrane uptake, and metabolism. The serotonin
transporter gene is located on chromosome 17, and variations in the promoter region of the gene
appear to have a role in serotonin membrane uptake and regulation (3,9). Several transporter
polymorphisms have been described, including one in the promoter region; the long L allele
increases effectiveness of the promoter and hence leads to reduced serotonin concentrations at
nerve endings compared to the short S allele. The L/L genotype is associated with increased
serotonin transporters on neuroimaging, and increased postmortem binding. Although no studies
have been done in AOP or ALTE subjects, recent data in white, African-American, and Japanese
infants indicate that SIDS victims are more likely than controls to have the L allele (9,10).
Fetal Breathing
Fetal sheep make regular breathing movements during REM sleep (11). Studies in animal models
and human fetuses have demonstrated that there is a maturational progression in fetal breathing
movements over gestation and that this respiratory pattern can be altered by a variety of
pharmacologic and physiologic inputs.
In the human fetus, ultrasound evidence of fetal breathing movements was first noted with
development of high-speed ultrasound in the mid-1970s. These fetal breathing movements have
been characterized throughout gestation
P.537
in response to maternal condition and as a potential indicator of overall fetal condition (12). From
these studies, several characteristics of fetal breathing are evident. First, regardless of gestation,
fetal breathing is not a continuous process. Significant periods of apnea, up to 122 minutes, are
seen even in near-term fetuses (13). These periods of apnea are generally more frequent and of
longer duration in younger gestational age fetuses (14), but significant respiratory pauses are
still seen in presumed healthy near-term fetuses. Second, during periods of frequent respiratory
motion, there are patterns of both regular and irregular breathing documented by chest/
abdominal wall movements and by Doppler sonography assessing tracheal fluid flow (15). Third,
there does appear to be a circadian rhythmicity to fetal breathing movements, with welldocumented increases during certain periods of the day as compared to others (13). Fourth,
maternal condition, particularly maternal glucose status, can have significant effects on fetal
breathing frequency, with well-documented increases in fetal breathing frequency after maternal
glucose loading following maternal fasting. This response to maternal glucose loading is not as
pronounced when the mother is not fasting but rather ingesting normal meals during continuous
24-hour ultrasound monitoring (16). Fifth, fetal breathing does appear to be a potential measure
of fetal well-being when used in conjunction with other ultrasound parameters. Several studies
have documented diminished fetal breathing activity in association with poor fetal health, and
this decreased activity, along with other measures of fetal well-being, can be helpful in guiding
obstetric actions in terms of need for urgent delivery (17,18).
The fetal breathing pattern responds to a variety of pharmacologic and physiologic manipulations.
Fetal breathing has been noted to increase three-fold if the mother inhales 5% CO2 for 15
minutes (19). Maternal hyperoxia does not alter fetal breathing movements or pattern in nearterm normally grown fetuses (20), but growth-restricted fetuses do exhibit an increase in
respiratory rate with maternal hyperoxia (21). Tocolytics, indomethacin, and terbutaline increase
fetal breathing movements almost two-fold when administered to women in preterm labor
between 26 and 32 weeks of gestation (22). Despite the increased sensitivity of ultrasound to
detect and characterize fetal breathing movements, almost all of the studies described have been
observational in nature and have lacked characterization of several other very important fetal
physiologic correlates, such as diaphragmatic electrical activity and sleep state. These human
ultrasound studies have also not addressed the effects of pharmacologic and physiologic
manipulation on fetal breathing in a controlled fashion. Animal studies, however, do augment the
human ultrasound data and provide a clearer picture of the maturational development of
breathing control.
Fetal breathing activity in sheep starts early in gestation, arises from centrally mediated stimuli,
and occurs primarily during periods of low-voltage electrocortical activity (REM sleep) (11). REM
sleep comprises about 40% of fetal life during the last trimester in sheep. Breathing does occur
during periods of high-voltage electrocortical activity (quiet sleep), but there is no significant
pattern to this breathingit is only episodic in nature and generally occurs after tonic muscular
discharges, i.e., body movements (23).
Additional insights have been obtained using chronically instrumented fetal lambs with direct
visualization of fetal breathing and state at baseline and after physiologic manipulations (23).
Fetal sheep do not show any periods of wakefulness as indicated by eye opening and purposeful
movement of the head. The fetus alternates between REM sleep (low-voltage electrocortical
activity) and quiet sleep (high-voltage electrocortical activity). Breathing movements, swallowing,
licking, and other body movements occur almost exclusively during REM sleep. The fetus
responds to an increase in arterial carbon dioxide pressure (PaCO2) with an increase in breathing
as determined by increases in tracheal pressure, breathing frequency, and integrated
diaphragmatic activity, suggesting that both breathing rate and tidal volume increase in
response to increasing hypercapnia (24,25,26). This breathing activity only occurs in REM sleep
and during transitions to quiet sleep. Hypoxic levels of maternal oxygen abolish fetal breathing
movements (27). These observations suggest that the relative low oxygen levels in the fetus at
baseline contribute to the relative lack of fetal breathing movements throughout gestation and
that the several-fold rise in oxygen partial pressure at birth is a contributing factor to the
initiation of a regular respiratory pattern.
The use of pharmacologic agents in the fetal sheep model has also enhanced our understanding
of fetal breathing. Indomethacin, pilocarpine, 5-hydroxytryptophan, and morphine all cause
continuous breathing in both REM and quiet sleep. After these breakthroughs of breathing
activity into quiet sleep, the intensity eventually diminished even when the fetus transitions back
to REM sleep in association with decreasing drug levels (1,28).
In summary, animal experiments confirm that fetal breathing does occur. Under baseline
conditions and nonpharmacologic manipulation (i.e., hypercarbia, hypoxia, hyperoxia), it occurs
primarily during REM sleep (low-voltage electrocortical activity). The fetus does respond to
central chemoreceptor input as evidenced by its response to hypercapnia. The fetal response to
hypoxia appears to be centrally mediated and results in diminished normal neuronal activity and
diminished or absent fetal breathing movements. The peripheral chemoreceptors may also be
immature in the fetus and hence further contribute to absence of the adult response to hypoxia
of an increase in respiratory rate and tidal volume.
Initiation of Breathing at Birth

Despite the enhanced understanding of fetal maturation of breathing control, the understanding
remains incomplete regarding factors responsible for initiating and maintaining a regular pattern
at birth. The traditional view is that mild fetal asphyxia during labor stimulates peripheral
chemoreceptors leading to the first breath, and breathing is
P.538
then maintained via input of other sensory stimuli, e.g., cold and touch. However, this view has
been challenged by animal experiments showing that denervation of the carotid and aortic
chemoreceptors does not alter fetal breathing or the initiation/continuation of breathing at birth
(29). A stable respiratory pattern can also be initiated in utero with cord occlusion and hyperoxia,
independent of PaCO2. Postdelivery studies have also documented the importance of cooling in
establishing a regular respiratory pattern and of CO2 in maintaining that pattern (30). There are
some data to suggest that a placentally derived factor inhibits fetal breathing and removal of this
factor at birth then permits initiation of a regular respiratory pattern (1). This placental inhibitor
theory is intriguing, but it does not explain the totality of response needed to initiate and
maintain a normal respiratory pattern at birth; also, no candidate inhibitor has been suggested.
Development and maintenance of respiration in the newborn is more likely due to a complex
interaction of sensory stimuli and both central and peripheral chemoreceptor inputs (Fig. 28-1).
The degree of maturation in the central respiratory centers also appears to be important since
the responses to these stimuli in the term infant are more developed than in the preterm infant.
The Preterm Infant

The respiratory responses to hypercarbia are diminished in preterm infants less than 33 weeks'
postconceptional age (PCA) (31), unrelated to any mechanical limitations in achieving adequate
tidal volumes. The slope of the CO2 response curve (Fig. 28-2) increases significantly between 29
to 32 and 33 to 36 weeks of gestation. The cause of this decreased sensitivity of central
respiratory centers to CO2 in preterm infants is related to central nervous system immaturity as
indicated by decreased synaptic connections and incomplete dendritic arborization.
Preterm infants and full-term infants up to approximately 3 weeks postnatal age have a
characteristic response to hypoxia that is quite different than in older infants. In contrast to the
sustained hyperventilation in older infants, preterm infants have only a transient hyperventilation
lasting 30 seconds to a minute. Subsequently, minute ventilation returns to baseline over several
minutes and then progresses to ventilatory depression despite continued low inspired oxygen
concentrations (1). The initial hyperventilatory response may be completely blunted in extremely
premature infants (32). It is thought that this transient hyperventilation is in response to
peripheral chemoreceptor input that is subsequently overridden by depression of central
respiratory centers by hypoxia, and the hypoxic ventilatory depression is secondary to
descending inhibition from the upper brainstem, midbrain, or higher structures (33,34).
Figure 28-2 CO2 response curves for preterm infants with and without apnea of prematurity.
PaCO2, arterial carbon dioxide pressure. (Reprinted from Gerhardt T, Bancalari E. Apnea of
prematurity: I. Lung function and regulation of breathing. Pediatrics 1984; 74:58-62, with
permission.)
Upper airway reflexes also may play a role in inhibiting respiration, particularly in preterm
infants. Multiple sensory afferent fibers exist within the upper airways, and stimulation of these
fibers by various mechanisms can result in abnormal respiratory responses. Responses to upper
airway afferent fiber stimulation can change markedly with maturation. Newborn rabbits, for
example, can exhibit irreversible, fatal apnea with stimulation of the nares. This response does
not occur in mature animals (34). Negative pressure in the upper airways in human infants
results in depressed ventilation, and isolated application of negative pressure to the upper airway
in a rabbit model can result in diaphragmatic inhibition (35). This inhibition may contribute to the
centrally mediated apnea (central apnea) occurring in mixed apneas initiated by obstructed
breaths. As upper airway obstruction occurs, the infant makes respiratory efforts against this
obstruction and the resulting increased negative pressure in the upper airway may result in reflex
inhibition of diaphragmatic contraction. Due to a blunted response to hypercarbia and hypoxic
ventilatory depression, less mature preterm infants with apnea may be less able to recover
spontaneously and hence more likely to require active intervention by health care providers.
Laryngeal reflexes can also have significant effects on control of breathing. Apnea has been
associated with overt regurgitation of gastric contents into the upper airway (36). Either
regurgitation or reflux of the acidic contents of the stomach into the area of the larynx appears to
stimulate chemoreceptors in the receptor-rich area of the larynx. This leads to an inhibitory
afferent signal to the central respiratory centers. Animal studies have demonstrated that the area
of the larynx is rich in chemoreceptors connecting with afferent fibers leading back to the
brainstem. Stimulation of these chemoreceptors results in apnea in a lamb model (37,38).
Periodic Breathing
Periodic breathing is defined as a pattern of breathing alternating with respiratory pauses lasting
through three cycles of breathing, with the pauses or apneic periods lasting at least 3 seconds
(1,39). Periodic breathing is frequently seen in preterm infants and also in term neonates and
young infants.
Periodic breathing appears to occur predominantly during REM sleep, but it also occurs during
quiet sleep (32,40). During quiet sleep, periodic breathing is regular with consistent durations
of the apneic and breathing periods,

P.539
while during REM sleep periodic breathing tends to be irregular with inconsistent durations of
cycles of periodic breathing. These episodes of periodic breathing are reflective of immaturity of
the respiratory control system. While initially thought to be a benign respiratory pattern,
increasing evidence has indicated that periodic breathing can have pathologic consequences.
Particularly in more immature preterm infants, minute ventilation can diminish significantly
during episodes of periodic breathing, and oxygen saturations can decrease to hypoxemic levels
in association with longer apnea durations and increased time spent in periodic breathing. Since
more immature infants generally spend more time in sleep and more of this sleep time is
characterized by periodic breathing, these infants can spend a significant amount of time with low
oxygen saturations. These extreme degrees of periodic breathing likely reflect a greater general
immaturity of the central respiratory control centers; the more severe the periodic breathing,
the more likely an infant is to demonstrate overt clinical symptoms related to AOP.
It is very likely that both periodic breathing and AOP arise from the same pathophysiology, i.e.,
inability of the central respiratory control centers to adequately modulate multiple afferent inputs
and resulting inhibitory outputs to the muscles of respiratory control. AOP may simply reflect a
more severe variant of this dysmodulation. Not all studies have linked periodic breathing with
more prolonged apneas (41), but preterm infants that demonstrate frequent intervals of periodic
breathing also demonstrate an increased incidence of AOP and hence a general immaturity of the
respiratory control system. While periodic breathing can thus be a normal respiratory pattern in
both term and preterm infants, the higher the percent sleep time spent in periodic breathing and
the greater the length of apnea during apneic periods, the more likely it is that these intervals
represent a clinically significant immaturity of the central respiratory system associated with
pathologic disturbances in central respiratory control.
The prevalence of periodic breathing is not precisely known since most studies have been based
on convenience samples not necessarily representative of asymptomatic as well as symptomatic
infants. Differing definitions for periodic breathing are an additional concern. In preterm infants
less than 1,000 grams, many authors have reported the prevalence to be 100% (1,41). In term
infants, the prevalence has been reported to be lower, but still as high as 80% (42,43). The high
initial prevalence in both term and preterm neonates diminishes over time, most significantly in
the first 6 to 8 weeks of life.
CLINICAL DISORDERS IN CONTROL OF BREATHING

Apnea
Apnea is defined as the cessation of breathing for longer than 20 seconds, or a shorter duration
in the presence of pallor, cyanosis (See Color Plate), or bradycardia (44). There are multiple
possible etiologic factors leading to symptomatic apnea in the preterm and full-term infant
(Tables 28-1 and 28-2). The occurrence of apneic episodes thus requires clinical and laboratory
assessments to rule out conditions requiring definitive treatment. There are no systematic
prevalence data for symptomatic apnea in full-term infants, but most occurrences will have an
identifiable medical cause (Table 28-2). In preterm infants less than 1,500 grams at birth,
approximately 70% will have at least one clinically observed episode of symptomatic apnea while
in the neonatal intensive care unit (NICU), and about 20% of these infants will have a specific
medical cause (Table 28-1). The other 80% of preterm infants with symptomatic apnea do not
have a specific medical cause and by exclusion are then diagnosed as having AOP, the most
important and prevalent disorder of respiratory control occurring in preterm infants.
TABLE 28-1 ETIOLOGY OF APNEA IN PRETERM INFANTS

Cause
Comment
Idiopathic
Apnea of prematurity with immaturity of respiratory centers;

modified by sleep state
Central Nervous System
Intracranial hemorrhage, seizures, depressant drugs,

hypoxemia, hypothermia, hyperthermia
Respiratory
Pneumonia, obstructive airway lesions, Respiratory Distress

Syndrome, laryngeal reflex, phrenic or vocal cord paralysis,
pneumothorax, hypoxemia, hypercarbia, nasal occlusion
caused by phototherapy eye patches, tracheal occlusion
caused by neck flexion
Cardiovascular
Gastrointestinal
Infection
Metabolic
Heart failure, hypotension, hypertension, hypovolemia,

increased vagal tone
Gastroesophageal reflux, abdominal distension, peritonitis
Pneumonia, sepsis, meningitis
Acidosis, hypoglycemia, hypocalcemia, hyponatremia,
hypernatremia
Hematological
Anemia
Adapted from Hunt, CE. Apnea and sudden infant death syndrome in strategies in
pediatric diagnosis and therapy. Editors: RM Kliegman. W.B. Saunders Co. Philadelphia,
PA 1996.
Apnea of Prematurity
Pathophysiology
The maturation of cardiorespiratory control and the clinical course of premature infants with AOP
parallel each other and are consistent with the data reported in fetal animal studies. AOP is a
direct consequence of immaturity of brainstem respiratory control centers. Both the presence
P.540
and severity progressively increase the lower the gestational age. This autonomic immaturity is
not unique to respiratory control centers, however. Although there is no objective way to
clinically quantify degree of immaturity for other brainstem autonomic control systems,
brainstem auditory maturation can be quantified by serial measurements of brainstem conduction
time from the auditory-evoked response (wave VI interval) (45). Both improved synaptic
efficiency and myelination appear to be responsible for shortening of brainstem auditory
conduction times with advancing gestational age. The brainstem auditory nuclei are located in
close proximity to the cardiorespiratory centers, and there is a strong association between long
brainstem conduction times for the auditory-evoked responses and clinical episodes of AOP. Since
not all very-low-birth-weight preterm infants develop AOP and the severity varies even among
affected infants of the same gestational age, other genetic and/or environmental risk factors are
also likely to be important. Although no genetic, neurotransmitter, or neuropathologic studies
have been performed in preterm infants with AOP, performance of such studies may be as
informative in AOP as they appear to be in other control-of-breathing disorders.
TABLE 28-2 ETIOLOGY OF APNEA AT BIRTH IN FULL TERM INFANTS
Cause
Comment
Intrapartum asphyxia
Placental transfer of CNS-depressant
Hypoxemia, acidosis, brain stem depression

Narcotics, magnesium sulfate, general
Airway obstruction
anesthetics
Choanal atresia, macroglossia-mandibular
hypoplasia (Pierre-Robin Sequence),
tracheal web or stenosis, airway mass
lesions
Neuromuscular disorders
Absent or uncoordinated sucking/

swallowing, uncoordinated sucking and
Trauma
breathing, congenital myopathies or

neuropathies
Intracranial hemorrhage, spinal cord
Infection
transaction, phrenic nerve palsy

Pneumonia, sepsis, meningitis
Seizures, congenital central hypoventilation
syndrome, Arnold-Chiari malformation,
Dandy-Walker malformation
Adapted from Hunt, CE. Apnea and sudden infant death syndrome in strategies in
pediatric diagnosis and therapy. Editors: RM Kliegman. W.B. Saunders Co. Philadelphia,
PA 1996.
Apneic episodes in AOP are subclassified as central, obstructive, or mixed (34,46). Figure 28-3
demonstrates the respiratory patterns during these events. Central apneas result from lack of
respiratory effort. Obstructive apneas (obstructed breaths) are also central in origin but are
related to absence of neuromuscular control of upper airway patency rather than absence of
inspiratory diaphragmatic stimulation. Hence, obstructive apneas are characterized
P.541
by cessation of inspiratory air flow into the lungs despite persisting respiratory effort. Mixed
apneas represent a combination of missed breaths (central apnea) and obstructed breaths. Mixed
episodes of AOP can begin either with obstructed breaths or with central apnea, and there can be
multiple alternations between obstructed breaths and central apnea within any single episode. In
terms of relative frequency of each type, mixed apnea generally accounts for 53% to 71% of
events, with obstructive apnea 12% to 20% and central apnea 10% to 25% (34). Preterm infants
with AOP have significantly lower ventilatory responses to CO2 than preterm infants matched for
gestational age without AOP (31) (Fig. 28-2).
Figure 28-3 Examples of mixed, obstructive, and central apnea episodes occurring in apnea of
prematurity. A: Mixed apnea. Obstructed breaths precede and follow a central respiratory pause.
B: Obstructive apnea. Breathing efforts continue, although no nasal airflow occurs. C: Central
apnea. Both nasal airflow and breathing efforts are absent. BPM, beats per minute. (Reprinted
from Miller MJ, Martin RJ, Carlo WA. Diagnostic methods and clinical disorders in children. In
Edelman NH, Santiago TV, eds. Breathing disorders of sleep. New York: Churchill Livingstone,
1986:157-180, with permission.)
Incidence and Diagnosis

Threshold criteria for diagnosis have not been objectively defined and there is no established
diagnostic test. Reported incidence data are therefore not based on any standardized criteria for
diagnosis, but range from a low of less than 10% in preterm infants born at 34 weeks of
gestation or more to approximately 60% at birth weights less than 1,500 grams and to a high of
greater than 85% among infants born at less than 28 weeks of gestation (44,47,48). The first
episodes can manifest themselves within the first 24 hours of life in spontaneously breathing
infants without respiratory distress but are typically delayed in infants still receiving mechanical
ventilatory support (49).
The identification of individual episodes of AOP is routinely based on detection of: a) central
apnea lasting at least 20 seconds using standard impedance-based cardiorespiratory monitoring,
b) bradycardia of less than 80 to 100 beats per minute (bpm) depending on gestational age, or c)
arterial oxygen desaturation using a pulse oximeter (SpO2). These symptoms are typically
associated with clinically observed cyanosis. To be certain that monitor alarms are clinically
significant, episodes initially detected by a monitor alarm need to be confirmed by visual
observation (34,50). Since impedance-based cardiorespiratory monitors cannot detect
obstructive apneas unless and until associated with bradycardia, however, episodes of AOP will
not be detected by the monitor until central apnea has persisted long enough to trigger a monitor
alarm or until the bradycardia threshold has been reached. Isolated respiratory pauses of less
than 20 seconds without associated bradycardia or desaturation may not be of clinical
significance and are not sufficient to justify a diagnosis of AOP. Periodic breathing is typically
prominent in infants with AOP but is not sufficient for diagnosis in the absence of associated
bradycardia and/or desaturation.
In infants with AOP, the periodic respiratory pauses associated with periodic breathing may be
longer than in infants without AOP, and the total time spent in periodic breathing may be longer.
Both of these factors contribute to the occurrence and severity of associated bradycardia and/or
desaturation (39). Visually confirmed episodes of apnea with associated heart rate decelerations
and desaturation are classified as symptomatic episodes. There is no consensus-based threshold
number of clinical episodes required for a diagnosis of AOP, but the minimum is at least one
visually confirmed symptomatic episode unrelated to another medical cause (Table 28-1). The
severity of AOP in individual patients and the threshold for initiating treatment are determined by
the frequency and severity of subsequent clinically observed episodes. Although a multihour and
multichannel recording that includes monitoring for both central and obstructive apneas,
bradycardia, and desaturation might provide more objective threshold criteria for diagnosis and
treatment, such recordings are not part of routine clinical care and no criteria have been
established.
Treatment
Once other medical causes of apnea have been excluded (Table 28-1) and a diagnosis of AOP
established, multiple treatment strategies are available. The first step is to ensure that borderline
levels of baseline hypoxemia, anemia, hypocalcemia, and hypoglycemia have been corrected.
Infants with intermittent hypoxemia due to episodes of AOP will be at increased risk for reduced
tissue oxygen delivery (if they are also anemic) and for further exacerbation of AOP symptoms.
Blood transfusions can be effective in reducing the severity of episodes (52,53), but some studies
have not shown any correlation between degree of hypoxemia and degree of anemia (54,55 and
56). Some studies suggest that transfusion may be helpful in ameliorating symptoms of AOP only
if the baseline hematocrit is less than 25 %. Clinicians should address the following questions
when considering transfusion to improve symptoms related to AOP (57):
Are the symptoms related to AOP new in origin?
Has the severity of AOP increased as anemia has worsened?
Is the child receiving oxygen or other ventilatory support?
Has the heart rate increased?
Are other respiratory rhythm changes, such as periodic breathing, more
prominent?
Affirmative answers to these questions increase the likelihood that a transfusion may improve
clinical symptoms attributed to AOP.
The criteria for pharmacologic or ventilatory support vary among neonatologists, and there are
no established clinical guidelines. Treatment is indicated, however, whenever clinical episodes are
recurring, do not resolve spontaneously or in response to minimal stimulation, and are associated
with bradycardia and intermittent hypoxemia. The first line of therapy is generally a
methylxanthine, either caffeine or aminophylline.
Methylxanthines
Methylxanthines are central respiratory stimulants that increase CO2 sensitivity and hence lead to
improved tidal
P.542
and minute volumes and blood gas values. Methylxanthines also increase diaphragmatic function,
decrease muscle fatigue, and increase metabolic rate and catecholamine activity. These
medications can be given by either oral or intravenous route. Both aminophylline and caffeine
have been studied in placebo-controlled trials, and both have been shown to reduce the incidence
of apnea (44,58,59 and 60). Table 28-3 reviews the recommended starting doses and dosing
intervals for caffeine and aminophylline. The recommended therapeutic blood levels are 5 to 10
g/mL for aminophylline and 8 to 20 g/mL for caffeine. Methylxan-thine side effects are
secondary to catecholamine stimulation and include tachycardia, jitteriness, and irritability.
Caffeine has a wider therapeutic index and side effects are uncommon at the recommended blood
levels. There is consequently less need to check frequent blood levels. Direct comparisons of
aminophylline and caffeine indicate similar efficacy at equivalent levels of central ventilatory
stimulation. At equivalent degrees of central chemostimulation, however, side effects are less
with caffeine than with aminophylline (61,62,63). Many neonatologists thus recommend caffeine
as the preferred methylxanthine for treatment of AOP.
TABLE 28-3 DOSING REGIMENS FOR INITIATING METHYLXANTHINE THERAPY (IV/po)

FOR TREATMENT OF SYMPTOMATIC EPISODES IN PRETERM INFANTS WITH APNEA OF
PREMATURITY
Drug
Loading
Maintenance
4 to 6 mg/kg
1 to 3 mg/kg every 8 to 12
hours
10 to 20 mg/kg
2.5 to 4.0 mg/kg per day

(base)
Caffeine Citratec
20 to 40 mg/kg
5.0 to 8.0 mg/kg per day
Aminophylline
Caffeine
Start maintenance dose 8 to 12 hours after loading dose. If changing from IV to po

aminophylline, increase dose by 20%. If changing to po theophylline, no change in dose
is required.
b
Start maintenance dose 24 hours after loading dose. Caffeine citrate may be given IV
or po.
IV, intravenous; po, oral.
Regardless of which methylxanthine is chosen, most infants will have a significant reduction in
central apnea events and hence in the severity of episodes. While individual studies measure
efficacy in different ways, the number of clinical episodes was generally reduced by 50% to 90%
(58,59 and 60,64,65,66 and 67). A Cochrane collaborative review found that both caffeine and
aminophylline reduce the number of episodes of apnea and decrease the need for intermittent
positive pressure ventilation (68).
The necessary duration of treatment with a methylxanthine is highly variable. AOP improves as
brainstem respiratory control centers progressively mature, but there is considerable individual
variation in the PCA at which maturation is sufficient to eliminate clinically documented episodes.
By 32 to 36 weeks' PCA, most infants with AOP can be weaned from treatment. Since there is no
objective threshold to define resolution of AOP, neonatologists will usually use a combination of
PCA, length of time since the last documented episode, and general clinical status to decide when
to stop treatment. Once therapy is stopped, most physicians will continue to monitor an infant for
a variable period of time (3 to 8 days) during drug washout before concluding that the infant is
ready for discharge. Infants with recurrent clinical symptoms of bradycardia or cyanosis related
to AOP will require reinstitution of medical treatment.
Continuous Positive Airway Pressure

Continuous positive airway pressure (CPAP), usually administered using nasal prongs (NCPAP), is
also an effective treatment for AOP and is typically used when clinically significant episodes
persist despite optimal methylxanthine therapy. At CPAP levels of 2- to 5-cm water, infants with
AOP will have fewer episodes (69). This reduction is primarily related to significant reductions in
episodes of obstructive and mixed apneas (Fig. 28-4) and has been attributed to splinting open of
the upper airways by the positive airway pressure. CPAP levels of 2- to 5-cm water are generally
sufficient to reduce upper airway resistance in sleeping premature infants and eliminate clinical
symptoms of AOP related to mixed and obstructive apneas.
If NCPAP, in conjunction with a methylxanthine, is not sufficient to prevent recurring episodes
associated with bradycardia and intermittent hypoxemia, some investigators have suggested
using nasal intermittent positive pressure ventilation (70,71,72). Combined NCPAP and
intermittent positive pressure ventilation may most likely be needed when episodes are related
predominately to central apneas that are not being effectively eliminated by the NCPAP.
Figure 28-4 Effect of continuous positive airway pressure (CPAP) on a number of apneic
episodes 10 seconds or longer in 10 infants with apnea of prematurity. Mixed and obstructive
apnea decreased significantly during both periods of CPAP, with no evident effect on central
apneas. (Reprinted from Miller MJ, Carlo WA, Martin RJ. Continuous positive airway pressure
selectively reduces obstructive apnea in preterm infants. J Pediatr 1985;106: 91-94, with
permission.)
P.543
Doxapram
Doxapram is an intravenous analeptic agent used in central hypoventilation syndrome and acute
respiratory failure, and as a postoperative respiratory stimulant (44,73). Treatment with
doxapram should be considered whenever methylxanthine therapy and NCPAP have been
optimized but significant bradycardia and/or hypoxemia episodes persist. Its mechanism of action
appears to be related to stimulation of carotid chemoreceptors at lower doses and to direct
stimulation of central respiratory control neurons at higher doses. It is effective in diminishing the
number of clinical episodes in small clinical trials and may be most effective when used in low
doses in combination with low-dose methylxanthine treatment (74,75). The intravenous loading
dose for doxapram is 2.5 to 3.0 mg/kg over 15 minutes followed by a continuous infusion of 1.0
mg/kg titrated to the lowest responsive dose (maximum dose 2.5 mg/kg/hour). At these low
doses of doxapram used in AOP, significant side effects are uncommon but increased blood
pressure and other side effects of catecholamine stimulation can occur, including lowering seizure
threshold.
Intubation and Assisted Ventilation

If an infant is still having clinically significant episodes of bradycardia and/or cyanosis (See Color
Plate) secondary to AOP despite optimal pharmacologic and noninvasive positive airway pressure,
then endotracheal intubation and positive pressure ventilation is the treatment of last resort. It is
infrequent that such invasive treatment is necessary, but no data as to frequency of need are
available. Since intubation for treatment of AOP is uncommon, infants with AOP requiring
intubation also need to be evaluated for intercurrent illnesses such as sepsis/meningitis and
seizures.
Gastroesophageal Reflux
Significant controversy exists as to whether gastroesophageal reflux (GER) makes a significant
contribution to episodes of AOP. Multiple studies indicate that stimulation of upper airway and
especially laryngeal receptors leads to inhibitory influences on central respiratory centers and
subsequent apnea. However, the preponderance of data from human studies indicates no causal
link between GER and AOP (76,77 and 78). Some studies using intraluminal impedance
techniques have shown a positive correlation between GER and AOP (79), but other studies using
the same technique have not shown this correlation (78). GER and AOP may thus coexist in
preterm infants and both may require treatment; however, in the absence of specific symptoms
of GER, there is no evidence that treatment for GER will ameliorate symptoms of AOP.
Feeding-Related Symptoms
Both the establishment of adequate oral intake and resolution of any apneic or hypoxemic
episodes related to persisting AOP are developmentally regulated. Just as resolution of AOP can
be tracked by measuring maturation of brainstem auditory conduction using brainstem auditoryevoked responses, so can the attainment of coordinated oral feeding (45). Many infants will
resolve their AOP and begin successful oral feeding at approximately the same time
developmentally, but it is also possible that oral feedings may further aggravate episodes of AOP
if the brainstem areas related to ventilatory control and oropharyngeal coordination are not
sufficiently mature.
Feeding is a complex motor task that involves three coordinated steps: sucking, swallowing, and
breathing. While full-term infants will breathe while sucking, the preterm infant has greater
difficulty with this motor task. In addition, the preterm infant is very likely at the beginning of a
feeding session to spend an exaggerated period of time in the initial sucking stage and not breath
at all while in this stage. The full-term infant with no lung disease and a mature suck/swallow
mechanism may have a brief period of mild desaturation during oral feeding, but rarely does this
lead to bradycardia and apnea. The preterm infant with borderline respiratory status and an
immature suck/swallow response, however, is more likely to hypoventilate while feeding due to
immature suck/ swallow pattern and response (80,81). If a preterm infant with AOP is
demonstrating desaturations with feeding that are leading to bradycardia or apnea, decreased
oral intake may be necessary pending further maturational development in oropharyngeal and
ventilatory control as related to AOP. In addition to continuing methylxanthine treatment,
additional therapeutic intervention may be as simple as providing oxygen immediately prior to
and during the feeding (80).
Natural History: Implications for Discharge Planning and

Home Care
Most preterm infants with AOP progressively improve with increasing PCA and are clinically
normal without treatment by 34 to 36 weeks' PCA. Even after apparent resolution of AOP
symptoms, however, symptoms can recur under well-defined conditions. Infection with
respiratory syncytial virus (RSV) places infants at risk for apneic episodes, and apnea can be the
presenting sign of an RSV infection (82,83). It is unclear why RSV places premature infants at
risk for recurrent episodes of apnea, but RSV may alter laryngeal chemoreceptor sensitivity. With
recovery from the RSV infection, the apnea episodes also resolve.
Exposure to general anesthesia also places premature infants at risk for recurrent apneic events
(84). This is particularly true when inhalational anesthetic agents are used. Apnea-related
symptoms can be related to anemia, especially at hematocrit levels below 30% (85). Unrelated
to anemia, the incidence of apnea-related symptoms after discharge from the recovery room may
be as high as 5% at 48 to 50 weeks' PCA and does not decrease to less than 1% until 54 to 56
weeks' PCA (84). Most clinically significant episodes, however, occur at less than 44 weeks' PCA,
which is very consistent with Collaborative Home Infant Monitoring Evaluation (CHIME) data (46).
Premature
P.544
infants below these PCA limits, especially at lower hematocrit levels, thus need to receive
adequate monitoring postoperatively to prevent life-threatening events.
Persisting or recurring symptoms at home associated with a prior history of AOP in former
premature infants has become an issue of increasing clinical importance. At least in part related
to discharge at younger PCAs, a significant minority of AOP patients have persisting symptomatic
episodes related to apnea, bradycardia, cyanosis, or a combination thereof, when otherwise
ready for discharge. In the CHIME study, for example, 17% of 443 infants less than 1,750 grams
and 34 weeks or less of gestational age at birth were still symptomatic at the time of discharge
home and 29% of these 443 infants were discharged with a methylxanthine (46).
Numerous studies have attempted to establish the minimum number of symptom-free hospital
days without treatment required to ensure that symptoms would not recur after discharge home.
Based on retrospective chart reviews in otherwise healthy preterm infants 1,500 grams or less
and 32 weeks or less at birth and noting frequency of clinical episodes of apnea, bradycardia, and/
or color change related to AOP, infants may be symptom-free for up to 8 days and still have
another event (86). These investigators thus concluded that 8 symptom-free days were required
to ensure resolution of AOP. In a different study based on monitor alarms, preterm infants with
bedside monitor alarms had very infrequent apnea 20 seconds or more in duration when
approaching discharge but had a higher frequency of desaturation and bradycardia with short
apneas compared to preterm infants free of monitor alarms for at least 2 days (87). Among
preterm infants with birth weights less than 1,250 grams, not having any clinically apparent
episodes, and otherwise ready for discharge, 24-hour recordings of oxygen saturation, heart rate,
respiratory impedance, and end-tidal CO2 partial pressure revealed significant apnea in 91%, the
majority of which were obstructive (88). Significant apnea was defined as apnea of longer than
12 seconds in association with a heart rate decrease of at least 10% or a decrease of at least 10
percentage points in oxygen saturation. Since there was no apparent correlation between
severity of these predischarge recorded events and apparent life-threatening events or sudden
unexpected death in the first 6 months of age, however, these data do not clarify whether one
should delay discharge or prescribe home-based intervention for events detected only by
overnight recording and not evident by bedside observation. In the absence of a clear standard of
care in this regard, individual practice is highly variable. Some centers discharge preterm infants
after as few as 3 days or as long as 5 to 7 days or more without a clinical event. There are no
data to suggest that the relative risk of having an ALTE or dying of SIDS is any greater in infants
discharged after just 3 days versus 5 to 7 or more days or that the relative risk is greater in
infants in whom the event-free period is determined only by clinical observation versus overnight
recordings.
The CHIME provides new insights regarding how long AOP-related events can persist after
discharge home and the potential value of home-based memory monitor recordings in assessing
risk for subsequent life-threatening or fatal events (46). The CHIME study was performed with a
specially designed memory monitor utilizing respiratory inductance plethysmography to detect
obstructed breaths as well as central apneas. Healthy full-term infants, premature infants,
subsequent siblings of prior SIDS victims, and infants with a prior idiopathic ALTE were monitored
until 6 months' postnatal age. The events stored for analysis included events exceeding
conventional alarm thresholds (apnea of at least 20 seconds or heart rate less than 60 to 80 bpm
for at least 5 seconds, depending on PCA) and extreme events (apnea of at least 30 seconds or
heart rate less than 50 to 60 bpm for at least 5 seconds, depending on PCA).
Conventional events were common in all groups and at least one occurred in 41% of infants. Of
those conventional events with apnea of at least 20 seconds, 50% included three or more
obstructed breaths. Extreme events occurred in 10% of all infants. Among extreme events with
apnea of at least 30 seconds, 70% included three or more obstructed breaths. In general, the
degree of hypoxemia increased with increasing duration of apnea or bradycardia, and 25% of
extreme events were associated with a decrease in oxygen saturation (SpO2) of at least 10%.
Compared to healthy term infants, preterm infants were more likely to have at least one extreme
event, especially preterm infants younger than 34 weeks of gestation and with birth weights of
less than 1,750 grams. Among these preterm groups, the relative risk of at least one extreme
event was 18.0 in those who had persisting symptoms related to AOP within the last 5 days prior
to NICU discharge and was 10.1 in those who had no clinical AOP-related events for at least 5
days before discharge. In these two groups, the risk of at least one extreme event and at least
one conventional event remained higher than in the healthy term infants until approximately 43
weeks' PCA.
The CHIME study does not establish whether it is clinically important to detect conventional or
extreme cardiorespiratory events by home memory monitoring. It is not known, for example,
whether infants with a prior history of AOP are at greater risk for SIDS than gestational agematched preterm infants without such a history, and the CHIME study was not designed to
determine to what extent the presence and severity of extreme events in preterm infants after
discharge home may be contributing to the greater prevalence of SIDS in preterm infants.
Neurodevelopmental Outcome
Most reports have found little evidence of any neurodevelopmental risk directly attributed to a
history of AOP or ALTE (39,89). All of these AOP studies are limited due to nonstandardized
criteria for diagnosis and for quantifying severity of apnea-related events in the NICU, and due to
variable treatment strategies (89,90). Precisely measured predischarge apnea related to AOP,
however, has been
P.545
reported to be predictive of lower developmental indices at two years of age (91).
No outcome studies in preterm infants with AOP have considered the potential impact of
cardiorespiratory events occurring at home during early infancy even though persistence beyond
term gestation appears to be common, especially at lower birth weights (46,92). Using treatment
with doxapram as a surrogate for severe AOP in preterm infants with birth weights of less than
1,250 grams, the duration of doxapram treatment and total dose received are significantly
greater in children with isolated mental delay at 18 months compared to matched controls not
receiving doxapram (93).
The CHIME study also provides some insights regarding risk for neurodevelopmental sequelae in
infants with events documented using home memory monitoring (94). Among 256 infants (46%
preterm) who used the home monitor and returned at 92 weeks' PCA for performance of the
Bayley Scales of Infant Development-Revised (BSID-II), there was an inverse relationship
between number of conventional events detected by home cardiorespiratory monitoring and
neurodevelopmental outcome. This decrease in neurodevelopmental performance was observed
both in full-term and preterm infants but was limited to the Mental Development Index (MDI).
The adjusted difference in mean MDI scores with at least five events compared to no events was
5.6 points lower in full-term infants and 4.9 points lower in preterm infants. There was a trend
toward a lower adjusted difference in mean Psychomotor Development Index scores with at least
five events compared to no events in full-term infants (P = 0.07) but not in preterm infants. A
dose effect is suggested by the tendency for mean BSID-II values with 1 to 4 events to be
intermediate between 0 and at least 5 events. These findings are consistent with recent studies
of sleep-disordered breathing in children showing diminished school performance several years
later and in adults showing neuropsychologic deficits and gray matter loss by high-resolution
magnetic resonance imaging in multiple sites (95,96 and 97). These CHIME data do not clarify,
however, whether lower developmental performance at 1 year of age in preterm infants is caused
by the events documented at home or whether the events and developmental delay are both the
consequence of antecedent events in the NICU or antenatally. Since the findings in preterm
infants are similar to those observed in full-term infants, however, the developmental outcomes
appear at least in part attributable to the home-based events. It is unknown to what extent these
developmental delays in preterm infants can be reduced or eliminated by improvements in
hospital-based treatment of AOP or in home-based treatment of persisting events.
IDIOPATHIC APPARENT LIFE-THREATENING EVENTS

An ALTE is defined as a sudden, unexpected change in an infant that is frightening to the
caregiver but does not lead to death or persistent collapse (98). Most patients are younger than
12 months and generally younger than 6 months of age. Episodes are characterized by some
combination of apnea, color change, change in muscle tone, choking, and gagging (39).
Approximately 40% to 50% of ALTE cannot be explained by the history and clinical evaluation
and are hence classified as idiopathic. The percent of idiopathic cases is likely higher when
symptoms are sleep related. Incidence rates for an ALTE vary from 0.5 to 10.0 per 1,000 live
births; this 20-fold difference is likely related to different case definitions and methods of
ascertainment. Most studies of ALTE do not include information about birth weight and
gestational age, but the incidence in preterm infants has been estimated to be in the range of 8%
to 10% (88). In the CHIME study, 30% of 152 infants with an idiopathic ALTE were 37 weeks of
gestation or less at birth. Preterm infants thus are at increased risk for an idiopathic ALTE, but it
is unknown to what extent this might be related to a prior history of AOP or to persisting
symptoms of AOP.
Infants with an idiopathic ALTE are at increased risk for SIDS, but there is no consensus as to the
magnitude of this risk or the extent to which prematurity or AOP might be a related factor.
History of an idiopathic ALTE has been reported in 5% to 9% of SIDS victims (98,99). The risk of
SIDS appears to increase with two or more idiopathic events, but no definitive incidence rates are
available. In the CHIME study, the relative risk of having at least one extreme event was
increased in preterm infants with a history of an idiopathic ALTE, but only until about 43 weeks'
PCA (46).
Infants experiencing an idiopathic ALTE are candidates for home memory monitoring. In such
patients, memory monitors can be useful in determining risk for, and timely identification of,
recurrent life-threatening episodes, and perhaps useful in reducing the risk for
neurodevelopmental sequelae from subsequent events (94). There are no data to indicate,
however, that the risk for SIDS can be reduced by using a home monitor (100).
SUDDEN INFANT DEATH SYNDROME

SIDS is defined as the sudden death of an infant that is unexpected by history and unexplained
by a thorough postmortem examination, which includes a complete autopsy, investigation of the
scene of death, and review of the medical history (101). An autopsy is essential in all sudden and
unexpected infant deaths because the history and scene investigation do not exclude all causes,
including unsuspected congenital abnormalities and fatal child abuse. Comprising more than 7%
of all infant mortality in the United States, SIDS was the third leading cause of infant mortality in
2000, ranked below congenital anomalies (21%) and disorders relating to short gestation/low
birth weight (15%). About 2,650 infants died of SIDS in the United States in 1999, a rate of 0.67
in 1,000 live births; preliminary data suggest a lower rate for 2000 (102).
P.546
SIDS is the most common cause of postneonatal infant mortality in developed countries,
generally accounting for 40% to 50% of infant deaths between 1 month and 1 year of age and
about 20% of all deaths in infants discharged from a NICU. In full term infants, SIDS is rare
before 1 month of age; the peak incidence occurs at 2 to 4 months of age, and 95% of all cases
have occurred by 6 months of age. Compared to full-term infants, SIDS in preterm infants occurs
at a younger PCA but older postnatal age (103).
Pathology
There is no autopsy finding pathognomonic of SIDS and no finding required for the diagnosis.
There are, however, some common pathologic observations (104). Petechial hemorrhages are
found in more than 90% of cases and may be more extensive than in other causes of infant
mortality. Pulmonary edema is often present and may be substantial. Autopsy studies
demonstrate structural evidence (tissue markers) indicative of preexisting, chronic low-grade
asphyxia in nearly two-thirds of SIDS subjects (105). These tissue markers include persistence of
adrenal brown fat, hepatic erythropoiesis, brainstem gliosis, and increased substance P. SIDS
infants as a group have both prenatal and postnatal growth retardation, again consistent with
prenatal and postnatal low-grade asphyxia. Vascular endo-thelial growth factor, which is
upregulated by hypoxia, is elevated in cerebrospinal fluid of SIDS victims compared to control
cases, confirming that SIDS is preceded by hypoxia and suggesting that it is prolonged hypoxia
(106). The elevated levels of hypoxanthine in vitreous humor observed in SIDS infants further
indicate a relatively long period of tissue hypoxia preceding death (105,107). Since adenosine, a
precursor of hypoxanthine, is a respiratory inhibitor, these observations suggest that asphyxia
from any cause could cause a secondary acceleration of adenosine monophosphate catabolism
and adenosine accumulation and hence a vicious cycle leading to progressive hypoventilation and
worsening asphyxia.
Brainstem structure and function have been a major focus of postmortem studies in SIDS
victims. In addition to gliosis, findings include a persistent increase of dendritic spines and
delayed maturation of synapses in the medullary respiratory centers, decrease of tyrosine
hydroxylase immunoreactivity, and decreased catecholaminergic neurons in the brainstem (108).
Significant decreases in serotonin receptor immunoreactivity in vagal medullary centers
important in control of breathing have also been observed (109). Collectively, these results
suggest that a maturational delay or malfunction related to cardiorespiratory autonomic
regulation is one mechanism leading to SIDS.
The arcuate nucleus in the ventral medulla has been a particular focus for studies in SIDS
victims. It is an integrative site for vital autonomic functions, including breathing and arousal,
and is integrated with other regions that regulate arousal and autonomic chemosensory function.
Quantitative three-dimensional anatomic studies indicate that a small subset of SIDS victims
have hypoplasia of the arcuate nucleus, and histopathologic studies have observed bilateral,
partial, and monolateral hypoplasia in as many as 57% of SIDS victims (110,111).
Neurotransmitter studies of the arcuate nucleus have also identified receptor abnormalities in
some SIDS victims that involve several receptor types relevant to autonomic control overall and
to ventilatory and arousal responsiveness in particular. These deficits include significant
decreases in binding to kainate, muscarinic cholinergic, and serotonergic receptors (8). Protein
kinase C (PKC) and neuronal nitric oxide synthase (NOS) in the brainstem are critical components
of respiratory drive, and abnormalities have been implicated as causal in SIDS (112). Prenatal
cigarette smoke exposure is an important risk factor for SIDS, and decreased immunoreactivity
to selected PKC and NOS isoforms have been observed in rats exposed prenatally to cigarette
smoke.
Environmental Risk Factors

Numerous maternal and infant risk factors for SIDS have been identified (Table 28-4). Some of
these risk factors may be surrogates for more fundamental factors (99). Some risk factors that
are modifiable have been the successful focus of risk-reduction campaigns, resulting in
P.547
decreased SIDS rates (100). Preterm infants are approximately 3 to 6 times more likely to die of
SIDS than full-term infants (100,103,113,114). The environmental risk factors associated with
SIDS in preterm infants are not substantially different than those observed in full-term infants
(Table 28-5), and there is an inverse relationship between risk for SIDS and birth weight/
gestational age. The postnatal age of preterm infants dying of SIDS is 5 to 7 weeks later than
that of full-term infants, and the PCA is 4 to 6 weeks earlier than that for full-term infants (Table
28-6).
TABLE 28-4 ENVIRONMENTAL FACTORS ASSOCIATED WITH INCREASED RISK FOR SIDS
Maternal and antenatal risk factors

Smoking
Drug exposure (e.g., cocaine, heroin)
Nutritional deficiency
Inadequate prenatal care
Low socioeconomic status
Decreased age, education
Single marital status
Increased parity
Shorter interpregnancy interval
Intrauterine hypoxia
Fetal growth retardation
Infant risk factors
Age (peak 2-4 months)
Male gender
Race/ethnicity (e.g., African American and American Indian)
Growth failure
No pacifier (dummy)
Prematurity
Prone (and side) sleep position
Recent (febrile) illness
Smoking exposure (prenatal and postnatal)
Soft sleeping surface, soft bedding
Thermal stress/overheating
Colder season, noncentral heating
Reprinted from Hunt CE, Hauk F: Sudden infant death syndrome, in Nelson textbook of
pediatrics, 17th Edition, ed. by RE Behrman, RM Kliegman and HB Jenson, Philadelphia:
W.B. Saunders Company 2003:1380, with permission.
TABLE 28-5 EPIDEMIOLOGICAL FACTORS ASSOCIATED WITH RISK FOR SIDS IN
PRETERM INFANTS
Factor
Birth weight
500 to 999 g
Adjusted Odds Ratio
3.1
1000 to 1499 g
3.8
1500 to 2499 g
2.5
>2,499 g
Gestational age
1.0
17 to 28 wks
2.9
29 to 32 wks
2.8
33 to 36 wks
1.8
>36 wks
Race
Non-Black
Black
1.7
Gender
Male
1.5
Female
Maternal age
<18 y
18 to 35 y
>35 y
Maternal education
<12 y
12 y
Pregnancies
1
1.0
1.0
1.0
1.7
1.0
0.5
1.7
1.0
a
2 to 3
0.6
1.0
>3
1.1
Maternal smoking
No
Yes
1.0
2.8
Statistically significant difference compared to reference group

SIDS, sudden infant death syndrome. Adapted from Malloy MH, Freeman DH Jr. Birth
weight- and gestational age-specific sudden infant death syndrome mortality: United
States, 1991 versus 1995. Pediatrics 2000;105:1227-1231, with permission.
In response to international back-to-sleep campaigns introduced in numerous countries in the

1990s, SIDS rates have declined by 50% or more as the prevalence of placing infants prone for
sleep has decreased dramatically. These dramatic decreases in SIDS rates have generally not
been associated with significant decreases in other risk behaviors. The 40% reduction in SIDS
rates in the United States in 1992-1998 as prone prevalence decreased by approximately 50%,
however, was associated with a 25% decrease in maternal reporting of smoking during
pregnancy.
TABLE 28-6 POSTCONCEPTIONAL AND POSTNATAL AGES (WEEKS) OF SIDS VICTIMS
Gestational Age
Postconceptional Age
24 to 28
44
29 to 32
45
33 to 36
47
11
50
11
>36
a
Postnatal Age
18
16
P <0.05 compared to >36 weeks' gestational age
P <0.05 compared to the two older groups

Adapted from Malloy MH, Hoffman HI. Prematurity, SIDS, and age of death. Pediatrics
1995;96:464-471, with permission.
The odds ratio for SIDS in prone-sleeping infants in a metanalysis of case-control studies was 2.8
(95% confidence intervals [CI] 2.1-3.6) (99). The highest risk for SIDS may occur in infants
usually placed nonprone but placed prone for last sleep (unaccustomed prone) or found prone
(secondary prone). Unaccustomed prone is more likely to occur in day care or other settings
outside the home and highlights the importance of educating all infant caretakers about
appropriate sleep positioning. As prone sleeping rates have declined following SIDS risk-reduction
campaigns, side sleeping has also emerged as an independent risk factor for SIDS, with a relative
risk of 2.0 compared to supine sleeping. The American Academy of Pediatrics has recommended
since 1996 that supine sleeping be the preferred sleeping position for all infants (100), but they
did not explicitly include preterm infants in this recommendation. As prone- and side-sleeping
prevalence have decreased, unsafe sleeping practices, including soft sleep surfaces and soft
bedding, have emerged as significant risk factors (99). Bed sharing has been implicated as
another risk factor for SIDS, but studies have not classified risk for SIDS according to reason for
bed sharing and have only partially adjusted for other pertinent risk factors such as breastfeeding.
Despite initial concern that supine sleeping would increase the risk for adverse events such as
aspiration, vomiting, and trouble sleeping, studies have not identified any adverse consequences
of supine sleeping compared to prone sleeping (115). No symptom or illness was increased
among nonprone sleepers during the first 6 months of age, and some illnesses were less
common, especially ear infections. Aspiration has not been observed to occur more frequently in
infants sleeping supine, and infants sleeping supine do not appear to be at increased risk for
episodes of apnea or cyanosis (See Color Plate).
Preterm infants were initially excluded from back-to-sleep campaigns. This exclusion was based
on data from the NICU that ventilation and oxygenation, especially with persisting respiratory
symptoms, were better when positioned prone compared to supine. When preterm infants are
sufficiently mature and healthy to be discharged home without supplemental oxygen, however,
P.548
there is no clinically significant impairment in respiratory status when sleeping supine compared
to prone. The American Academy of Pediatrics has therefore recommended since 2000 that all
infants should sleep supine at home regardless of gestational age at birth (100), but adherence
with this recommendation remains limited. Infants with birth weights of less than 1,500 grams
are more likely to be placed to sleep in the prone position at home than infants with a birth weigh
of 1,500 to 2,499 grams (26% versus 14%) even though risk for SIDS is greater at birth weights
less than 1,500 grams compared to greater than 1,500 grams (113). Strategies are needed to
reassure health care providers that the supine position for sleeping is as safe and effective in
preterm as in full-term infants and to model safe sleeping practices in the NICU as the discharge
planning process begins.
Maternal smoking during pregnancy has consistently been associated with increased risk of SIDS
(99). The relative risk is in the range of 4.7 and represents one of the most significant modifiable
risk factors following declines in prone sleeping. There appears to be a small independent effect
of paternal smoking during pregnancy, but studies examining the influence of other household
members have been inconsistent. Infants dying from SIDS tend to have higher concentrations of
nicotine in their lungs than control infants, regardless of reported smoking exposure (116).
Elimination of prenatal exposure to cigarette smoke could theoretically reduce the risk of SIDS by
approximately 30% to 40% (117). It is difficult to assess the independent effect of postnatal
exposure to cigarette smoke because smoking exposure during and after pregnancy are highly
correlated, but studies do suggest that eliminating postnatal exposure to environmental cigarette
smoke might further reduce risk for SIDS. In some studies of bed sharing as a risk factor for
SIDS, this association has been linked with postnatal maternal smoking.
There are several potential mechanisms to explain why cigarette smoke exposure is a risk factor
for SIDS. Maternal smoking can potentiate hyperplasia of pulmonary neuroendocrine cells, and
dysfunction of these cells may contribute to the pathophysiology of SIDS (3). Both animal and
clinical studies indicate decreased ventilatory and arousal responsiveness to hypoxia following
fetal exposure to nicotine (118). The age-specific attenuation of hypoxic defenses following
nicotine exposure suggests impaired brain catecholamine metabolism. In vitro studies suggest
that smoking increases risk for SIDS due to greater susceptibility to viral and bacterial infections
and enhanced bacterial binding after passive coating of mucosal surfaces with smoke components
(119).
Most studies have not identified an association between SIDS and maternal alcohol use during or
following pregnancy after adjusting for cigarette smoke exposure (99). A recent study, however,
identified an increased frequency of binge drinking during pregnancy of 73% in mothers of SIDS
victims compared to 45% in control mothers (120).
Maternal drug use during pregnancy is a risk factor for SIDS, with a two-fold increased risk of
SIDS observed in the National Institute of Child Health and Human Development Cooperative
Epidemiological Study after adjusting for birth weight, race, and age (121). Another study
identified a seven-fold increased risk for SIDS among infants of substance-abusing mothers
compared with drug-free mothers (122). Relative risks vary from 3.1 (CI 0.43-21.74) for
phencyclidine and 6.9 (4.04-11.68) for cocaine to 15.1 (6.30-36.20) for opiates. The variable and
sometimes conflicting results appear related at least in part to failure to control for confounding
variables and sometimes to inadequate sample size.
No clear association has been identified between SIDS and specific viral or bacterial pathogens
(99). It has been suggested that upper respiratory infections or other minor illnesses in
conjunction with other factors, such as prone sleeping, may play a role in the pathogenesis of
SIDS. Deficient inflammatory responsiveness to infection has also been hypothesized to be a
mechanism for SIDS (3,119,123). Partial deletions in the C4 gene may contribute to this
apparent link between upper respiratory infection and SIDS. Mast cell degranulation has been
reported in SIDS victims; this is consistent with an anaphylactic reaction to a bacterial toxin, and
some family members of SIDS victims also have mast cell hyperreleasability and degranulation,
suggesting a genetic component to risk for an anaphylactic reaction (124).
Gene-Environment Interactions
The actual risk for SIDS in individual infants is determined by complex interactions between
genetic and environmental risk factors (3). There appears, for example, to be an interaction
between prone/side sleep position and impaired ventilatory and arousal responsiveness. Facedown or nearly face-down sleeping does occasionally occur in prone-sleeping infants and can
result in episodes of airway obstruction and asphyxia in healthy full-term infants (125). Healthy
infants will arouse before such episodes became life threatening, but infants with insufficient
arousal responsiveness to asphyxia would be at risk for sudden death. There may also be links
between modifiable risk factors such as soft bedding, prone-sleep position, and thermal stress,
and links between genetic risk factors such as ventilatory and arousal abnormalities and
temperature or metabolic regulation deficits. The increased risk for SIDS associated with fetal
and postnatal exposure to cigarette smoke also appears at least in part to depend on genetic risk
factors (3).
There are substantial data indicating that both genetic and environmental factors contribute to an
increased risk for death from most natural causes in siblings. The next-born siblings of first-born
infants dying of any noninfectious cause, for example, are at significantly increased risk for infant
death from the same cause, including complications of prematurity and SIDS (3). The relative
risk is 9.1 for concordance of cause of recurrent death versus 1.6 for a discordant cause of death,
and the relative risk for recurrence of each cause of infant death is similar for SIDS (5.4-5.8) and
for each of the other causes (range 4.6-12.5). The risk
P.549
for recurrent infant mortality in subsequent siblings from the same cause as in the index sibling
thus appears to be increased to a similar degree for both explained causes and for SIDS. These
increased risks in SIDS families are consistent with genetic risk factors associated with brainstem
abnormalities in autonomic control interacting with environmental risk factors.
TABLE 28-7 PHYSIOLOGIC ABNORMALITIES THAT HAVE BEEN ASSOCIATED WITH SIDS
Deficient regulation of brainstem autonomic control, affecting

Arousal/gasping
Ventilatory responsiveness
Respiratory pattern
Heart rate
Temperature
Vagal tone
Blood pressure
Sleep/Circadian Rhythm
Metabolic (e.g.; fatty acid defects)
Abnormal inflammatory/immune response to infection
Prolonged QT syndrome
SIDS, sudden infant death syndrome.
Adapted from Hunt, CE, Hauck FR. Sudden infant death syndrome.
In: Behrman RE, Kliegman RM, Jenson HB, eds. Nelson textbook of pediatrics, 17th ed.
Philadelphia: Elsevier, 2004: 1380, with permission.
Physiologic Studies
Physiologic studies have been performed in healthy infants during early infancy, a few of whom
later died of SIDS (99). Physiologic studies have also been performed in infant groups at
increased risk for SIDS, especially in infants having experienced an unexplained ALTE and in
subsequent siblings of SIDS victims (Table 28-7). In the aggregate, these physiologic studies are
indicative of a brainstem abnormality related to neuroregulation of cardiorespiratory control and
other autonomic functions and are consistent with the autopsy findings in SIDS victims. The
observed physiologic abnormalities include respiratory pattern, chemoreceptor sensitivity, control
of heart and respiratory rate or variability, and asphyxic arousal responsiveness. A deficit in
arousal responsiveness may be a necessary prerequisite for SIDS to occur but may be insufficient
to cause SIDS in the absence of other genetic or environmental risk factors. Autoresuscitation
(gasping) is a critical component of the asphyxic arousal response, and a failure of
autoresuscitation in victims of SIDS may be the final and most devastating physiologic failure
(126). Most full-term infants less than 9 weeks of age arouse in response to mild hypoxia, but
only 10% to 15% of normal infants older than 9 weeks of age arouse (127). These data thus
suggest that as full-term infants mature, their ability to arouse to mild/moderate hypoxic stimuli
diminishes as they reach the age range of greatest risk for SIDS.
The ability to shorten Q-T interval as heart rate increases appears to be impaired in some SIDS
victims, suggesting that such infants may be predisposed to ventricular arrhythmia (100). Infants
studied physiologically and later dying of SIDS have higher heart rates in all sleep-waking states,
diminished heart rate variability during wakefulness, and significantly lower heart rate variability
at the respiratory frequency across all sleep-waking cycles. Even in early infancy, therefore,
future SIDS victims differ in the extent to which cardiac and respiratory activity are coupled. No
heart rate variability data are available in preterm infants with AOP compared to gestational agematched infants without AOP or to full-term infants.
Part of the decreased heart rate variability and increased heart rate observed in infants who later
die of SIDS may be related to decreased vagal tone. This decreased tone appears at least in part
to be related to vagal neuropathy or to brainstem damage in areas responsible for
parasympathetic cardiac control. Comparing heart rate power spectra before and after
obstructive apneas in infants, future SIDS victims do not have the decreases in low-frequency to
high-frequency power ratios observed in control infants (128). Some future SIDS victims thus
have different autonomic responsiveness to obstructive apnea, perhaps indicating impaired
autonomic nervous system control associated with higher vulnerability to external or endogenous
stress factors and hence to reduced electrical stability of the heart.
Home cardiorespiratory monitors with memory capability have recorded the terminal events in
some SIDS victims (129). These recordings, however, have not included SpO2 and do not permit
identification of obstructed breaths due to reliance on transthoracic impedance for breath
detection. In most instances, there has been sudden and rapid progression of severe bradycardia
that is either unassociated with central apnea or appears to occur too soon to be explained by the
central apnea. These observations are consistent with an abnormality in autonomic control of
heart rate variability, or with obstructed breaths and associated bradycardia or hypoxemia.
Reducing Risk for Sudden Infant Death Syndrome

The international back-to-sleep campaigns have significantly reduced environmental risks for
SIDS and have been associated with reductions in SIDS rates of 50% or more since the early
1990s. Despite the successes with risk reduction, however, there is no intervention that can be
guaranteed to prevent SIDS in individual infants. Home electronic surveillance for apnea and
bradycardia has been utilized for more than two decades, but no suitably designed prospective
studies have been performed and hence no data exist to support a role for home
cardiorespiratory monitors in preventing SIDS (130).
Limited data are available from the National Maternal and Infant Health Survey regarding extent
of home monitor
P.550
use in preterm and full-term SIDS victims and live controls (132). The prevalence estimates for
monitor use for birth weights of 500 to 1,499, 1,500 to 2,499, and 2,500 grams or more were
20%, 3%, and 1% in African-American infants versus 44%, 9%, and 1% for other infants. In no
instance was there a significant difference in prevalence of home monitor use in SIDS versus
case controls, and in African-American infants with birth weights of 500 to 1,499 grams, the
adjusted odds ratio for SIDS was significantly higher (3.9) with a home monitor. This higher odds
ratio likely reflects a higher risk for SIDS in those infants monitored in this retrospective
nonrandomized observational study.
Preventing extreme events such as observed in the CHIME study may be necessary if home
monitors are to have any potential for reducing risk for SIDS (47). Even if prolonged apnea
including obstructed breaths were not a primary autonomic abnormality contributing to risk for
SIDS, home monitoring could still be potentially helpful if apnea of any type causing bradycardia
or hypoxemia as part of the terminal event could be reliably detected sufficiently early to be
amenable to intervention before the onset of life-threatening hypoxemia. However, this
hypothesis remains untested and neither home electronic surveillance using current technology
nor any other intervention can be recommended as a strategy to prevent SIDS. Despite absence
of any proven prospective intervention to prevent SIDS, dramatic decreases in population-based
risk can be achieved by eliminating modifiable risk factors associated with SIDS.
REFERENCES
1. Rigatto H. Maturation of breathing. Clin Perinatol 1992;19: 739-756.
2. Haddad G. Respiratory control in the newborn: comparative physiology and clinical

disorders. In: Mathew OP, ed. Respiratory control and disorders in the newborn. New York:
Marcel Dekker, 2003:1-13.
3. Hunt CE. Sudden infant death syndrome and other causes of infant mortality: diagnosis,
mechanisms, and risk for recurrence in siblings. Am J Respir Crit Care Med 2001;164:346-357.
4. Kahn A, Franco P, Scaillet S, et al. Development of cardiopulmonary integration and the role
of arousability from sleep. Curr Opin Pulm Med 1997;3:440-444.
5. Katz DM, Balkowiec A. New insights into the ontogeny of breathing from genetically
engineered mice. Curr Opin Pulm Med 1997;3:433-439.
6. Fortin G, del Toro ED, Abadie V, et al. Genetic and developmental models for the neural
control of breathing in vertebrates. Respir Physiol 2000;122:247-257.
7. Dauger S, Renolleau S, Vardon G, et al. Ventilatory responses to hypercapnia and hypoxia in

Mash-1 heterozygous newborn and adult mice. Pediatr Res 1999;46:535-542.
8. Panigrahy A, Filiano J, Sleeper LA, et al. Decreased serotonergic receptor binding in rhombic
lip-derived regions of the medulla oblongata in the sudden infant death syndrome. J
Neuropathol Exp Neurol 2000;59:377-384.
9. Narita N, Narita M, Takashima S, et al. Serotonin transporter gene variation is a risk factor
for sudden infant death syndrome in the Japanese population. Pediatrics 2001;107:690-692.
10. Weese-Mayer DE, Berry-Kravis EM, Maher BS, et al. Sudden infant death syndrome:
association with a promoter polymorphism of the serotonin transporter gene. Am J Med Genet
2003;117A:268-274.
11. Dawes GS. Breathing before birth in animals and man. An essay in developmental
medicine. N Engl J Med 1974;290:557-559.
12. Kaplan M. Fetal breathing movements. An update for the pediatrician. Am J Dis Child
1983;137:177-181.
13. Patrick J, Campbell K, Carmichael L, et al. A definition of human fetal apnea and the
distribution of fetal apneic intervals during the last ten weeks of pregnancy. Am J Obstet
Gynecol 1980;136:471-472.
14. Natale R, Nasello-Paterson C, Connors G. Patterns of fetal breathing activity in the human
fetus at 24 to 28 weeks of gestation. Am J Obstet Gynecol 1988;158:317-321.
15. Kalache KD, Chaoui R, Marcks B, et al. Differentiation between human fetal breathing
patterns by investigation of breathing-related tracheal fluid flow velocity using Doppler
sonography. Prenat Diagn 2000;20:45-50.
16. Goodman JD. The effect of intravenous glucose on human fetal breathing measured by
Doppler ultrasound. Br J Obstet Gynaecol 1980;87:1080-1083.
17. Trudinger BJ, Lewis PJ, Petit B. Fetal breathing patterns in intrauterine growth retardation.
Br J Obstet Gynaecol 1979;86:432-436.
18. Manning FA, Platt LD. Fetal breathing movements: antepartum monitoring of fetal
condition. Clin Obstet Gynaecol 1979;6: 335-349.
19. Ritchie JW, Lakhani K. Fetal breathing movements in response to maternal inhalation of
5% carbon dioxide. Am J Obstet Gynecol 1980;136:386-388.
20. Devoe LD, Abduljabbar H, Carmichael L, et al. The effects of maternal hyperoxia on fetal
breathing movements in third-trimester pregnancies. Am J Obstet Gynecol 1984;148:790-794.
21. Bekedam DJ, Mulder EJ, Snijders RJ, et al. The effects of maternal hyperoxia on fetal
breathing movements, body movements and hear rate variation in growth retarded fetuses.
Early Hum Dev 1991;27:223-232.
22. Hallak M, Moise K Jr, Lira N, et al. The effect of tocolytic agents (indomethacin and
terbutaline) on fetal breathing and body movements: a prospective, randomized, double-blind,
placebo-controlled clinical trial. Am J Obstet Gynecol 1992;167: 1059-1063.
23. Rigatto H, Moore M, Cates D. Fetal breathing and behavior measured through a double-wall
Plexiglas window in sheep. J Appl Physiol 1986;61:160-164.
24. Boddy K, Dawes GS. Fetal breathing. Br Med Bull 1975;31:3-7.
25. Dawes GS, Gardner WN, Johnston BM, et al. Effects of hypercapnia on tracheal pressure,
diaphragm and intercostal electromyograms in unanaesthetized fetal lambs. J Physiol
1982;326:461-474.
26. Moss IR, Scarpelli EM. Generation and regulation of breathing in utero: fetal CO2 response
test. J Appl Physiol 1979;47:527-531.
27. Clewlow F, Dawes GS, Johnston BM, et al. Changes in breathing, electrocortical and muscle
activity in unanaesthetized fetal lambs with age. J Physiol 1983;341:463-476.
28. Condorelli S, Scarpelli EM. Fetal breathing: induction in utero and effects of vagotomy and
barbiturates. J Pediatr 1976;88:94-101.
29. Jansen AH, Ioffe S, Russell BJ, et al. Effect of carotid chemoreceptor denervation on
breathing in utero and after birth. J Appl Physiol 1981;51:630-633.
30. Gluckman PD, Gunn TR, Johnston BM. The effect of cooling on breathing and shivering in
unanaesthetized fetal lambs in utero. J Physiol 1983;343:495-506.
31. Gerhardt T, Bancalari E. Apnea of prematurity. II. Respiratory reflexes. Pediatrics

1984;74:63-66.
32. Alvaro R, Alvarez J, Kwiatkowski K, et al. Small preterm infants (less than or equal to 1500
g) have only a sustained decrease in ventilation in response to hypoxia. Pediatr Res
1992;32:403-406.
33. Gluckman PD, Johnston BM. Lesions in the upper lateral pons abolish the hypoxic
depression of breathing in unanaesthetized fetal lambs in utero. J Physiol 1987;382:373-383.
34. Miller MJ, Martin RJ. Apnea of prematurity. Clin Perinatol 1992;19:789-808.
35. Thach BT, Schefft GL, Pickens DL, et al. Influence of upper airway negative pressure reflex
on response to airway occlusion in sleeping infants. J Appl Physiol 1989;67:749-755.
36. Menon AP, Schefft GL, Thach BT. Frequency and significance of swallowing during
prolonged apnea in infants. Am Rev Respir Dis 1984;130:969-973.
37. Lawson EE. Prolonged central respiratory inhibition following reflex-induced apnea. J Appl
Physiol 1981;50:874-879.
38. Perkett EA, Vaughan RL. Evidence for a laryngeal chemoreflex in some human preterm
infants. Acta Paediatr Scand 1982;71:969- 972.
P.551
39. National Institutes of Health Consensus Development Conference on Infantile Apnea and
Home Monitoring, Sept 29 to Oct 1, 1986. Pediatrics 1987;79:292-299.
40. Rigatto H. Breathing and sleep in preterm infants. In: Loughlin GM, Carroll JL, Marcus CL,
eds. Sleep and breathing in children. a developmental approach. New York: Marcel Dekker,
2000:495-523.
41. Glotzbach SF, Baldwin RB, Lederer NE, et al. Periodic breathing in preterm infants:
incidence and characteristics. Pediatrics 1989; 84:785-792.
42. Richards JM, Alexander JR, Shinebourne EA, et al. Sequential 22-hour profiles of breathing
patterns and heart rate in 110 full-term infants during their first 6 months of life. Pediatrics
1984;74: 763-777.
43. Kelly DH, Stellwagen LM, Kaitz E, et al. Apnea and periodic breathing in normal full-term
infants during the first twelve months. Pediatr Pulmonol 1985;1:215-219.
44. Hunt CE. Apnea and sudden infant death syndrome. In: Kliegman RM, Nieder ML, Super
DM, eds. Practical strategies in pediatric diagnosis and therapy. Philadelphia: WB Saunders,
1996: 135-147.
45. Henderson-Smart DJ, Pettigrew AG, Campbell DJ. Clinical apnea and brain-stem neural
function in preterm infants. N Engl J Med 1983;308:353-357.
46. Ramanathan R, Corwin MJ, Hunt CE, et al. Cardiorespiratory events recorded on home
monitors. Comparison of healthy infants with those at increased risk for SIDS. JAMA 2001;285:
2199-2207.
47. Hunt CE. Sudden infant death syndrome. In: Spitzer AR, ed. Intensive care of the fetus
and neonate, 2nd ed. St. Louis: Elsevier, 2005:.
48. Alden ER, Mandelkorn T, Woodrum DE, et al. Morbidity and mortality of infants weighing
less than 1,000 grams in an intensive care nursery. Pediatrics 1972;50:40-49.
49. Carlo WA, Martin RJ, Versteegh FG, et al. The effect of respiratory distress syndrome on
chest wall movements and respiratory pauses in preterm infants. Am Rev Respir Dis
1982;126:103-107.
50. Kattwinkel J, Nearman HS, Fanaroff AA, et al. Apnea of prematurity. Comparative
therapeutic effects of cutaneous stimulation and nasal continuous positive airway pressure. J
Pediatr 1975; 86:588-592.
51. Joshi A, Gerhardt T, Shandloff P, et al. Blood transfusion effect on the respiratory pattern
of preterm infants. Pediatrics 1987;80: 79-84.
52. Ross MP, Christensen RD, Rothstein G, et al. A randomized trial to develop criteria for
administering erythrocyte transfusions to anemic preterm infants 1 to 3 months of age. J
Perinatol 1989;9: 246-253.
53. Sasidharan P, Heimler R. Transfusion-induced changes in the breathing pattern of healthy

preterm anemic infants. Pediatr Pulmonol 1992;12:170-173.
54. Bifano EM, Smith F, Borer J. Relationship between determinants of oxygen delivery and
respiratory abnormalities in preterm infants with anemia. J Pediatr 1992;120:292-206.
55. Keyes WG, Donohue PK, Spivak JL, et al. Assessing the need for transfusion of premature
infants and role of hematocrit, clinical signs, and erythropoietin level. Pediatrics 1989;84:412417.
56. Poets CF, Pauls U, Bohnhorst B. Effect of blood transfusion on apnoea, bradycardia and
hypoxaemia in preterm infants. Eur J Pediatr 1997;156:311-316.
57. Lawson EE. Nonpharmacological management of idiopathic apnea of the premature infant.
In: Mathew OP, ed. Respiratory control and disorders in the newborn. New York: Marcel
Dekker, 2003;335-354.
58. Peliowski A, Finer NN. A blinded, randomized, placebo-controlled trial to compare

theophylline and doxapram for the treatment of apnea of prematurity. J Pediatr 1990;116:648653.
59. Aranda JV, Gorman W, Bergsteinsson H, et al. Efficacy of caffeine in treatment of apnea in
the low-birth-weight infant. J Pediatr 1977;90:467-472.
60. Erenberg A, Leff RD, Haack DG, et al. Caffeine citrate for the treatment of apnea of
prematurity: a double-blind, placebo-controlled study. Pharmacotherapy 2000;20:644-652.
61. Fuglsang G, Nielsen K, Kjaer NL, et al. The effect of caffeine compared with theophylline in
the treatment of idiopathic apnea in premature infants. Acta Paediatr Scand 1989;78:786-788.
62. Bairam A, Boutroy MJ, Badonnel Y, et al. Theophylline versus caffeine: comparative effects
in treatment of idiopathic apnea in the preterm infant. J Pediatr 1987;110:636-639.
63. Brouard C, Moriette G, Murat I, et al. Comparative efficacy of theophylline and caffeine in
the treatment of idiopathic apnea in premature infants. Am J Dis Child 1985;139:698-700.
64. Sims ME, Yau G, Rambhatla S, et al. Limitations of theophylline in the treatment of apnea
of prematurity. Am J Dis Child 1985; 139:567-570.
65. Murat I, Moriette G, Blin MC, et al. The efficacy of caffeine in the treatment of recurrent
idiopathic apnea in premature infants. J Pediatr 1981;99:984-989.
66. Gupta JM, Mercer HP, Koo WW. Theophylline in treatment of apnoea of prematurity. Aust
Paediatr J 1981;17:290-291.
67. Jones RA. Apnoea of immaturity. 1. A controlled trial of theophylline and face mask
continuous positive airways pressure. Arch Dis Child 1982;57:761-765.
68. Henderson-Smart DJ, Subramaniam P, Davis PG. Continuous positive airway pressure
versus theophylline for apnea in preterm infants. Cochrane Database Syst Rev 2001;4:
CD001072.
69. Miller MJ, Carlo WA, Martin RJ. Continuous positive airway pressure selectively reduces
obstructive apnea in preterm infants. J Pediatr 1985;106:91-94.
70. Khalaf MN, Brodsky N, Hurley J, et al. A prospective randomized, controlled trial comparing
synchronized nasal intermittent positive pressure ventilation versus nasal continuous positive
airway pressure as modes of extubation. Pediatrics 2001;108: 13-17.
71. Courtney SE, Pyon KH, Saslow JG, et al. Lung recruitment and breathing pattern during
variable versus continuous flow nasal continuous positive airway pressure in premature infants:
an evaluation of three devices. Pediatrics 2001;107:304-308.
72. Lemyre B, Davis PG, De Paoli AG. Nasal intermittent positive pressure ventilation (NIPPV)
versus nasal continuous positive airway pressure (NCPAP) for apnea of prematurity. Cochrane
73. Hunt CE, Inwood RJ, Shannon DC. Respiratory and nonrespiratory effects of doxapram in
congenital central hypoventilation syndrome. Am Rev Respir Dis 1979;119:263-269.
74. Barrington KJ, Finer NN, Peters KL, et al. Physiologic effects of doxapram in idiopathic
apnea of prematurity. J Pediatr 1986; 108:124-129.
75. Brion LP, Vega-Rich C, Reinersman G, et al. Low-dose doxapram for apnea unresponsive to
aminophylline in very low birthweight infants. J Perinatol 1991;11:359-364.
76. Arad-Cohen N, Cohen A, Tirosh E. The relationship between gastroesophageal reflux and
apnea in infants. J Pediatr 2000;137: 321-326.
77. Kahn A, Rebuffat E, Sottiaux M, et al. Lack of temporal relation between acid reflux in the
proximal oesophagus and cardiorespiratory events in sleeping infants. Eur J Pediatr 1992;151:
208-212.
78. Peter CS, Sprodowski N, Bohnhorst B, et al. Gastroesophageal reflux and apnea of
prematurity: no temporal relationship. Pediatrics 2002;109:8-11.
79. Wenzl TG, Schenke S, Peschgens T, et al. Association of apnea and nonacid
gastroesophageal reflux in infants: investigations with the intraluminal impedance technique.
Pediatr Pulmonol 2001;31:144-149.
80. Mathew OP. Respiratory control during nipple feeding in preterm infants. Pediatr Pulmonol
1988;5:220-224.
81. Shivpuri CR, Martin RJ, Carlo WA, et al. Decreased ventilation in preterm infants during
oral feeding. J Pediatr 1983;103:285-289.
82. Bruhn FW, Mokrohisky ST, McIntosh K. Apnea associated with respiratory syncytial virus
infection in young infants. J Pediatr 1977;90:382-386.
83. Kneyber MC, Brandenburg AH, de Groot R, et al. Risk factors for respiratory syncytial virus
associated apnoea. Eur J Pediatr 1998;157:331-335.
84. Cote CJ, Zaslavsky A, Downes JJ, et al. Postoperative apnea in former preterm infants after
inguinal herniorrhaphy. A combined analysis. Anesthesiology 1995;82:809-822.
85. Welborn LG, Hannallah RS, Luban NL, et al. Anemia and postoperative apnea in former
preterm infants. Anesthesiology 1991; 74:1003-1006.
86. Darnall RA, Kattwinkel J, Nattie C, et al. Margin of safety for discharge after apnea in
preterm infants. Pediatrics 1997;100:795-801.
87. Di Fiore JM, Arko MK, Miller MJ, et al. Cardiorespiratory events in preterm infants referred
for apnea monitoring studies. Pediatrics 2001;108:1304-1308.
P.552
88. Barrington KJ, Finer N, Li D. Predischarge respiratory recordings in very low birth weight
newborn infants. J Pediatr 1996;129: 934-940.
89. Schmidt B. Methylxanthine therapy in premature infants: sound practice, disaster, or

fruitless byway? J Pediatr 1999;135:526-528.
90. Martin RJ, Fanaroff AA. Neonatal apnea, bradycardia, or desaturation: does it matter? J
Pediatr 1998;132:758-759.
91. Cheung PY, Barrington KJ, Finer NN, et al. Early childhood neurodevelopment in very low
birth weight infants with predischarge apnea. Pediatr Pulmonol 1999;27:14-20.
92. Eichenwald EC, Aina A, Stark AR. Apnea frequently persists beyond term gestation in
infants delivered at 24 to 28 weeks. Pediatrics 1997;100:354-359.
93. Sreenan C, Etches PC, Demianczuk N, et al. Isolated mental developmental delay in very
low birth weight infants: association with prolonged doxapram therapy for apnea. J Pediatr
2001;139:832-837.
94. Hunt CE, Baird T, The Collaborative Home Infant Monitoring Evaluation (CHIME) Study
Group, et al. Cardiorespiratory events detected by home memory monitoring and
neurodevelopmental outcome at one year of age. J Pediatr 2004;145:465-471
95. Gozal D. Sleep-disordered breathing and school performance in children. Pediatrics

1998;102:616-620.
96. Macey PM, Henderson LA, Macey KE, et al. Brain morphology associated with obstructive
sleep apnea. Am J Respir Crit Care Med 2002;166:1382-1387.
97. Kim HC, Young T, Matthews CG, et al. Sleep-disordered breathing and neuropsychological
deficits. A population-based study. Am J Respir Crit Care Med 1997;156:1813-1819.
98. Samuels MP. Apparent life-threatening events: pathogenesis and management. In:
Loughlin GM, Carroll JL, Marcus CL, eds. Sleep and breathing in children: a developmental
approach. New York: Marcel Dekker, 2000:423-441.
99. Hunt CE, Hauck FR. Sudden Infant Syndrome. In: Behrman RE, Kliegman RM, Jenson HB,
eds. Nelson textbook of pediatrics, 17th ed. Philadelphia: Elsevier, 2004:1380-1385.
100. American Academy of Pediatrics. Task Force on Infant Sleep Position and Sudden Infant
Death Syndrome. Changing concepts of sudden infant death syndrome: implications for infant
sleeping environment and sleep position. Pediatrics 2000;105: 650-656.
101. Willinger M, James LS, Catz C. Defining the sudden infant death syndrome (SIDS):
deliberations of an expert panel convened by the National Institute of Child Health and Human
Development. Pediatr Pathol 1991;11:677-684.
102. Hoyert DL, Freedman MA, Strobino DM, et al. Annual summary of vital statistics: 2000.
Pediatrics 2001;108:1241-1255.
103. Malloy MH, Hoffman HJ. Prematurity, sudden infant death syndrome, and age of death.
Pediatrics 1995;96:464-471.
104. Valdes-Dapena M. The sudden infant death syndrome: pathologic findings. Clin Perinatol
1992;19:701-716.
105. Naeye RL. Sudden infant death syndrome, is the confusion ending? Mod Pathol
1988;1:169-174.
106. Jones KL, Krous HF, Nadeau J, et al. Vascular endothelial growth factor in the
cerebrospinal fluid of infants who died of sudden infant death syndrome: evidence for
antecedent hypoxia. Pediatrics 2003;111:358-363.
107. Rognum TO, Saugstad OD. Hypoxanthine levels in vitreous humor: evidence of hypoxia in
most infants who died of sudden infant death syndrome. Pediatrics 1991;87:306-310.
108. Obonai T, Yasuhara M, Nakamura T, et al. Catecholamine neurons alteration in the

brainstem of sudden infant death syndrome victims. Pediatrics 1998;101:285-288.
109. Ozawa Y, Okado N. Alteration of serotonergic receptors in the brain stems of human
patients with respiratory disorders. Neuropediatrics 2002;33:142-149.
110. Filiano JJ, Kinney HC. Arcuate nucleus hypoplasia in the sudden infant death syndrome. J
Neuropathol Exp Neurol 1992;51:394-403.
111. Matturri L, Biondo B, Suarez-Mier MP, et al. Brain stem lesions in the sudden infant death
syndrome: variability in the hypoplasia of the arcuate nucleus. Acta Neuropathol (Berl) 2002;
104:12-20.
112. Hasan SU, Simakajornboon N, MacKinnon Y, et al. Prenatal cigarette smoke exposure
selectively alters protein kinase C and nitric oxide synthase expression within the neonatal rat
brainstem. Neurosci Lett 2001;301:135-138.
113. Vernacchio L, Corwin MJ, Lesko SM, et al. Sleep position of low birth weight infants.
Pediatrics 2003;111:633-640.
114. Malloy MH, Freeman DH Jr. Birth weight- and gestational age-specific sudden infant death
syndrome mortality: United States, 1991 versus 1995. Pediatrics 2000;105:1227-1231.
115. Hunt CE, Lesko SM, Vezina RM, et al. Infant sleep position and associated health
outcomes. Arch Pediatr Adolesc Med 2003;157: 469-474.
116. McMartin KI, Platt MS, Hackman R, et al. Lung tissue concentrations of nicotine in sudden
infant death syndrome (SIDS). J Pediatr 2002;140:205-209.
117. Wisborg K, Kesmodel U, Henriksen TB, et al. A prospective study of smoking during
pregnancy and SIDS. Arch Dis Child 2000;83: 203-206.
118. Froen JF, Akre H, Stray-Pedersen B, et al. Adverse effects of nicotine and interleukin1beta on autoresuscitation after apnea in piglets: implications for sudden infant death
syndrome. Pediatrics 2000;105:E52.
119. Gordon AE, El Ahmer OR, Chan R, et al. Why is smoking a risk factor for sudden infant
death syndrome? Child Care Health Dev 2002;28(Suppl 1):23-25.
120. Iyasu S, Randall LL, Welty TK, et al. Risk factors for sudden infant death syndrome among
northern plains Indians. JAMA 2002;288:2717-2723.
121. Hoffman HJ, Hillman LS. Epidemiology of the sudden infant death syndrome: maternal,
neonatal, and postneonatal risk factors. Clin Perinatol 1992;19:717-737.
122. Ward SL, Bautista D, Chan L, et al. Sudden infant death syndrome in infants of substanceabusing mothers. J Pediatr 1990;117:876-881.
123. Blackwell CC, Weir DM, Busuttil A. Infection, inflammation and sleep: more pieces to the
puzzle of sudden infant death syndrome (SIDS). APMIS 1999;107:455-473.
124. Gold Y, Goldberg A, Sivan Y. Hyper-releasability of mast cells in family members of

infants with sudden infant death syndrome and apparent life-threatening events. J Pediatr
2000; 136:460- 465.
125. Waters KA, Gonzalez A, Jean C, et al. Face-straight-down and face-near-straight-down

positions in healthy, prone-sleeping infants. J Pediatr 1996;128:616-625.
126. Poets CF, Meny RG, Chobanian MR, et al. Gasping and other cardiorespiratory patterns
during sudden infant deaths. Pediatr Res 1999;45:350-354.
127. Ward SL, Bautista DB, Keens TG. Hypoxic arousal responses in normal infants. Pediatrics
1992;89:860-864.
128. Franco P, Szliwowski H, Dramaix M, et al. Decreased autonomic responses to obstructive

sleep events in future victims of sudden infant death syndrome. Pediatr Res 1999;46:33-39.
129. Meny RG, Carroll JL, Carbone MT, et al. Cardiorespiratory recordings from infants dying
suddenly and unexpectedly at home. Pediatrics 1994;93:44-49.
130. American Academy of Pediatrics. Committee on Fetus and Newborn. Apnea, sudden infant
death syndrome, and home monitoring. Pediatrics 2003;111:914-917.
131. Malloy MH, Hoffman HJ. Home apnea monitoring and sudden infant death syndrome. Prev
Med 1996;25:645-649.
132. Martin RJ, Miller MJ, Carlo WA. Pathogenesis of apnea in preterm infants. J Pediatr
1986;109:733-741.
133. Gerhardt T, Bancalari E. Apnea of prematurity: I. Lung function and regulation of

breathing. Pediatrics 1984;74:58-62.
134. Miller MJ, Martin RJ, Carlo WA. Diagnostic methods and clinical disorders in children. In
Edelman NH, Santiago TV, eds. Breathing disorders of sleep. New York: Churchill Livingstone,
1986: 157-180.
Chapter 29
Acute Respiratory Disorders
Jeffrey A. Whitsett
Ward R. Rice
Barbara B. Warner
Susan E. Wert
Gloria S. Pryhuber
Successful adaptation to air breathing at the time of birth is the culmination of an orderly process of growth and
differentiation of pulmonary cells, leading to alveolar and capillary surfaces capable of providing oxygen and
eliminating carbon dioxide. Failure to achieve adequate gas exchange at birth represents a major cause of
perinatal morbidity and mortality. This chapter reviews the common disorders of neonatal respiratory
adaptation, including respiratory distress syndrome (RDS), pulmonary meconium aspiration syndrome (MAS),
pulmonary hypertension, pneumonia, air leak, pulmonary hemorrhage, transient tachypnea of the newborn
(TTN), and other causes of acute respiratory dysfunction in the perinatal period. The clinical manifestations and
therapy of these disorders are discussed in the context of the morphologic, biochemical, and physiologic factors
critical to normal pulmonary growth, maturation, and function in the newborn.
HUMAN LUNG DEVELOPMENT

Human lung development is normally divided into five distinct stages of organogenesis, which describe the
histologic changes that the lung undergoes during morphogenesis and maturation of its structural elements
(1,2). These five stages include the embryonic, pseudoglandular, canalicular, saccular, and alveolar stages of
lung development. Human lung development is initiated during the early embryonic period of gestation (3-7
weeks postcoitum) as a small saccular outgrowth of the ventral foregut called the respiratory diverticulum.
During the subsequent pseudoglandular stage of lung development (5-17 weeks postcoitum), formation of the
conducting airways, i.e., the tracheobronchial tree, occurs by elongation and repetitive branching of the
primitive bronchial tubules. Vascularization of the surrounding mesenchyme with formation of the air-blood
barrier, i.e., the alveolar-capillary membrane, occurs during the canalicular stage of lung development (16-26
weeks postcoitum). Cytodifferentiation of bronchiolar and alveolar epithelial cells is also initiated during this
stage. Enlargement and expansion of the peripheral air spaces occurs during the saccular stage of lung
development (24-38 weeks postcoitum), resulting in the formation of primitive sac-like alveoli and thick
interalveolar septa. Formation of thin secondary alveolar septa and remodeling of the capillary bed occurs during
the alveolar stage of lung development (36 weeks postcoitum to 8 years of age), giving rise to the mature
alveolar organization of the adult lung (Fig. 29-1).
The human lung is a derivative of the primitive foregut endoderm and the surrounding splanchnic mesoderm.
The primitive respiratory diverticulum appears at 3 weeks of gestation as an enlargement of the caudal end of
the laryngotracheal sulcus located in the median pharyngeal groove, which is an outgrowth of the ventral wall of
the primitive esophagus. During the fourth week of gestation, the respiratory diverticulum enlarges and
subdivides into the left and right primary bronchi (see Fig. 29-1A,B). The primitive lung grows caudally,
expanding into the mesenchyme surrounding the primitive foregut, although the trachea becomes separated
from the esophagus by a band of mesenchymal tissue called the tracheoesophageal septum. Between 4 and 5
weeks of gestation, the left and right primary bronchi subdivide to produce secondary, or lobar, bronchi (see Fig.
29-1C,D). Further subdivision of lobar bronchi into tertiary or segmental bronchi occurs during the sixth week of
gestation, with the lung taking on a lobulated appearance as the segmental buds are formed (see Fig. 29-1E,F).
The developing respiratory tract is lined by endodermally derived epithelium that forms the conducting airways
and alveoli. The surrounding mesoderm is composed of mesenchymal cells that will differentiate into connective
tissue components,
P.554
P.555
including blood vessels, fibroblasts, smooth muscle cells, and cartilage.
Figure 29-1 Lung development during the embryonic (A-F) and pseudoglandular (G, H) stages of
organogenesis. The overall branching pattern of the primitive lung (left panels) results in the development of
the bronchial tree. The histologic organization of the fetal lung becomes more complex as branching
morphogenesis progresses through these stages (right panels).
Preacinar blood vessels first appear at the end of week 4. Pulmonary arteries arise from the sixth pair of aortic
arches and grow into the mesenchyme, in which they accompany the developing airways, segmenting with each
bronchial subdivision. Pulmonary veins develop as outgrowths of the left atrium of the heart and subdivide
several times before connecting to the pulmonary vascular bed. Intraacinar vessels develop later, in parallel with
alveolar formation.
The pseudoglandular stage of fetal lung development extends from about 5 to 17 weeks of gestation and is
marked by the formation of the bronchial portion of the lung. This occurs through a process known as branching
morphogenesis, during which the segmental tubules of the developing lung undergo repetitive lateral and
terminal dichotomous branching to form the primitive bronchial tree (see Fig. 29-1G,H). By week 17 of
gestation, the segmental bronchi have subdivided to produce approximately 23 generations of bronchial tubules
ending in the terminal bronchioles. These bronchial tubules are lined initially by a pseudostratified columnar
epithelium containing large pools of glycogen. A prominent basement membrane underlies the epithelium, and
mesenchymal cells adjacent to these tubules differentiate into fibroblasts, which become organized in a
circumferential orientation, perpendicular to the long axis of the tubules. As branching progresses,
pseudostratified columnar epithelium is reduced to a tall columnar epithelium, especially in distal regions of the
bronchial tree. During this period, cytodifferentiation of the airway epithelium occurs in a centrifugal direction
with ciliated, nonciliated, goblet, neuroendocrine, and basal cells appearing first in the more proximal airways.
Cartilage, smooth muscle cells, and mucous glands are also found in the trachea during the pseudoglandular
stage of development and extend as far as the segmental bronchi.
The canalicular stage of fetal lung development extends from week 16 to 26 of gestation. By the end of week
16, the terminal bronchioles have divided into two or more respiratory bronchioles that have subdivided into
small clusters of short acinar tubules and buds lined by cuboidal epithelium. These structures undergo further
differentiation and maturation to become the adult respiratory unit, or pulmonary acinus, consisting of the
alveolated respiratory bronchiole, alveolar ducts, and alveoli. Clusters of acinar tubules and buds continue to
grow by lengthening, subdividing, and widening at the expense of the surrounding mesenchyme (Fig. 29-2A).
This peripheral growth is accompanied by the formation of intraacinar capillaries, which align themselves around
the air spaces, establishing contact with the overlying cuboidal epithelium. During this stage of lung
development, type II epithelial cell differentiation occurs in acinar tubules with formation of intracellular
multivesicular bodies and multilamellar bodies, the storage form of pulmonary surfactant phospholipids. Type I
epithelial cell differentiation occurs in conjunction with development of the air-blood barrier, wherever
endothelial cells of the developing capillary system come into contact with the overlying epithelial cells.
During the saccular stage of fetal lung development, which extends from week 24 to 38 of gestation, the
terminal clusters of acinar tubules and buds begin to dilate and expand into thin, smooth-walled, transitory
ducts and saccules that later become the true alveolar ducts and alveoli of the adult (Fig. 29-2B). During this
stage, there is a marked reduction in the amount of interstitial tissue. Intersaccular and interductal septa
develop, which contain a delicate network of collagen fibers and the intraacinar capillary bed. Near the end of
this stage, elastin is deposited in regions in which future interalveolar septa will form. Increasing amounts of
tubular myelin, the secretory form of pulmonary surfactant, are seen in the air spaces.
The alveolar stage, which extends from 36 weeks of gestation to between 2 and 8 years of age, is the last stage
of lung development and is marked by the formation of secondary alveolar septa, which partition the transitory
ducts and saccules into true alveolar ducts and alveoli (Fig. 29-2C,D). This process of alveolarization greatly
increases the surface area of the lung available for gas exchange. At the beginning of this period, the secondary
interalveolar septa consist of short buds or projections of connective tissue that contain a double capillary
network and interstitial cells that are actively synthesizing collagen and elastic fibers. By 5 months of age, these
secondary interalveolar septa have lengthened and thinned, and now contain only a single capillary network.
Although definitive alveoli can be found in the human lung by 36 weeks of gestation, 85 to 90% of all alveoli are
formed within the first 6 months of life. Overall the number of alveoli increases by about six-fold between birth
and adulthood, i.e., from an average of 50 million alveoli in the term lung to 300 million in the adult human
lung. After the first 6 months of life, alveolar formation occurs at a slower pace until about 2 to 8 years of age,
when further growth of the lung becomes proportional to growth of the body. The surface area available for gas
exchange, and its diffusion capacity, increases linearly with body weight up to about 18 years of age. The
conducting airways also increase in length and diameter, although airspace and capillary volume increase
coordinately at the expense of interstitial volume.
DEVELOPMENTAL ANOMALIES
Each of these stages of lung development includes distinct changes in tissue organization and cellular
differentiation that are important for subsequent growth and maturation of the lung. Structural and functional
defects in lung development at birth can often be traced to arrested or aberrant development during one of
these periods of organogenesis, often as a result of mutations in genes critical for patterning and growth of the
lung, such as the GLI gene (Pallister Hall Syndrome), which is a component of the
P.556
Sonic Hedgehog signaling pathway, and the fibroblast growth factor receptor, FGFR2, gene (Pfeiffer's, Apert's,
and Crouzon's syndromes). Developmental anomalies of the lung occur through defective division and
differentiation of the lung bud, of the left or right bronchial bud or of the trachea and esophagus. Pulmonary
agenesis, tracheal and bronchial malformations, tracheoesophageal fistulas, ectopic lobes, and bronchogenic and
pulmonary cysts arise during the embryonic and pseudoglandular stages of lung development. Clinical disorders
related to pulmonary hypoplasia, acinar dysplasia, alveolar capillary dysplasia, and respiratory insufficiency are
associated with later periods of development. Pulmonary hypoplasia can be caused by a reduction of space
within the pleural cavity, usually as a consequence of another primary developmental defect, such as congenital
diaphragmatic hernia, or by a reduction in the amount of amniotic fluid following premature rupture of
membranes or in association with renal dysgenesis Potter's syndrome. Respiratory distress syndrome and
bronchopulmonary dysplasia are associated with premature birth at a time when biochemical functions (e.g.,
surfactant production) and structural functions (e.g., elasticity) of the lung are still underdeveloped.
Figure 29-2 Lung development during the canalicular (A), saccular (B), and alveolar stages of organogenesis
(C, D). Dramatic histologic changes in tissue organization occur during these periods. The adult alveolar
epithelium is composed of squamous type I cells and cuboidal type II cells (inset).
Hereditary-Genetic Causes of Acute Respiratory Failure in the

Newborn
Mutations in the surfactant protein B and C genes are rare causes of acute respiratory failure in neonates (3,4).
Hereditary SP-B deficiency causes respiratory distress, generally in full-term infants, within the first day of life.
Respiratory failure progresses despite ventilatory support, surfactant replacement or ECMO, generally causing
death from respiratory failure in the first months of life. Mutations in the SP-B gene are inherited as autosomal
recessive traits, resulting in the lack of SP-B in the airways and severe surfactant dysfunction. Hereditary SP-B
deficiency is a lethal disease, although some infants have been successfully treated with lung transplantation.
The definitive diagnosis is made by genotyping. Mutations in the SP-C gene are generally inherited as autosomaldominant, and cause both acute and chronic interstitial lung disease in
P.557
newborns, infants, and adults. This disorder is associated with the lack of SP-C protein in alveolar lavage and
with mutations in the proSP-C protein. The diagnosis of SP-C deficiency is made by genotyping. Mutations in a
transport protein, ABCA3, also cause neonatal respiratory failure. Like hereditary SP-B deficiency, ABCA3 defects
are inherited as autosomal recessive genes and cause respiratory failure that is refractory to ventilation and
surfactant therapy (5).
THE SURFACTANT SYSTEM

The unique physical-chemical boundary between the alveolar gases and the highly solvated molecules at the
apical surface of the respiratory epithelium generates a region of high surface tension produced by the unequal
distribution of molecular forces among water molecules at an air-liquid interface. Surface-active material at this
interface in the alveoli provides surface-tension-lowering activity that contributes to the remarkable pressurevolume associations characteristic of the lung. This surface-active material, called surfactant, has been subject
to intense study in recent decades (6,7 and 8).
Deficiency or dysfunction of pulmonary surfactant plays a critical role in the pathogenesis of respiratory diseases
in the newborn period. Pulmonary surfactant exists in a variety of physical forms when isolated from the alveolar
wash of the lung. These physical forms include lamellated and vesicular forms and highly organized tubular
myelin. Tubular myelin is highly surface active and, although composed predominately of phospholipids, its
unique structure depends on Ca2+ and lung surfactant proteins A (SP-A), B (SP-B), and D (SP-D). Tubular
myelin represents the major extracellular pool of surfactant from which a lipid monolayered/multilayered film is
generated to produce an interface between the hydrated cellular surfaces and alveolar gas (Fig. 29-3).
Lamellated and vesicular forms of surfactant represent nascent or catabolic forms of surfactant material
respectively; the latter is taken up by type II epithelial cells and recycled. Surfactant proteins A, B, C, and D
play important roles in the organization and function of the surfactant complex regulating surfactant
homeostasis. Alveolar surfactant concentrations are tightly controlled by a variety of mechanisms that modulate
lipid and protein synthesis, storage, secretion, and recycling.
Figure 29-3 Surfactant phospholipids are synthesized in the endoplasmic reticulum, transported through the
Golgi apparatus to multivesicular bodies, and ultimately packaged in lamellar bodies before secretion. After
exocytosis of the lamellar bodies, surfactant phospholipids are organized into a complex lattice called tubular
myelin phospholipid that provides material for a monolayer-multilayer at the air-fluid interface in the alveolus.
Surfactant phospholipids and proteins are taken up by type II cells, probably transported by endosomal
multivesicular bodies, and then catabolized or transported to lamellar bodies for recycling. A fraction of
surfactant proteins and lipids are also catabolized by alveolar macrophages. Surfactant proteins are synthesized
in polyribosomes and extensively modified in the endoplasmic reticulum, Golgi apparatus, and multivesicular
bodies. Surfactant proteins are detected within lamellar bodies or in secretory vesicles closely associated with
lamellar bodies before secretion into the alveolus.
Composition of Surfactant
Pulmonary surfactant is composed primarily of the phospholipids phosphatidylcholine and phosphatidylglycerol
(Fig. 29-4). These lipid molecules are enriched in dipalmitoyl acyl groups attached to a glycerol backbone that
pack tightly and generate low surface pressures (Fig. 29-5). Rapid spreading and stability of pulmonary
surfactant are achieved by the interactions of surfactant proteins and phospholipids. Surfactant is synthesized
and secreted by
P.558
type II epithelial cells in the alveolus. Synthesis of phosphatidylcholine, surfactant proteins, and lamellar bodies,
an intracellular storage form of pulmonary surfactant, increases with advancing gestation. Lamellar bodies are
secreted into the lung liquid that contributes to the amniotic fluid. The measurement of amniotic fluid
phosphatidylcholine, disaturated phosphatidylcholine, phosphatidylglycerol, or the surfactant proteins has
provided useful biochemical markers that predict lung maturation and the adequacy of lung function at birth (e.
g., lecithin-sphingomyelin [L-S] ratio and phosphatidylglycerol values). Surfactant function can be assessed by a
variety of physical and physiologic tests that measure its ability to reduce surface tension at an air-liquid
interface and to spread rapidly during dynamic compression and expansion. The Wilhelmy balance, Langmuir
trough, pulsating bubble meter, and a variety of animal models have been used to assess the efficacy of
surfactant and surfactant replacements.
Figure 29-4 Pulmonary surfactant components are expressed as a percentage of the total weight. Chol,
cholesterol; DG, diacylglycerol; DPPC, dipalmitoylphosphatidylcholine; PA, phosphatidic acid; PC,
phosphatidylcholine; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; PI, phosphatidylinositol; SM,
sphingomyelin. Adapted from Possmayer F. Pulmonary surfactant. Can J Biochem Cell Biol 1984; 62:1121, with
permission.
Control of Surfactant Synthesis and Secretion

Synthesis of pulmonary surfactant is closely linked to the morphologic and biochemical differentiation of alveolar
type II cells in the peripheral respiratory epithelium. Interactions between mesenchymal and epithelial cells,
mediated by direct cell-cell contact or by paracrine factors, contribute to the differentiation process. Endocrine
factors also modulate the differentiation of type II epithelial cells and the synthesis of surfactant components. In
vivo and in vitro evidence supports the role of glucocorticoids in the modulation of morphologic differentiation
and production of phospholipids and surfactant proteins by the lung.
Phospholipid Synthesis
Phosphatidylcholine is produced by type II epithelial cells using extracellular substrate and the glycogen stores
that accumulate in the pretype II cells of the fetal lung. Metabolic pathways producing phosphatidylcholine
depend on the production of phosphatidic acid and a glycerophosphate backbone (see Fig. 29-5); the latter is
produced as an intermediate of the glycolytic pathway (9). The synthesis of phosphatidylcholine involves the
deacylation of phosphatidic acid and its reaction with cytidine diphosphocholine (CDP-choline).
Figure 29-5 Dipalmitoylphosphatidylcholine (DPPC) is the most abundant phospholipid in pulmonary

surfactant. The acyl chains of DPPC pack tightly to form the surfactant monolayer, reducing surface tension in
the alveolus.
Disaturated forms of phosphatidylcholine may be formed de novo, using disaturated acyl precursors or by
remodeling, i.e., salvage pathway, of phospholipids by deacylation and reacylation reactions. Production of CDPcholine is critical to phosphatidylcholine synthesis and is achieved by phosphorylation of choline and transfer to
cytidine triphosphate in a reaction dependent on choline kinase and choline phosphate cytidylyltransferase. The
activities of many of the enzymes in the synthetic pathway for phosphatidylcholine increase with advancing
gestation in the lung, generally increasing in the last third of gestation (9,10).
Glucocorticoid Enhancement of Pulmonary Maturation

A variety of hormonal factors influence the rate of production of the enzymes controlling phosphatidylcholine
synthesis in the developing lung (9,10). Glucocorticoids are the most clinically relevant and useful of these
agents. Studies in fetal lambs and humans demonstrated that administration of glucocorticoid to the dam or
mother resulted in precocious respiratory function in prematurely
P.559
born offspring. The initial clinical studies of Liggins and Howie demonstrated that maternal administration of
glucocorticoid decreased the incidence of respiratory distress in premature infants (11). Although the precise
mechanisms by which glucocorticoids induce pulmonary maturation and lung function in premature infants have
not been discerned, increased phosphatidylcholine synthesis and morphologic remodeling of the alveolar
architecture, including the thinning of interstitial components of the fetal lung, are observed after glucocorticoid
treatment. Glucocorticoids regulate several genes that are associated with the differentiation of the fetal lung,
including the genes encoding enzymes involved in the synthesis of phosphatidylcholine and the surfactant
proteins. The effects of glucocorticoid on lung cell differentiation are mediated in part by glucocorticoid
receptors, which, when occupied by hormones, influence gene transcription and mRNA stability, altering the
abundance of the proteins synthesized by pulmonary cells. Prenatal glucocorticoid therapy is useful in prevention
of RDS in preterm infants.
Other Hormonal Influences

Thyroid hormones i.e., T3, T4, thyrotropin-releasing hormone (TRH), estrogens, prolactin, epidermal growth
factor, -adrenergic agents, and other agents that enhance cellular cAMP levels influence pulmonary maturation
or biochemical indices of pulmonary maturation. Both T3 and T4 increase the synthesis of phospholipids in
mammalian lung but do not readily cross the placenta.
Surfactant Secretion
Surfactant is stored within type II cells in large lipid-rich organelles called lamellar bodies. Secretion of lamellar
bodies occurs by a process of exocytosis that is regulated by a number of physical and hormonal factors.
Stretch, the mode of ventilation, and the labor process enhance surfactant secretion and extracellular surfactant
pool sizes at birth. Catecholamines, purinoceptor agonists (e.g., adenosine triphosphate) that activate protein
kinases, and Ca2+ ionophores enhance phospholipid secretion by type II cells in vitro (9). Hyperglycemia and
hyperinsulinemia inhibit surfactant phospholipid secretion. Newly secreted surfactant enters the extracellular
space and undergoes dramatic structural reorganization to form tubular myelin, a process dependent on SP-A,
Ca2+, phospholipids, and SP-B. Phospholipids must move from tubular myelin to form monolayers and
multilayers at the air-liquid interface, thereby reducing collapsing forces in the alveoli.
Surfactant Recycling and Catabolism

The process of inflation and deflation produces spent forms of surfactant phospholipids that are taken up by type
II cells and reused or catabolized. Surfactant proteins B and C enhance the reuptake of phospholipids in vitro.
Surfactant phospholipid is recycled. In the adult rabbit lung, the half-life of surfactant phospholipids is
approximately 8 hours, and in newborn animals, the half-life is 3.5 days. The intracellular and extracellular pools
of surfactant are generally larger in the newborn animal than in adults. A relatively small fraction of the alveolar
surfactant pool is cleared by catabolism and alveolar macrophages, with most of the surfactant phospholipid
recycled or catabolized by type II epithelial cells. Recent studies support an important role for granulocyte/
macrophage-stimulating factors and their receptors in the mediation of surfactant clearance, acting primarily on
the alveolar macrophage. Defects in granulocyte-macrophage cerebrospinal fluid (GM-CSF) signaling, caused by
autoantibodies to GM-CSF or mutations in its receptor, cause alveolar macrophage dysfunction leading to
marked lipid accumulation in the postnatal lung, in turn, causing the syndrome of pulmonary alveolar
proteinosis. Exogenously administered surfactant is reused efficiently by adult and newborn lungs (12). The
effects of surfactant replacement therapy are therefore related both to the direct surface-tension-lowering
properties of surfactant introduced into the airway and to the recycling of the exogenously delivered
phospholipids by type II cells.
The Role of Surfactant in Lung Disease

Quantitative and qualitative abnormalities of pulmonary surfactant contribute to the pathogenesis of lung
disease in the newborn infant. In premature infants, deficiencies in surfactant production and secretion decrease
intracellular and extracellular pools of surfactant, leading to alveolar surfactant insufficiency and atelectasis.
Qualitative abnormalities of surfactant are also associated with many types of lung injury. Alveolar-capillary
leak, hemorrhage, pulmonary edema, and alveolar cell injury fill the alveolus with proteinaceous material that
inactivates surfactant. Serum and nonserum proteins, including albumin, fibrinogen, hemoglobin, and
meconium, are potent inactivators of pulmonary surfactant in vivo and in vitro; SP-A, SP-B, and SP-C act
synergistically to stabilize the surface properties of phospholipids in the presence of these inactivating proteins.
Inhibitory factors associated with surfactant dysfunction in acute lung injury can be overcome by the
administration of exogenous surfactants that contain the surfactant proteins.
Surfactant Replacement
The first successful surfactant replacement therapy in humans was reported by Fujiwara and colleagues in 1980
(13). Natural synthetic and semisynthetic surfactants have been successfully administered into the lungs of
premature infants for treatment of RDS, for the treatment of meconium aspiration, and are being tested for
therapy of other lung diseases. Surfactant replacement has become standard for prevention and treatment of
RDS. Animal surfactant preparations containing phospholipids, SP-B, and SP-C (e.g., Survanta, Curosurf,
Infasurf) and synthetic preparations
P.560
composed primarily of phospholipids mixed with spreading agents (e.g., Exosurf) are in clinical use (7,8),
although the animal based preparations are most widely used. The surfactant preparations containing surfactant
proteins provide highly surface active material to the alveolus. Surfactant replacement also contributes to the
pool size of surfactant phospholipids, providing substrate for surfactant synthesis by means of the recycling
pathways.
RESPIRATORY DISTRESS SYNDROME

Respiratory distress syndrome, previously called hyaline membrane disease, is a common cause of morbidity
and mortality associated with premature delivery. Respiratory distress syndrome is a developmental disorder
rather than a disease process per se, and it is usually associated with premature birth. The incidence and
severity of RDS generally increase with decreasing gestational age at birth and are usually worse in male
infants. Infants of diabetic mothers with poor metabolic control and infants born after fetal asphyxia,
maternofetal hemorrhage, or after pregnancies complicated by multiple births are at higher risk for RDS.
Respiratory distress syndrome affects approximately 20,000 to 30,000 infants each year in the United States
and complicates about 1% of pregnancies. Approximately 50% of the infants born between 26 and 28 weeks of
gestation develop RDS, whereas fewer than 20% to 30% of premature infants at 30 to 31 weeks have the
disorder.
Clinical Presentation
Infants with RDS present at birth or within several hours after birth with clinical signs of respiratory distress that
include tachypnea, grunting, retractions, and cyanosis (See Color Plate) accompanied by increasing oxygen
requirements. Physical findings include rales, poor air exchange, use of accessory muscles of breathing, nasal
flaring, and abnormal patterns of respiration that may be complicated by apnea. Chest radiographs are
characterized by atelectasis, air bronchograms, and diffuse reticular-granular infiltrates, often progressing to
severe bilateral opacity characterized by the term white-out (Fig. 29-6). Radiographic patterns in RDS are
variable and may not reflect the degree of respiratory compromise.
The infant attempts to maintain alveolar volume by prolonging and increasing expiratory pressures by breathing
against a partially closed glottis, causing the grunting noise characteristic of RDS, but often seen in other
respiratory disorders as well. Increasing oxygen requirements and the need for ventilatory support often occur
rapidly in the first 24 hours of life and continue for several days thereafter. The clinical course depends on the
severity of RDS and the size and maturity of the infant at birth. In uncomplicated RDS, typically seen in more
mature infants, recovery occurs over several days, and infants generally no longer require oxygen or ventilatory
support after the first week of life. The most premature infants are at greatest risk for severe RDS and
frequently develop complications, including central nervous system (CNS) hemorrhage, patent ductus arteriosus
(PDA), air leak, and infection, which contribute to prolonged requirements for oxygen and ventilatory support.
Figure 29-6 This premature infant presented with grunting, retractions, and cyanosis after delivery. The
diffuse reticular-granular opacification, air bronchograms, and decreased lung volumes in the chest radiograph
film indicate respiratory distress syndrome.
Pathology
Pathologic findings early in the course of RDS include atelectasis, pulmonary edema, pulmonary vascular
congestion, pulmonary hemorrhage, and evidence of direct injury to the respiratory epithelium (Fig. 29-7).
Epithelial cell injury is especially evident in the bronchiolar region of the lung. Histologic findings include the
presence of hyaline membranes, the characteristic eosinophilic material derived from bronchial and bronchiolar
injury to epithelial cells. Alveolar spaces are generally not inflated, and at autopsy, the lungs of infants with RDS
are often airless on passive deflation. Leukocytic infiltration is not observed early in the course of RDS unless
complicated by infection. Pulmonary edema, hemorrhage, and hemorrhagic edema are common pathologic
features in RDS, especially if the clinical course is further complicated by PDA and congestive heart failure.
Pathophysiology
Avery and Mead first demonstrated the paucity of alveolar surfactant in the lungs of infants dying of RDS (14).
Quantitative and qualitative abnormalities of the pulmonary surfactant system are critical to the pathogenesis of
RDS in premature infants. Lack of pulmonary surfactant leads to progressive atelectasis, loss of functional
residual
P.561
capacity, alterations in ventilation-perfusion ratio, and uneven distribution of ventilation. The RDS is further
complicated by the relatively weak respiratory muscles and the compliant chest wall of the premature infant,
which impair alveolar ventilation. Diminished oxygenation, cyanosis (See Color Plate), and respiratory and
metabolic acidosis contribute to increased pulmonary vascular resistance (PVR). Right-to-left shunting through
the ductus arteriosus, foramen ovale, and intrapulmonary ventilation-perfusion mismatch further exacerbate
hypoxemia.
Figure 29-7 Dilated air spaces, hyaline membranes, and extensive atelectasis are seen throughout the lung of
an infant born at 28 weeks of gestation with severe respiratory distress syndrome (Hematoxylin and eosin
stain; original magnification 250). (Courtesy of Edgar Ballard, Cincinnati Children's Hospital, Cincinnati, OH.)
Prevention
Although the incidence of premature birth in the United States (approximately 7%) has not changed significantly
in recent decades, the incidence of severe RDS has decreased at each gestational age as advances in maternal
care and strict attention to avoidance of asphyxia and infection at birth have become standard treatment.
Careful fetal monitoring, treatment of underlying maternal disorders, determination of amniotic fluid lamellar
body number or other biochemical indicators of fetal lung maturity, and administration of tocolytics and
maternal glucocorticoids have decreased the incidence of RDS. Although a single course of antenatal steroids
improves lung function and reduces the risk of newborn death, current evidence from animal and clinical studies
suggests that additional courses of steroids do not further improve lung function and are associated with risks of
adverse consequences (15). Surfactant replacement can further decrease the incidence and severity of RDS.
Rapid restoration of blood volume after hemorrhage and correction and avoidance of anemia, acidosis, and
hypothermia improve the clinical outcomes in RDS as well. Positive-pressure ventilation and continuous positive
airway pressure (CPAP) improve the course of severe RDS, but do not prevent the disease itself.
Treatment
Postnatal therapy of RDS begins with careful assessment and resuscitation. Adequate ventilation, oxygenation,
circulation, and temperature must be assured before the infant is transferred from the delivery room to the
appropriate site of care. Surfactant replacement therapy may be initiated at birth in infants at risk for RDS or
thereafter, as symptoms of RDS are established and the diagnosis of RDS is confirmed. Ventilatory management
of neonatal respiratory disorders has been reviewed and is detailed in Chapter 30 (16).
Adequacy of ventilation and oxygenation must be established as soon as possible to avoid pulmonary
vasoconstriction, further ventilation-perfusion abnormalities, and atelectasis. Positive-pressure ventilation, CPAP,
and oxygen therapy may be required at any time during the course of RDS and must be readily available to the
infant. Close monitoring of pH, oxygen saturation, partial pressure of CO2 (PCO2), and partial pressure of
oxygen (PO2) by transcutaneous monitors and by arterial catheterization or sampling of arterialized capillary
blood is critical in guiding mechanical ventilation and ambient oxygen requirements. Surfactant replacement
therapy is provided through the endotracheal tube and is often used several times during the early course of
RDS to maintain pulmonary function. Exogenous surfactants are given by intratracheal instillation of doses of
approximately 100 to 150 mg of phospholipid per 1 kg of body weight. Animal-derived surfactants cause a more
rapid improvement in oxygenation and lung compliance when compared to currently available synthetic
surfactants (17).
Mild or moderate RDS can be managed by CPAP applied by mask, nasal cannula, nasal prongs, or endotracheal
or nasopharyngeal tubes. In general, 3 to 6 cm of water (H2O) pressure is applied to the infant's airway.
Oxygenation and effort of breathing are usually rapidly improved by CPAP. Rapid fluctuations in blood gases may
occur, requiring careful monitoring of PCO2 and PO2. As forced inspiratory oxygen requirements decrease during
recovery, airway pressure is decreased, and the infant is weaned to head hood or nasal cannula oxygen. Apnea,
inadequacy of ventilation, atelectasis, mucous plugging, hyperaeration, or air leak may complicate the care of
infants with RDS.
P.562
Careful attention to the mechanical details of the application of CPAP or mechanical respirators is required.
Mandatory ventilation should be instituted well in advance of respiratory failure and severe respiratory acidosis
to avoid severe hypoxemia and atelectasis. Ventilation is maintained through an endotracheal tube, which can
be placed nasally or orally, for delivery of oxygen and positive pressure. Pressure-cycled ventilators are most
frequently used in the neonatal intensive care unit (NICU) and are controlled by setting positive inspiratory
pressure, rate, inspiratory-expiratory times, and positive end-expiratory pressures (PEEP). Volume-cycled
ventilators, in which fixed volumes are delivered to define the respiratory cycle, are used less frequently in the
newborn. As in all respiratory therapy, critical attention to adequacy of ventilation, as assessed by PO2, PCO2,
pH, and transcutaneous oxygen saturation, is required on an almost continual basis to adjust to the rapid
changes in respiratory status occurring in these critically ill infants. Barotrauma and oxygen toxicity to the lung
represent significant pulmonary complications in the therapy of RDS. Excesses in ventilation, peak or mean
airway pressure, and oxygen therapy should be avoided. Because hyperoxia is associated with retrolental
fibroplasia, a major cause of blindness in premature infants, arterial PO2 must be carefully monitored, generally
maintaining PO2 between 50 to 80 mm Hg. Other forms of ventilation such as high-frequency or jet ventilators
are often used in combination with exogenous surfactant for the treatment of RDS. These therapies are often
considered for treatment of severely affected infants whose ventilation has not been adequately supported by
conventional mandatory ventilation and surfactant therapy. Although some controlled trials indicate high
frequency ventilation may reduce the risk of chronic lung disease in preterm infants, this mode of therapy may
increase the usual adverse neurologic outcomes and should therefore be utilized with caution (18). Nitric oxide
(NO) has also been used successfully in the treatment of respiratory failure in term infants. However, for
premature infants, there has been no clear demonstration of improvement in any clinically relevant outcome in
the randomized trials conducted to date (19).
Complications
CNS hemorrhage, intraventricular hemorrhage (IVH), and PDA represent significant clinical problems affecting
the care of infants with RDS. Patent ductus arteriosus and subsequent congestive heart failure and pulmonary
edema further compromise respiratory function, decreasing pulmonary compliance and perhaps inactivating
pulmonary surfactant. Prompt diagnosis and medical or surgical treatment of PDA are indicated during the
treatment of RDS. Acute CNS hemorrhage is often associated with shock, pulmonary compromise, and
pulmonary hemorrhage. Fluctuations in respiratory status may contribute to IVH and can be minimized by
careful attention to respiratory care and by judicious use of sedation. Intravenous fluids and administration of
oral feedings must be adjusted carefully during acute and convalescent care of infants with RDS. Excessive fluid
administration impairs pulmonary function and increases the risk of PDA.
MECONIUM ASPIRATION SYNDROME

Meconium-stained amniotic fluid (MSAF) occurs in approximately 12% of live births. The cause, pathophysiology,
and treatment of MSAF and MAS have been recently reviewed (20).
Meconium first appears in the fetal ileum between 10 and 16 weeks of gestation as a viscous, green liquid
composed of gastrointestinal secretions, cellular debris, bile and pancreatic juice, mucus, blood, lanugo, vernix
and approximately 72% to 80% water. The dry weight composition consists primarily of mucopolysaccharides,
with less protein and lipid. Although meconium appears in the intestine very early in gestation, MSAF rarely
occurs before 38 weeks of gestation. Incidence of MSAF increases thereafter, and approximately 30% of
newborns have MSAF if born after 42 weeks of gestation. The increased incidence of MSAF with advancing
gestational age probably reflects the maturation of peristalsis in the fetal intestine. Motilin, an intestinal peptide
that stimulates contraction of the intestinal muscle, is in lower concentrations in the intestine of premature vs.
postterm infants. Umbilical cord motilin concentration is higher in infants who have passed meconium than in
infants with clear amniotic fluid. Intestinal parasympathetic innervation and myelination also increase
throughout gestation and may play a role in the increased incidence of passage of meconium in late gestation.
In utero passage of meconium is also associated with fetal asphyxia and decreased umbilical venous blood PO2
(Fig. 29-8). Experimentally, intestinal ischemia produces a transient period of hyperperistalsis and relaxation of
anal sphincter tone, leading to the passage of meconium. The fetal diving reflex, which shunts blood
preferentially to the brain and heart and away from the visceral organs during hypoxia may enhance intestinal
ischemia. The gasping respiratory efforts that accompany fetal asphyxia contribute to the actual entry of
meconium into the respiratory tract, resulting in MAS.
Meconium in amniotic fluid is a hallmark of fetal compromise and demands critical evaluation of fetal well-being.
However, most infants with MSAF do not have lower Activity, Pulse, Grimace, Appearance, and Respiration
(APGAR) scores, more acidosis, or clinical illness than infants born with clear amniotic fluid. Fetal heart rate
patterns do predict those infants at greatest risk of significant meconium aspiration syndrome and poor
outcome. When normal fetal heart rate patterns are observed in cases of MSAF, the neonatal outcome is
generally comparable to deliveries with clear amniotic fluid. In contrast, perinatal morbidity is increased in
newborns with abnormal fetal heart rates and MSAF. Neonatal outcomes of
P.563
deliveries complicated by MSAF associated with fetal tachycardia or decreased fetal heart rate variability are
similar to those of nonmeconium-stained infants with similar abnormalities of fetal heart rate. Those babies with
abnormal fetal tracings are at greatest risk for the meconium aspiration syndrome. However, in infants with a
normal fetal heart rate pattern, MSAF generally carries a low risk for perinatal morbidity.
Figure 29-8 Pathogenesis of meconium aspiration syndrome.
Meconium found below the vocal cords defines MAS. MAS occurs in approximately 35% of live births with MSAF
or in approximately 4% of all live births. Meconium aspiration syndrome describes a wide spectrum of
respiratory disease, ranging from mild respiratory distress to severe disease and death despite mechanical
ventilation. MAS typically presents as respiratory distress, tachypnea, prolonged expiratory phase and
hypoxemia soon after birth in an infant heavily stained on the nails, hair, and umbilical cord with meconium or
born through thick meconium. Infants with severe MAS often have an increased anteriorposterior dimension of
the thorax, a barrel chest, secondary to obstructive airway disease. Persistent pulmonary hypertension is also
frequently observed in infants with severe MAS. Less severe meconium aspiration, typically of nonparticulate
meconium, may present with the appearance of a pneumonitis with mild increased work of breathing or peaceful
tachypnea reaching a peak at one to three days and resolving slowly over the first week of life.
The chest radiographs of infants with MAS, especially when associated with thick particulate meconium, are
heterogeneous and demonstrate coarse infiltrates, with widespread consolidation and areas of hyperaeration
(Fig. 29-9). Pleural effusions are detected in approximately 30% of infants with MAS. There is an increased risk
of pneumothorax or pneumomediastinum, which occur in approximately 25% of severely affected infants. Chest
radiographs are abnormal in more than one-half of infants with meconium detected below the vocal cords,
although fewer than 50% of the infants with abnormal radiographs have significant respiratory distress. The
severity of chest radiographic abnormalities may not correlate well with the severity of clinical disease.
Figure 29-9 This full-term infant was born with fetal bradycardia and thick meconium in the amniotic fluid.
Cyanosis and respiratory distress were evident within minutes of delivery. The chest radiograph film
demonstrates coarse, irregular infiltrates, hyperinflation (left and right diaphragms at ribs 10 to 11) and right
pleural effusion indicative of meconium aspiration syndrome. Endotracheal and nasogastric tubes are in position.
Pathology
Postmortem examination of lungs from infants with severe MAS reveals meconium, vernix, fetal squamous cells
and cellular debris in the air spaces from the airways to the alveoli. An inflammatory response with
polymorphonuclear leukocytes, macrophages, and alveolar edema may be observed, but large quantities of
meconium may be present without histologic signs of inflammation. Hyaline membrane formation, pulmonary
hemorrhage, and necrosis of pulmonary microvasculature and parenchyma can occur. Platelet-rich microthrombi
in small arterioles and increased muscularization of distal arterioles have been described in some infants dying
of MAS.
Pathophysiology
The pulmonary abnormalities in MAS are related primarily to acute airway obstruction, decreased lung tissue
compliance, and parenchymal lung damage (see Fig. 29-8). Instillation of meconium into adult rabbit and
newborn dog tracheas causes acute mechanical obstruction of proximal and distal airways (21). A ball-valve
mechanism producing partial airway obstruction contributes to air trapping, which results in increased anteriorposterior chest diameter, increased expiratory lung resistance, and increased functional residual capacity.
Complete obstruction of small airways may result in regional atelectasis and
P.564
ventilation-perfusion inequalities. Disruption of surfactant function by serum and nonserum proteins and fatty
acids contributes to atelectasis, decreased compliance, and resulting hypoxia. Additionally, meconium may be
toxic to pulmonary epithelial cells and may itself contain, and stimulate, the production of pro-inflammatory
cytokines including IL-8 and TNF-.
In more than one-half of the infants with severe MAS, pulmonary hypertension with right-to-left shunting
contributes to the characteristically severe and sometimes refractory hypoxemia. Such clinical pulmonary
hypertension correlates with increased muscularization of distal pulmonary vessels pathologically and,
experimentally, with chronic intrauterine hypoxia. Perinatal asphyxia is a critical underlying factor in the
pathogenesis of MAS, increasing the risks for pulmonary hypertension and meconium aspiration.
Prevention
Before the late 1970s, it was thought that aspiration of amniotic fluid and meconium occurred during the first
few breaths after delivery. Therapy was aimed at preventing MAS at the time of delivery by DeLee suctioning of
the nasopharynx before delivery of the shoulders and before the first breath and, after delivery, immediate
intubation and suctioning of the trachea to limit aspiration of meconium from the oropharynx and trachea into
the distal lung. Mortality from MAS was decreased when the trachea was suctioned immediately after birth (22).
DeLee suctioning of the nasopharynx while the infant was at the perineum also decreased morbidity and
mortality from MAS. However, meconium aspiration syndrome continues to occur in infants who are adequately
suctioned in the delivery room. Aspiration of meconium or amniotic fluid in utero probably occurs in some infants
with MAS, particularly in those with perinatal asphyxia. Generally, fetal lung fluid flows outward from the lungs
into the amniotic sac. However, studies with radiopaque contrast and 51Cr-labeled erythrocytes injected into the
amniotic sac demonstrated that some amniotic fluid enters the fetal lung even in the nonasphyxiated human
fetus. Gasping associated with inhalation of amniotic fluid or meconium occurs in fetal lambs, rhesus monkeys,
and humans in response to fetal asphyxia induced by compression of the umbilical cord or maternal aorta. Fetal
gasping may be a critical factor in entry of meconium into the lung before birth. Antenatal diagnosis and
treatment of fetal asphyxia is therefore critical for prevention of MAS. Amnioinfusion, the infusion of saline into
the amniotic sac, decreases cord compression and dilutes meconium, potentially minimizing its toxicity after
aspiration (23). Despite clinical studies supporting the use of intrapartum amnioinfusion to decrease the rate of
emergency cesarean section and to decrease morbidity related to MAS, data regarding its efficacy have not
supported its widespread use in MSAF.
Based on the findings that meconium aspiration may occur in utero, is occasionally accompanied by asphyxia,
and that the benefit of immediate intubation in the delivery room of infants with MSAF may be out-weighed by
the risks of trauma to the airway, a more selective approach to intubation of neonates exposed to meconium in
utero is now recommended (24). The oropharynx and naso-pharynx of all meconium-exposed neonates should
be cleared on delivery of the head by a wall-mounted DeLee suction device. Immediate tracheal intubation and
suctioning is recommended only if the infant is depressed. The utilization of this type of protocol has decreased
the need for emergent intubation by up to 40% without an increase in incidence or severity of meconium
aspiration syndrome. Clearing the airway and establishing respiration and oxygenation remain however basic to
the resuscitation of all infants.
Treatment
Postnatal therapy for MAS begins with continuous observation and monitoring of infants at risk. Vigorous
treatment of nonrespiratory sequelae of neonatal asphyxia, including temperature instability, hypoglycemia,
hypocalcemia, hypotension, and decreased cardiac function are critical to promoting the fetal-to-newborn
physiologic transition. Attention also needs to be given to the potential effects of multi-organ hypoxemia and
ischemia, including reduced renal function, reduced liver production of clotting factors, hypoalbuminemia,
cerebral edema, and seizures. Pulmonary vasoconstriction is associated with MAS, and correction of hypoxemia
and acidosis is indicated. Chest physiotherapy and suctioning of particulate meconium may be useful if there is
airway obstruction and the infant maintains adequate oxygenation during such therapy. Exogenous surfactant
has been used successfully for the treatment of meconium aspiration, decreasing air leak and the need for
extracorporeal membrane oxygenation. Continuous monitoring of oxygenation by transcutaneous oxygen
monitoring or pulse oximetry and assessment of PaO2, PaCO2, and pH should be used to guide the application of
oxygen therapy and mechanical ventilation. The presence and severity of pulmonary hypertension should be
evaluated by echocardiography in the hypoxemic infant with meconium aspiration syndrome. Broad-spectrum
antibiotics are routinely used in the therapy of MAS in infants with abnormal radiographic findings and
respiratory distress, although their efficacy in MAS is unproven. Treatment of acute MAS with glucocorticoids has
not been beneficial.
Surfactant Treatment
Recent research suggests beneficial effects of surfactant replacement therapy, inhaled NO, and high-frequency
oscillatory ventilation for MAS. Meconium instilled into canine or piglet lungs or mixed with surfactant in vitro
inactivates surfactant function, decreasing lung compliance, lung volumes, and oxygenation. Surfactant
inactivation
P.565
can be overcome by addition of exogenous surfactant. Several studies suggest that surfactant therapy may
decrease respiratory failure associated with MAS (25). Findlay and associates (26) determined that surfactant
replacement (6 mL/kg, Survanta) improved oxygenation and reduced the incidence of air leaks and the severity
of pulmonary morbidity when begun within 6 hours after birth. Response to the first dose was moderate.
However, after the second and third doses, given 6 hours apart, improvements were documented in mean
arterial-to-alveolar PO2 ratio and in oxygenation index. Optimal dose, method, and timing of instillation of
surfactant in MAS remain to be determined.
Ventilatory Support
Mechanical ventilation is required in up to 30% of infants with severe MAS and must be managed carefully.
Although improvement in oxygenation was observed in patients with MAS treated with 4 to 7 cm H2O PEEP,
further studies to confirm the safety and efficacy of positive end-expiratory pressure in MAS are needed.
Continuous positive airway pressure or PEEP may aggravate hyperinflation associated with MAS and should be
used with caution. Pneumothorax or pneumomediastinum occur frequently during the course of MAS because of
the ball-valve effect of meconium and may occur before the application of positive-pressure ventilation.
Lengthening the expiratory time of the ventilatory cycle may minimize hyperinflation. Oscillation and highfrequency jet ventilation have been used in the treatment of MAS, but reports of their safety and efficacy in MAS
are conflicting.
High-frequency oscillatory ventilation (HFOV) has been studied alone and in combination with INO as treatment
for MAS. Kinsella and associates (27) found that the response rate of infants with MAS to HFOV plus INO was
greater than the response rate to either HFOV or INO with conventional ventilation alone. Extracorporeal
membrane oxygenation (ECMO) has been used successfully in rescue treatment of severe MAS refractory to
conventional ventilatory therapy. As of June 2003, more than 6,300 neonates with MAS were registered as
receiving ECMO in the Extracorporeal Life Support Organization (ELSO) database, with a 94% survival rate (28).
Surfactant therapy during ECMO reduced the perfusion time required and the overall rate of complications after
ECMO.
PERSISTENT PULMONARY HYPERTENSION OF THE NEWBORN

Gersony and colleagues (29) described hypoxemia in two infants with persistent physiologic characteristics of
the fetal circulation (PFC) in the absence of recognizable cardiac, pulmonary, hematologic, or central nervous
system disease. Because the placenta is no longer present and the ductus arteriosus may or may not be patent,
the term persistent pulmonary hypertension of the newborn (PPHN) is now used to describe this disorder. The
pathophysiology of PPHN is related to a failure to make the transition from high pulmonary vascular resistance
(PVR) and low pulmonary blood flow, characteristic of the fetus, to the relatively low PVR and high pulmonary
blood flow of the postnatal infant. The clinical syndrome of PPHN has been recently reviewed (30).
Pathophysiology
The pathophysiology of PPHN is best understood within the framework of the current knowledge of the
transitional circulation. Normal transition occurs in four phases: the in utero phase, the immediate phase
occurring in the first minutes after birth, the fast phase developing in the first 12 to 24 hours, and the final
phase, which requires days or months to complete.
In Utero Circulation
The in utero phase is characterized by PVR that exceeds systemic vascular resistance, resulting in right atrial
and ventricular pressures exceeding left atrial and ventricular pressures. As a result of this pressure differential,
more than one-third of the oxygenated blood returning from the placenta through the inferior vena cava streams
across the patent foramen ovale (PFO), is ejected from the left ventricle, and perfuses the head and neck
vessels and the lower body. Venous blood returning through the superior vena cava preferentially flows into the
right ventricle and main pulmonary artery. A small amount of this deoxygenated blood, comprising
approximately 8% of the cardiac output and with a PO2 less than 20 mm Hg, does perfuse the lungs, but
because of elevated PVR, most is shunted across the PDA to mix with the blood in the aorta distal to the cervical
and subclavian arteries. The lower body is therefore perfused with relatively less well oxygenated blood than the
head and neck. Because of the large right-to-left shunt at the PFO and the PDA, blood bypasses the lungs in
utero. Persistence of the elevated PVR after birth, without the benefit of placental oxygenation, results in the
profound hypoxemia characteristic of PPHN. The mechanisms that maintain the fetal state of high PVR are under
study. Pulmonary vasoconstriction induced by hypoxia, alterations in NO and arachidonic acid metabolism, and
systemic acidosis probably contribute to physiologic abnormalities in PPHN.
Immediate Phase
The second stage of normal transition, the immediate phase, is accomplished in the first minute after birth when
the fluid-filled fetal lungs are distended with air during the first breath. A rapid decrease in PVR occurs with the
mechanical distention of the pulmonary vascular bed, allowing more oxygenated blood to perfuse the lungs. The
entry of air into the alveoli improves the oxygenation of the pulmonary vascular bed, further decreasing PVR.
P.566
Fast Phase
The fast phase of the transitional circulation occurs for 12 to 24 hours after birth and accounts for the greatest
reduction in PVR. The drop in PVR has been associated with the production of vasodilators, such as prostacyclin
and endothelial-derived relaxing factor i.e., NO. Prostacyclin is produced in the neonatal lung in response to
rhythmic distension of the lungs. Pretreatment of the fetal lamb with cyclooxygenase inhibitor decreased
prostacyclin production and prevented the late fall in PVR. The role of cyclooxygenase and prostacyclin in the
transitional circulation may have clinical implications. Persistent pulmonary hypertension of the newborn has
been observed in infants of mothers receiving aspirin or nonsteroidal antiinflammatory agents that inhibit
cyclooxygenase activity. However, prostacyclin induction at birth is transient, and it does not account for
pulmonary vasodilation occurring in response to increasing oxygen tension. Likewise, indomethacin did not
reverse decreased PVR caused by hyperbaric oxygen. The role of pulmonary production of potent vasodilatory
leukotrienes, which also occurs during the initiation of ventilation, is unclear.
The pulmonary vasodilation and increase in pulmonary blood flow occurring in response to oxygenation can be
virtually masked by inhibitors of the endothelial-derived relaxing factor, NO. NO is induced by oxygen, adenosine
triphosphate (ATP), and sheer stress and is elevated in 1-day-old lamb pulmonary arteries and pulmonary veins
in comparison to near-term fetuses and few-week-old lambs. NO causes vasodilation by inducing the guanylate
cyclase enzyme. The resultant increase in cGMP in turn activates a kinase that decreases intracellular calcium,
allowing smooth muscle cell relaxation. The vascular effects of NO are specific and localized because of its great
affinity for hemoglobin, especially deoxyhemoglobin. Thus, INO can cause pulmonary vasodilation without
systemic hypotension. Prostaglandin, PGI2 in particular, and NO are believed to be the principal agents
responsible for the decrease in pulmonary vascular resistance in the fast phase of transition to air breathing.
Final Phase
The final phase of the neonatal pulmonary vascular transition involves remodeling of the pulmonary vascular
musculature (31). In the normal fetal and term lung, fully muscularized, thick-walled preacinar arteries extend
to the level of the terminal bronchioles. Intraacinar and alveolar wall arteries are not muscularized. Within days
after delivery, medial wall thickness of preacinar vessels smaller than 250 mm in diameter decreases, and within
months, medial wall thickness of vessels larger than 250 and smaller than 500 mm also decreases. Hypoxia at
birth prevents the remodeling and regression of the smooth muscle of the preacinar bronchiolar arteries. In
utero, or after birth, high-flow states and chronic hypoxia stimulate cells of the intraacinar and alveolar arteries
to differentiate into smooth muscle and connective tissue, resulting in abnormally thickened and reactive
arteriolar musculature. Distal extension of smooth muscle with increased numbers of adventitial fibroblasts and
extracellular matrix has been described in pulmonary arteries of infants dying of severe MAS with PPHN.
Etiology
Persistent pulmonary hypertension of the newborn has a variety of causes that can be classified by the
predominant abnormality involved (Table 29-1). Identification of the cause and subclass of PPHN is helpful in
predicting severity and reversibility of PPHN in the neonate. Assessment of the clinical severity of PPHN helps
determine the need for referral to nurseries with ECMO and NO capability.
Clinically, PPHN presents as labile hypoxemia that is often disproportionate to the extent of pulmonary
parenchymal disease. Infants with PPHN are commonly appropriate for gestational age and near term. The
perinatal history frequently includes factors associated with perinatal asphyxia. Clinical symptoms include
tachypnea, respiratory distress, and often rapidly progressive cyanosis (See Color Plate), particularly in response
to stimulation of the infant. The cardiovascular examination may be normal or may reveal a right ventricular
heave, closely split or single loud S2, and low-pitched systolic murmur of tricuspid regurgitation suggesting that
pulmonary arterial pressure is equal to or greater than systemic arterial pressure. A gradient of 10 mm Hg
between right arm and lower extremity oxygen pressures suggests right-to-left shunting at the ductus arteriosus
and is consistent with, although not required for, the diagnosis of PPHN. PPHN may occur without differential
oxygen saturations if the ductus arteriosus is closed and mixing of cyanotic and oxygenated blood is occurring
within the lungs or at other intracardiac sites. Differential diagnosis of PPHN includes severe pulmonary
parenchymal disease, such as severe MAS, RDS, pneumonia, or pulmonary hemorrhage, and congenital heart
disease, such as transposition of the great arteries. Critical pulmonic stenosis, hypoplastic left ventricle, or
severe coarctation should be considered in the differential diagnosis. Methods used to differentiate PPHN from
pulmonary parenchymal disease or cardiac disease are outlined in Table 29-2.
The oxygenation of infants with severe pulmonary parenchymal disease without PPHN generally improves after
treatment with oxygen or mechanical ventilation. Infants with PPHN often have little or no parenchymal lung
disease. They are easily ventilated but remain hypoxic despite high fraction of inspiratory oxygen (FiO2).
Oxygenation will frequently improve markedly with improved ventilation and/or correction of metabolic acidosis
in infants with PPHN. Cyanotic congenital heart disease (CCHD) is usually associated with fixed, structural
mixing of venous and arterial blood. In infants with CCHD,
P.567
hypoxemia is generally unresponsive to increased exogenous oxygen, mechanical ventilation, hyperventilation,
or alkalinization. Diagnosis of PPHN can be complicated by the coexistence of pulmonary hypertension,
parenchymal lung disease, or CCHD. Echocardiography is useful in the diagnosis of structural heart disease and
assessment of pulmonary hypertension in this clinical setting.

TABLE 29-1 CLASSIFICATION SYSTEM FOR PERSISTENT PULMONARY HYPERTENSION OF THE
NEWBORN
Proposed
Pathology
Associated Diseases
Functional vasoconstriction; Acute perinatal hypoxia

normal pulmonary vascular Acute meconium aspiration
development
Sepsis or pneumonia (especially
group B streptococci)
Mechanism
Prognosis
Response to acute Good; reversible

hypoxia, particularly
in the presence of
acidemia
Hypoventilation
a
CNS depression
Fixed decreased diameter;

abnormal extension and
hypertrophy of distal
Hypothermia
Hypoglycemia
Placental insufficiency
Prolonged gestation
In utero closure of ductus
Response to chronic Poor; fixed

hypoxia
structural lesion
Excessive
pulmonary blood
pulmonary vascular smooth arteriosus Aspirin

muscle
Nonsteroidal antiinflammatory
flow in utero
agents
Single ventricle without pulmonic
stenosis
Chronic pulmonary venous
Elevated pulmonary
hypertension
venous pressure
a
TAPVR
Obstructive left-sided heart
lesions
Decreased cross-sectional
area of the pulmonary
vascular bed
Idiopathic diseases
Space-occupying lesions
Diaphragmatic hernia
Lung dysgenesis
Pleural effusions
Hypoplasia of
alveoli and
associated vessels
Poor; fixed
structural lesion
Increased blood
viscosity
Good, unless
chronic
Congenital lung hypoplasia

Potter syndrome
Functional obstruction to
pulmonary blood flow
a
Thoracic dystrophies
Polycythemia
Hyperfibrinogenemia
CNS, central nervous system; TAPVR, total anomalous pulmonary venous return.
Therapy
Supportive medical management includes correction of underlying abnormalities that may include shock,
polycythemia, hypoglycemia, hypothermia, diaphragmatic hernia, or CCHD. Metabolic acidosis and hypotension
should be corrected.
Specific therapy for PPHN is aimed at increasing pulmonary blood flow, decreasing right-to-left shunting and
reducing ventilation/perfusion (V/Q) mismatch. High ambient oxygen and mechanical ventilation are the
pulmonary therapeutic interventions for treatment of PPHN. Ligation of the PDA is not useful, and it may be
detrimental. Cardiac failure may occur after PDA ligation as the right ventricle fails in the face of high pulmonary
resistance without the safety valve of the patent ductus. Shunting between the pulmonary and systemic
circulations, such as through the ductus arteriosus or patent foramen ovale, depends on the relative pressures
of each system. Therefore, optimal therapy decreases pulmonary artery pressure although increasing or not
changing systemic arterial pressure and cardiac output.
Infants with severe PPHN are often sensitive to activity and agitation. Stimulation should be minimized during
the care of these infants. Transcutaneous and intravascular monitoring equipment and temporal clustering of
interventions reduce agitation. Muscle relaxants (e.g., pancuronium) and sedatives are frequently beneficial but
should be used with caution. Paralysis may further compromise ventilation and may mask clinical signs of
respiratory insufficiency. Sedatives should be chosen to minimize cardiovascular side effects such as systemic
hypotension. Infants with PPHN, especially if asphyxiated or septic, frequently develop systemic hypotension and
signs of cardiac failure. The hematocrit should be maintained at or above 45%, and volume expansion may be
used to support the circulation. Elevated right heart pressure with increased PVR, poor venous return secondary
to high intrathoracic pressures during mechanical ventilation, and predisposing asphyxia may contribute to
myocardial dysfunction that may be responsive to dobutamine. Adrenergic pressors are commonly used for
refractory hypotension but should be
P.568
used with caution, as they may contribute to pulmonary vasoconstriction. There is anecdotal evidence
supporting efficacy of hydrocortisone, 10 mg/kg/dose, for the treatment of refractory hypotension in the
newborn with PPHN and right-to-left intravascular shunt.
TABLE 29-2 DIAGNOSTIC EVALUATION OF SEVERE NEONATAL HYPOXEMIA
Test
Hyperoxia
Method
Result
Expose to 100% FiO2 for 5- PaO2 increases to >100
Pulmonary
10 min
parenchymal disease
Persistent
pulmonary
hypertension or
cyanotic congenital
heart disease
Pulmonary
parenchymal disease
mm Hg
PaO2 increases to <20
mm Hg
Hyperventilation-hyperoxia Mechanical ventilation with PaO2 increases to >100

100% FiO2 and respiratory mm Hg without
rate 100-150 breaths/min hyperventilation
PaO2 increases at a
critical PCO2, often to
<25 mm Hg
No increase in PaO2
despite hyperventilation
Simultaneous-preductalpostductal PO2
Echocardiography
Suggested Diagnosis
Persistent
pulmonary
hypertension
Cyanotic congenital
heart disease or
severe, fixed
pulmonary
hypertension
Patent ductus
arteriosus with rightto-left shunt
Compare PO2 of right arm
Preductal PO2 15 +
or shoulder to that of lower

abdomen or extremities
M-mode
Venous contrast injection
Two-dimensional
postductal PO2
Increased RVPEP and
RVET
Simultaneously appears
Right ventricular
systolic time interval
ratio (RVSTI =
echocardiography
Doppler
in PA and LA
Deviation of intraatrial
RVPEP/RVET > 0.5)

predicts PPHN
septum to left; rule out

congenital heart defect
Failure of acceleration of
systolic blood flow
between large main
pulmonary artery and
small peripheral
Patent foramen ovale

Increased
pulmonary arterial
pressure
Suggests right-toleft PDA or
intracardiac shunt
pulmonary artery
a
LA, left atrium; PA, pulmonary artery; PDA, patent ductus arteriosus; PPHN, persistent pulmonary
hypertension of the newborn; RVET, right ventricular ejection time; RVPEP, right ventricular ejection
period.
Respiratory and Metabolic Alkalosis

The often dramatic response of infants with PPHN to respiratory or metabolic alkalosis has lead to their use in
the care of infants with severe PPHN, however the sometimes transient benefit of alkalosis must be balanced
against the risk of secondary lung injury, neurologic compromise from hypocapnic alkalosis and cellular effects
of infusion of hypertonic solutions. The degree of pulmonary parenchymal disease and risk of barotrauma may
affect the clinical choice of inducing respiratory or metabolic alkalosis. Although it may be necessary to
transiently raise arterial pH with increased ventilation and base to reverse severe pulmonary vasoconstriction,
because of pulmonary and neurologic concerns, sustained hypocapnia and metabolic acidosis are not advocated.
INO and normalization of acid-base and ventilatory status are preferred. Excessive mechanical ventilation with
overdistension of the lung may increase right-to-left shunting. Pulmonary barotrauma associated with
aggressive ventilation should be avoided. Hypocarbic alkalosis, by shifting the hemoglobin-oxygen dissociation
curve, may also compromise the release of oxygen at the tissue level. Hyperoxia and hypocarbia may adversely
affect cerebral blood flow. When used, weaning from ventilation NO and oxygen must proceed with caution,
because dramatic liability of PO2 is often observed in infants with PPHN.
Nitric Oxide and High-Frequency Oscillatory Ventilation

Pharmacotherapy of PPHN was recently reviewed (32). Various vasodilators such as tolazoline and
prostaglandins D2 and E1 have been studied for treatment of PPHN (32). At doses required to decrease PVR,
these agents often cause undesirable systemic vasodilation and hypotension. Tolazoline was previously used for
therapy for PPHN but has not been shown to improve outcome.
At low pharmacologic doses, <40 parts per million, INO specifically dilates the pulmonary vasculature. Avid
binding to hemoglobin prevents INO from dilating systemic blood vessels. Additionally, INO preferentially
vasodilates vessels of alveoli that are patent, thus improving V/Q matching. Other less specific vasodilators may
P.569
increase blood flow to atelectatic alveoli, increasing V/Q mismatch. Inhaled NO combined with conventional
ventilation increased oxygenation and decreased the oxygen index of approximately 30% of infants with PPHN.
The likelihood of response to INO appeared inversely related to the severity of parenchymal disease (27). In
patients with moderate PPHN, (alveolar-arterial oxygenation gradient, A-aDO2 = 500-599), 15% of those
treated with INO, vs. 58% of controls, progressed to severe PPHN (33). Approximately 25% of PPHN infants
failing to respond to INO responded to INO combined with HFOV, suggesting improved response to INO if
ventilation strategies are optimized to recruit atelectatic alveoli. The combined use of exogenous surfactant,
HFOV and INO to coordinately enhance alveolar recruitment and perfusion appears to provide at least additive if
not synergistic effects, especially in RDS and meconium aspiration cases. A systematic review of twelve
randomized controlled trials in term or near-full term infants with hypoxic respiratory failure, concluded that INO
reduced the need for ECMO. Approximately 50% of infants demonstrated improved oxygenation with INO
treatment, those with diaphragmatic hernia were least likely to respond (34). No adverse effects of INO were
reported.
Extracorporeal Membrane Oxygenation

Extracorporeal membrane oxygenation has been useful for the treatment of severe PPHN refractory to medical
management. The first newborn survivor of ECMO therapy was reported by Bartlett and colleagues in 1975 (35).
In 1999, the National Registry of Neonatal ECMO, and in 2003, the International Summary data of the ELSO
Registry reported an 80% survival rate for the infants with the primary diagnosis of PPHN who were treated with
ECMO (28,36). Most of these infants met criteria for greater than 80% risk of mortality with conventional
medical management. However, criteria used to determine risk of mortality vary from institution to institution
and frequently are based on retrospective chart reviews, reflecting older methods of medical management. It is
impossible to determine how these infants would have done with modern conventional therapy.
Long-Term Outcome
Most infants treated for PPHN have few residual respiratory symptoms, neurologic or developmental sequelae by
1 year of age (37). Of infants with more severe parenchymal disease, qualifying for INO or ECMO, approximately
25% have persistent BPD or recurrent reactive airway disease at 1 and 2 years of age. Of 133 children with
moderately severe persistent PPHN, with oxygenation index of 24 9 at study entry, randomized to receive INO
or placebo, approximately 13% had major neurologic abnormalities, 30% had cognitive delays, and 19% had
hearing loss (38). There was no difference between INO treated and control infants. Particularly, infants with
severe MAS or congenital diaphragmatic hernia and PPHN have an increased risk for chronic pulmonary sequelae
(39). Continued oxygen therapy, bronchodilators, diuretics, and enhanced nutrition may be necessary to treat
residual disease and establish adequate growth. Hearing, vision, and neurologic development should be followed
closely in infants treated for PPHN, especially if severely asphyxiated. Approximately 25% of infants treated with
INO or ECMO for PPHN remain below 5% for weight at 1 to 2 years of age. Approximately 10% to 12% are
diagnosed with severe neurodevelopmental disability. The risk of neurologic, growth, and pulmonary sequelae is
greatest in infants with PPHN secondary to congenital diaphragmatic hernia (39).
Pneumonia
Pneumonia remains a significant cause of morbidity and mortality for preterm and term infants. The incidence of
pneumonia in NICU patients exceeds 10% (40), with mortality of perinatally acquired pneumonia varying
between 4% and 20% (40,41). Pneumonia may be acquired transplacentally, during the birth process, or
postnatally, and it is caused by a variety of pathogens, including viruses, bacteria, and fungi (Table 29-3).
Unique environmental and host factors predispose the neonate to pulmonary infections. The increased
susceptibility of neonates for pneumonia may be related to immaturity of mucociliary clearance, small size of the
conducting airways, and lowered host defenses. Invasive procedures, such as tracheal intubation, barotrauma,
and hyperoxic damage to the respiratory tract may further impair resistance to pneumonia. The nosocomial flora
of the hospital nursery, whether derived from nursery equipment or the unwashed hands of caregivers, are
important vectors of pathogenic organisms.
Transplacental Viral Pneumonias

Pneumonia acquired through the transplacental route is most commonly of viral origin. Rubella, varicella-zoster,
cytomegalovirus (CMV), herpes simplex virus (HSV), and human immunodeficiency virus (HIV) are acquired by
this route. Transplacentally acquired pneumonitis is also associated with adenoviral, enteroviral, and influenza
viral infections. Viral pneumonia is usually part of a systemic illness, reflecting hematogenous spread from the
mother. Severity and onset of respiratory symptoms varies from respiratory failure at delivery to chronic
pneumonia evolving months after birth.
Fetal infection may result from a primary maternal infection acquired during pregnancy or from reactivation of a
latent infection. In utero transmission of rubella occurs as a result of primary infection acquired during
pregnancy. Transmission of varicella-zoster, HSV, and CMV occurs as a result of primary or recurrent maternal
infection. The timing of maternal infection in relation to birth is often a critical factor in outcome. Congenital
varicella typically develops when primary maternal chickenpox
P.570
occurs within the 21 days preceding parturition. If maternal varicella occurs less than 5 days before birth,
antepartum transfer of maternal antibody to the fetus is minimal, increasing the risk of systemic varicella
infection in the neonate. Varicella pneumonia usually presents 2 to 4 days after the onset of the exanthem.
Although varicella pneumonia is often self-limited, it can cause significant mortality and morbidity. Treatment
with varicella-zoster immune globulin within 72 hours of birth improves clinical outcome in neonates exposed to
varicella. Pneumonitis is not a common presentation in congenital CMV or herpes, but it is more common in
perinatally acquired CMV and HSV. Pulmonary disease caused by transplacental transfer of the HIV virus
generally presents after the neonatal period.
TABLE 29-3 PRIMARY PATHOGENS OF NEONATAL PNEUMONIA
Vector
Viruses
Bacteria
Other Agents
Transplacental
Rubella
Varicella-zoster
HIV
CMV
HSV
L. monocytogenes
M. tuberculosis
T. pallidium
Perinatal
HSV
CMV
Postnatal
CMV
HSV
Community based (i.e., RSV,
influenza, parainfluenza)
Group B streptococci
C. trachomatis
Gram-negative enteric (i.e., E. coli, U. urealyticum
Klebsiella)
S. aureus
C. albicans
P. aeruginosa
Flavobacterium
S. marcescens
HIV, human immunodeficiency virus; CMV, cytomegalovirus; HSV, herpes simplex virus; RSV,
respiratory syncytial virus.
Transplacental Bacterial Pneumonias

Transplacental bacterial infections are less common causes of pneumonia. Listeria monocytogenes,
Mycobacterium tuberculosis, and Treponema palladium are the most common organisms. Maternal listeriosis
classically presents with a flu-like syndrome, with fever and chills occurring up to 2 weeks before delivery.
Preterm labor and meconium staining of amniotic fluid, even in preterm infants, are common. Early onset
listeriosis generally presents soon after birth with respiratory distress and pneumonia. Radiographic findings are
nonspecific, consisting of peribronchial or widespread infiltrates. Congenital tuberculosis occurs most commonly
in infants born to women with primary infections. Respiratory symptoms in neonatal tuberculosis typically
present at 2 to 4 weeks of age. Transplacental transfer of T. palladium occurs most commonly during primary or
secondary maternal infection, usually after 20 weeks of gestation. Pneumonia alba is relatively uncommon in
congenital syphilis and refers to the pale, firm, and enlarged lungs seen at autopsy. Although the rate of
congenital syphilis has declined from its peak in 1991, clinicians must remain aware of all manifestations of
congenital syphilis.
Pneumonia Acquired in the Perinatal Period

Neonatal pneumonia is most commonly acquired during the process of labor and delivery. Infection occurs from
organisms ascending from the genital tract after rupture of fetal membranes or acquired during passage of the
infant through the birth canal. Respiratory symptoms are often present at delivery or have their onset in the first
few days of life. Despite the abundance and heterogeneity of organisms in the genital tract, only a few
commonly cause pneumonia. These organisms parallel pathogens causing early onset of sepsis.
In U.S. nurseries, group B Streptococcus (GBS) remains the most frequently identified organism causing
neonatal pneumonia and sepsis (42). The second most common group of organisms to produce early onset
sepsis or pneumonia are the gram-negative enteric bacilli: Escherichia coli, Klebsiella, Enterobacter, and Proteus
species. For term infants, as the rate of GBS disease has declined with the introduction of intrapartum antibiotic
prophylaxis, the rate of disease caused by other organisms has been relatively stable (42). In very low birth
weight (VLBW) infants, as the rate of Group B streptococcal disease has declined, there has been an increase in
E. coli, the majority of which is ampicillin resistant (42,43 and 44). Herpes simplex and CMV are the most
common viral agents causing early onset pneumonia. Pneumonia caused by Chlamydia trachomatis usually
begins at 2 to 8 weeks of age with upper respiratory tract symptoms, a staccato cough, and apnea. Antecedent
conjunctival infection is common but not always observed. Interstitial pneumonitis and hyperinflation are
associated with chlamydial pneumonia. U. urealyticum is a common inhabitant of the lower genital tract of
women and is frequently associated with histologic evidence of chorioamnionitis. U. urealyticum is a cause of
acute congenital
P.571
pneumonia and has also been associated with chronic lung disease in infants.
Pneumonia Acquired in the Postnatal Period
Newborns exposed to respiratory equipment or humidified incubators are at risk for respiratory infection by
Pseudomonas species, Flavobacterium, Klebsiella, or Serratia marcescens. Direct contamination by the hands of
caretakers as a result of inadequate hand washing is associated with outbreaks of Staphylococcus aureus and
gram-negative enteric organisms. Cytomegalovirus that is acquired postnatally through blood products or breast
milk commonly presents as a pneumonitis. With advances in transfusion technology, acquisition through blood
products is rare. Cytomegalovirus is shed intermittently in human milk, and can be transmitted to the infant. In
term infants there is typically no resulting clinical illness. In preterm infants clinical symptoms have been
reported, including development of systemic disease (45).
Neonatal HSV infection is most often associated with HSV type II. However, data from the National Institute of
Allergy and Infectious Disease indicate that 27% of symptomatic neonatal HSV infections were caused by HSV
type I (46). Postnatal infection from HSV generally occurs from orolabial, oropharyngeal, or breast lesions.
Community-based respiratory pathogens, including respiratory syncyticial virus, influenza, parainfluenza, and
enteroviruses, occur in the nursery. Pneumonia resulting from epidemic outbreaks of various enteroviral agents,
including echovirus 22 and coxsackievirus type B, is often associated with other clinical manifestation of
enteroviral disease. Risk factors for nosocomial fungal infections include VLBW, prolonged antibiotic therapy,
intubation, central line catheter placement, intravenous alimentation, and corticosteroids. Pneumonia caused by
Candida albicans usually presents in the context of disseminated disease. Mycoplasma species may also cause
pneumonia in the postnatal period.
Figure 29-10 Acute neutrophilic response with atelectasis and hyaline membranes (arrows) are seen in lung
tissue from a full-term infant who died at 2 days of age of group B streptococcal pneumonia. (Hematoxylin and
eosin stain; original magnification 200) (Courtesy of Edgar Ballard, Cincinnati Children's Hospital, Cincinnati,
OH.)
Pathologic Findings
Three common histopathologic patterns have been associated with neonatal pneumonia: hyaline membrane
formation, suppurative inflammation, and interstitial pneumonitis. Hyaline membrane formation is a nonspecific
response seen in lung injury associated with surfactant deficiency, pneumonia, and oxygen therapy. Damage to
the alveolar epithelium results in cell necrosis and leakage of cell and serum proteins into the alveolar space.
Hyaline membranes in neonatal pneumonia are often observed after GBS infection, but they are also associated
with fatal pneumonia caused by H. influenzae, gram-negative enteric organisms, and viral agents. Bacteria are
commonly seen within the hyaline membranes (Fig. 29-10). Disruption of alveolar capillary permeability and cell
injury results in leakage of proteins into the alveolus that further inactivate pulmonary surfactant, leading to
atelectasis. The decreased compliance, atelectasis, and hypoxemia seen in pneumonia are often
indistinguishable from findings in surfactant-deficient lungs in premature infants. The chest radiographic findings
in RDS and neonatal pneumonia may be identical, although bronchopneumonia and pleural effusions are more
common in GBS and other bacterial causes of neonatal pneumonia than in RDS.
Suppurative Pneumonia
Staphylococcus aureus, enteric bacilli such as Klebsiella pneumoniae, E. coli, and Pseudomonas species, and
fungi can cause
P.572
suppurative pneumonia. An intense inflammatory response often occurs in the lungs during these bacterial
infections. Necrosis of lung parenchyma, microabscess formation, and partial obstruction of terminal bronchioles
results in thin-walled, air-filled pneumatoceles. Spontaneous rupture of these structures can produce
pneumothorax. Microabscesses may consolidate into larger cavities or rupture to the pleural space, causing
emphysema. Pneumonia may be focal or may consolidate to produce large confluent abscesses. Perfusion of
consolidated lung tissue causes venous admixture and hypoxemia.
Interstitial Pneumonitis
Interstitial pneumonitis is typically caused by a virus and characterized by interstitial inflammation, edema,
mononuclear infiltration, and septal hyperplasia. Alveolar spaces may remain uninvolved, but in severe cases, a
serous exudate containing desquamated pneumocytes and macrophages may be associated with hyaline
membrane formation. Septal wall necrosis may occur, adding a component of hemorrhage to the inflammatory
exudate. Alveolar capillary block associated with the inflammation may impair respiratory function. CMV, HSV,
varicella-zoster, rubella, HIV, enteroviruses, and the community-based pathogens, such as respiratory syncytial,
influenza and parainfluenza viruses, are commonly associated with interstitial pneumonitis.
Group B Streptococcal Pneumonia

During the 1970s, GBS emerged as the predominant pathogen causing neonatal sepsis and pneumonia.
Colonization rates in pregnant women, because that time, have remained relatively stable at 10% to 30%
(41,47). Infants become infected by aspiration of contaminated amniotic fluid or acquisition during passage
through the birth canal. Vertical transmission rates for untreated GBS is approximately 50% (46). Invasive
disease is much less common, developing in 1% to 3% of colonized infants. Factors increasing the risk for
infection include maternal group B streptococcal bacteria, prolonged rupture of membranes (>18 hours),
maternal intrapartum temperature (>100.4), chorioamnionitis, and prematurity (41). Mortality from invasive
disease has declined from 50% to approximately 4% secondary to advances in neonatal care in the 1990s (48).
Intrapartum chemoprophylaxis has been shown to be effective in preventing early onset of GBS disease. With
the introduction of intrapartum antibiotic prophylaxis, the incidence of early onset disease has decreased by
70% from 18 cases per 1,000 live births to 0.5 cases per 1,000 live births (41,49). In 2002, the Center for
Disease Control and Prevention recommended universal screening of all pregnant women at 35 to 37 weeks'
gestation, and intrapartum antibiotics for all GBS colonized women (41).
Figure 29-11 A full-term infant of an uncomplicated pregnancy developed respiratory distress, cyanosis, and
periods of apnea within 6 hours of life. The blood culture and urine latex particle agglutination assay were
positive for group B streptococci. Diffuse reticulogranular pattern, air bronchograms, and right pleural effusion
without significant volume loss are consistent with common radiologic features of group B streptococcal
pneumonia.
Early onset GBS disease presents within the first week of life. Septicemia (30%-40%), meningitis (20%-30%),
and pneumonia (30%-40%) are the most common presentations. Exposure to intrapartum antibiotics has not
changed the clinical presentation of illness with up to 95% of infants presenting within 24 hours and 90% of
affected infants present with respiratory distress (49). Radiographic features of GBS infection may be
indistinguishable from RDS, although pleural effusions may help differentiate GBS from RDS (Fig. 29-11). In twothirds of affected infants, increased vascular markings or patchy infiltrates are observed on the initial chest
radiographs. Respiratory distress in the absence of radiographic abnormalities may be associated with
pulmonary vascular hypertension and hypoxemia. Late-onset GBS usually presents from 1 to 6 weeks after birth
and is commonly associated with meningitis.
Respiratory failure in GBS pneumonia results from hyaline membrane formation, atelectasis, and pulmonary
hypertension. Pulmonary hypertension is proposed to be mediated by high-molecular-weight polysaccharide
exotoxin. In animals, infusion of GBS exotoxin results in an initial increase in pulmonary vascular pressures and
fever, followed by a second phase characterized by granulocytopenia, granulocyte trapping in the lung, and
increased pulmonary vascular permeability (50).
Isolation of GBS from cultures of blood, cerebrospinal fluid, or suppurative foci is diagnostic of GBS infection.
Surface cultures of skin or mucous membranes are not useful because of the large number of infants colonized
but not infected. Antimicrobial therapy is usually instituted before an organism is identified and consists of a
penicillin and an aminoglycoside. After the organism is identified and meningitis is excluded, therapy can
continue with penicillin alone at 200,000 U/kg/day, usually
P.573
for 10 to 14 days. Extensive supportive care, including oxygen, mechanical ventilation, and cardiovascular
support, may also be required in treating an overwhelming infection.
Herpes Simplex Pneumonia

The incidence of neonatal HSV is increasing in the United States, affecting 1,500 to 2,200 infants per year.
Classification of HSV infection is based on the site and extent of involvement; HSV may be classified as a
cutaneous infection; as encephalitic, with or without cutaneous infection; and as a disseminated infection.
Mortality is highest among infants with disseminated disease. The mortality rate of infants with HSV pneumonitis
is approximately 79% (46,51). Transmission of HSV infection to the neonate is most frequently related to direct
contact with virus at delivery. Most women delivering infants with symptomatic HSV disease shed virus
asymptomatically at delivery, and fewer than 20% of these women reported a history of previous infection (51).
Whether the maternal infection is primary or recurrent is important in the pathogenesis of neonatal HSV
infection. Infants of women with serologic evidence of a recent primary infection are more likely to develop
disease than infants of women with recurrent herpes. Primary infections are associated with higher titers of viral
shedding and with low levels of protective antibodies, placing the neonate at higher risk. Rupture of the amniotic
sac for longer than 6 hours is associated with increased risk of neonatal HSV infection, although infection after
cesarean section with intact membranes has also been reported.
Herpes simplex virus pneumonitis usually presents as part of a disseminated HSV infection. Many severely
involved infants have no skin lesions. Infants typically develop signs of infection at 4 to 5 days of life. Fever,
tachypnea, jaundice, and irritability, progressing to respiratory failure, shock, and disseminated intravascular
coagulation are commonly seen. Diffuse interstitial pneumonitis and hemorrhagic pneumonitis are characteristic
of HSV infection of the lung.
Diagnosis of HSV may be made by identification of HSV DNA through PCR or viral isolation by culture. Viral
cultures should be obtained from skin lesions, cerebrospinal fluid, stool, urine, throat, nasopharynx, and
conjunctiva. HSV deoxyribonucleic acid (DNA) may be identified in cerebrospinal fluid (CSF), aspirated skin
vesicles, or blood. Hepatic dysfunction, neutropenia, bleeding diathesis, and interstitial pneumonitis are
commonly associated with HSV but are not diagnostic. Serologic assays are not of clinical value, because they
do not differentiate between antibodies to HSV types 1 and 2 or between maternal IgG and endogenously
produced antibodies. Intranuclear inclusions and multinucleated giant cells observed in scrapings from
cutaneous lesions are supportive but not diagnostic of HSV infection.
Early antiviral therapy decreases the progression from localized, cutaneous HSV infection to disseminated
disease, and decreases the mortality from HSV infection. High-dose parental acyclovir is now widely used in
treatment of HSV infection in the neonate and is currently considered the treatment of choice (52).
Transient Tachypnea of the Newborn

TTN was first described in 1966 by Avery and colleagues in a group of eight patients, seven of whom were
delivered vaginally at or near term (53). All of the infants presented at or shortly after birth with grunting,
retractions, and an increased respiratory rate. Respiratory rates of the original infants ranged from 80 to 140
breaths/min, and the symptoms persisted for 2 to 5 days. These infants could be differentiated from infants with
other acute lung diseases by their clinical course and radiographic findings. Although the precise cause of TTN
remains unknown, it was originally postulated that infants with TTN had reduced lung compliance because of
delayed resorption of lung fluid at the time of birth. Many clinicians support the original proposal of Avery and
colleagues that TTN results from distention of interstitial spaces by fluid, leading to alveolar air trapping and
decreased lung compliance. Because the original description, others have postulated that TTN may result from
mild immaturity of the surfactant system. Lack of phosphatidylglycerol in amniotic fluid samples obtained from
infants with TTN supports the latter concept.
The incidence of TTN is approximately 11 per 1,000 live births. The risk factors for TTN include prematurity,
maternal sedation, maternal fluid administration, maternal asthma, exposure to b-mimetic agents, and fetal
asphyxia. Delivery by cesarean section also predisposes the term infant to TTN (54). Children with TTN as
neonates are also more likely to develop asthma during ages zero to 4 (55).
Infants with TTN initially present with grunting, retractions, and increased respiratory rate. The symptoms of
tachypnea may persist for several days, and most infants require less than 40% oxygen to maintain adequate
systemic oxygenation. The radiographic findings in this disorder are ill defined but include increased central
vascular markings, hyperaeration, evidence of interstitial and pleural fluid, prominent interlobar fissures, and
cardiomegaly (Fig. 29-12). Furosemide has been utilized to treat TTN, but a recent meta-analysis failed to
demonstrate any benefit of this therapy (56). Because TTN is self-limited, no specific therapy is indicated,
although adequate ventilation and oxygenation must be maintained. Because the symptoms of TTN are
nonspecific and consistent with neonatal sepsis or pneumonia, most infants with TTN are evaluated for infection
and are treated with broad-spectrum antibiotics pending a definitive diagnosis.
Pulmonary Hemorrhage
Pulmonary hemorrhage in the newborn may vary from a focal, self-limited disorder to massive, lethal
hemorrhage.
P.574
The incidence of pulmonary hemorrhage in the neonatal period ranges from 0.8 to 1.2 per 1,000 live births,
although the incidence was 74% of all liveborn infants in one autopsy series of 70 newborns and is as high as
10% in babies less than 30 weeks gestation (57). Asphyxia, prematurity, intrauterine growth retardation,
infection, hypothermia, oxygen therapy, severe Rh hemolytic disease, and coagulopathy are associated risk
factors. In some studies, surfactant therapy has been associated with an increased incidence of pulmonary
hemorrhage, although this remains controversial. Although disseminated intravascular coagulation may precede
pulmonary hemorrhage, most infants with pulmonary hemorrhage do not have a coagulopathy. Pulmonary
hemorrhage generally presents within the first week of life, and the mortality rate after pulmonary hemorrhage
is estimated to be 75% to 90%. Although most infants who develop pulmonary hemorrhage have the
predisposing factors of extreme prematurity and underlying asphyxia and stress, there are a few case reports
describing previously healthy, term infants with pulmonary hemorrhage associated with an inborn error of the
urea cycle and elevated blood ammonia.
Figure 29-12 This full-term infant was born by cesarean section and developed tachypnea and grunting that
resolved 48 hours after birth. Perihilar vascular densities, streaky opacities of interstitial edema, fluid in the
interlobar fissures, small pleural effusions, and cardiomegaly are observed on the radiograph. These features
are indicative of transient tachypnea of the newborn.
Clinical Findings
The observation that the hematocrit of lung effluent in pulmonary hemorrhage is lower than the hematocrit of
blood supports the concept that most of these infants have hemorrhagic pulmonary edema. Neonatal pulmonary
hemorrhage is therefore thought to result from shock, hypoxia, and acidosis, which lead to left ventricular failure
and increased pulmonary capillary pressure with subsequent hemorrhagic pulmonary edema. Chest radiographic
findings in pulmonary hemorrhage depend on whether the hemorrhage is focal or massive. Because blood or
hemorrhagic edema fluid has tissue density, hemorrhagic tissue appears opacified. It is often difficult to
differentiate focal hemorrhage from atelectasis or pneumonia by chest radiographs. In the case of massive
pulmonary hemorrhage, the lungs can be atelectatic and opacified. The clinical course of massive pulmonary
hemorrhage usually involves rapid deterioration of ventilatory function. Affected infants develop progressive
hypoxia and hypercarbia with resultant respiratory acidosis and may rapidly succumb to this disorder.
Treatment
Early detection and aggressive intervention improve the outcome of massive pulmonary hemorrhage, an
otherwise lethal syndrome. Positive-pressure ventilation and oxygen are critical components of therapy. Blood
volume and hematocrit should be vigorously restored and maintained with erythrocyte transfusions. Careful
correction of hypotension, hypoxemia, and acidosis is also indicated. Coagulation abnormalities should be
assessed and may be corrected with fresh-frozen plasma or appropriate clotting factors. Pressors and diuretics
are indicated if congestive heart failure develops. Surfactant therapy has also been suggested as a useful
adjunct in neonates with a clinically significant pulmonary hemorrhage.
AIR LEAKS
Air leaks include pneumothorax, pneumomediastinum, pneumopericardium, and pulmonary interstitial
emphysema (PIE).
Pathophysiology
Pulmonary interstitial emphysema, pneumomediastinum, pneumothorax, and pneumopericardium are closely
related clinical entities. Air leak begins with formation of PIE in which alveoli rupture into the perivascular and
peribronchial spaces. Air may be trapped in the interstitium of the lung, leading to PIE, but it may also dissect
into the mediastinum along the perivascular and peribronchial spaces, producing pneumomediastinum.
Mediastinal air ruptures into the pleural space, producing pneumothorax, or into the pericardial space, producing
pneumopericardium. In some instances, air can form blebs on the surface of the lung that rupture to produce
pneumothorax. Rupture of the lung directly into the pleural space is thought to occur rarely.
Risk Factors
Air leaks occur in 1% to 2% of all newborn infants, but they are thought to cause symptoms in only 0.05% to
0.07%. Mechanical ventilation and CPAP are important risk factors contributing to air leak in infants with lung
disease. Cystic adenomatoid malformation also predisposes infants to spontaneous pneumothorax. Summarizing
data
P.575
from 11 studies, Madansky found that air leak occurred in 12% of infants with RDS who were not on assisted
ventilation, 11% of infants on CPAP, and 26% of infants on mechanical ventilation (58). The incidence of air leak
in infants admitted to the NICU is approximately 2% to 8%, but it is higher if only low-birth-weight infants are
considered. As of 1986, of infants weighing 500 to 999 g at birth who developed air leak, 35% had PIE, 20%
had pneumothorax, 3% had pneumomediastinum, and 2% had pneumopericardium (58). Infants developing air
leak are at higher risk of death, but the risk changes with postnatal age at the time of the air leak. Aspiration
syndromes, including MAS, are frequently complicated by air leak.
Radiographic Evaluation
The chest radiographs of infants with PIE have been described as demonstrating a salt-and-pepper pattern in
which the radiolucent interstitial air is juxtaposed to lung parenchyma (Fig. 29-13). Radiolucent air is present in
the pleural space in a pneumothorax. Because chest radiographs of neonates are usually performed in the
supine position, pleural air of a pneumothorax may accumulate in the anterior chest and may be visible only on
a cross-table lateral or decubitus radiograph. In a tension pneumothorax, the lung and mediastinal organs may
be displaced away from the side of the pneumothorax (Fig. 29-14). The thymus may be outlined in
pneumomediastinum, seen on radiographs. Pneumopericardium results in a characteristic outline of the heart by
radiolucent air.
Figure 29-13 Chest x-ray film of pulmonary interstitial emphysema (PIE). A premature infant with severe
respiratory distress syndrome requiring mechanical ventilation developed worsening respiratory acidosis and
hypoxia refractory to increased ventilatory support. An anteroposterior chest x-ray film demonstrates a saltand-pepper pattern resulting from radiolucent interstitial air surrounding compressed lung tissue. A left chest
tube was placed to treat pneumothorax, a common complication of pulmonary interstitial emphysema.
Figure 29-14 A full-term infant born by a difficult breech delivery presented shortly after birth with crepitus in
the neck area, tachypnea, grunting, and retractions. An anteroposterior chest x-ray film demonstrates bilateral
pneumothorax under tension on the left. The heart and mediastinum are compressed and shifted to the right.
The left pleural air herniates across the midline. The left diaphragm is depressed and inverted. Subcutaneous
emphysema is seen in the soft tissues of the neck.
Pulmonary Interstitial Emphysema

Pulmonary interstitial emphysema occurs most frequently in smaller infants being treated by mechanical
ventilation for primary lung disease. In this clinical setting, PIE is associated with a mortality rate of more than
50%. Unilateral PIE can be managed by placing the infant with the affected side down for 24 to 48 hours.
Selective bronchial intubation and high-frequency or jet ventilation have been used to treat unilateral PIE.
Careful attention to peak and mean inspiratory pressures may be beneficial in preventing and treating PIE. Highfrequency ventilation may be helpful. Bronchopulmonary dysplasia is a frequent sequel in infants surviving PIE.
Pneumothorax, Pneumomediastinum, and Pneumopericardium

Infants with pneumothorax often present with grunting, tachypnea, cyanosis (See Color Plate), and retractions.
Accumulated air may collapse the lung or shift the mediastinum to the side opposite the air leak. A shift of the
trachea or point of maximal impulse and decreased breath sounds on the affected side may be found on clinical
examination. Pneumothoraces fall into two major groups: spontaneous pneumothorax in otherwise healthy, fullterm infants, which most often occurs within minutes of birth, and pneumothorax in infants with significant
pulmonary disease, which frequently occurs several days after birth, during therapy for pulmonary disease.
P.576
Prompt recognition of air leak is essential for effective therapy. Unexpected changes in ventilatory requirements
or status and abrupt fall in blood pressure, heart rate, respiratory rate, and PO2 may indicate an air leak.
Transillumination of the thorax can be useful in the diagnosis of pneumothorax and the response to therapy.
Treatment of tension pneumothorax requires immediate surgical drainage and placement of a chest tube. For
treatment of pneumothorax that does not involve tension or cardiovascular compromise, inhalation of 100%
oxygen, usually for 4 to 6 hours, can be used as a nitrogen washout method in term infants. Premature infants
should not be treated with hyperoxia because they are at risk for retrolental fibroplasia. Pneumomediastinum
and pneumopericardium that does not unduly stress the infant can also be managed with 100% oxygen therapy.
Tension pneumopericardium is life threatening, must be drained surgically, and is associated with a high
incidence of morbidity and mortality.
REFERENCES
1. Burri PH. Postnatal development and growth. In: Crystal RG, West JB, Barnes PJ, Cherniack NS, Weibel ER,
eds. The lung: scientific foundations. New York: Raven Press, 1991:677.
2. Randell SH, Young SL. Structure of alveolar epithelial cells and the surface layer during development. In:
Polin RA, Fox WW, eds. Fetal and neonatal physiology. Philadelphia: WB Saunders, 1992:962.
3. Nogee LM, Wert SE, Proffit SA, et al. Allelic heterogeneity in hereditary surfactant protein B (SP-B)
deficiency. Am J Respir Crit Care Med 2000;161:973-981.
4. Nogee LM, Dunbar AE 3rd, Wert SE, et al. A mutation in the surfactant protein C gene associated with
familial interstitial lung disease. N Engl J Med 2001;344:573-579.
5. Shulenin S, Nogee LM, Annilo T, et al. The ABCA3 gene is frequently mutated in human newborns with fatal
surfactant deficiency. N Engl J Med 2004;350:1296-1303.
6. Van Golde LM, Batenburg JJ, Robertson B. The pulmonary surfactant system: biochemical aspects and
functional significance. Physiol Rev 1988;68:374-455.
7. Whitsett JA. RDS in the premature infant. In: Crystal RG, West JB, Barnes PJ, Cherniack NS, Weibel ER,
eds. The lung: scientific foundations. New York: Raven Press, 1991:1723.
8. Shapiro DL, Notter RH. Surfactant replacement therapy. New York: Alan R Liss, 1989.
9. Rooney S. Regulation of surfactant associated phospholipid synthesis and secretion. In: Polin RA, Fox WW,
eds. Fetal and neonatal physiology. Philadelphia: WB Saunders, 1992:986.
10. Ballard PL. Hormonal regulation of pulmonary surfactant. Endocrinol Rev 1989;10:165-181.
11. Liggins GC, Howie RN. A controlled trial of antepartum glucocorticoid treatment for prevention of the
respiratory distress syndrome in premature infants. Pediatrics 1972;50:515-525.
12. Jobe A. Phospholipid metabolism and turnover. In: Polin RA, Fox WW, eds. Fetal and neonatal physiology.
Philadelphia: WB Saunders, 1992:986.
13. Fujiwara T, Maeta H, Chida S, et al. Artificial surfactant therapy in hyaline membrane disease. Lancet
1980;1:55-59.
14. Avery ME, Mead J. Surface properties in relation to atelectasis and hyaline membrane disease. Am J Dis
Child 1959;97:517-523.
15. Newnham JP, Moss TJ, Nitsos I, et al. Antenatal corticosteroids: the good, the bad and the unknown. Curr
Opin Obstet Gynecol 2002;14:607-612.
16. Spitzer AR, Shaffer TH, Fox WW. Assisted ventilation: physiologic implications and application. In: RA
Polin, WW Fox, eds. Fetal and neonatal physiology. Philadelphia: WB Saunders, 1991:894.
17. Suresh GK, Soll RF. Lung surfactants for neonatal respiratory distress syndrome: Animal-derived or
synthetic agents? Paediatr Drugs 2002;4:485-492.
18. Henderson-Smart DJ, Bhuta T, Cools F, et al. Elective high frequency oscillatory ventilation versus
conventional ventilation for acute pulmonary dysfunction in preterm infants. Cochrane Database Syst Rev
2003;1:CD000104.
19. Barrington KJ, Finer NN. Inhaled nitric oxide for respiratory failure in preterm infants. Cochrane Database
Syst Rev 2001;4: CD000509.
20. Wiswell TE. Handling the meconium-stained infant. Semin Neonatol 2001;6:225-231.
21. Tran N, Lowe C, Sivieri EM, et al. Sequential effects of acute meconium obstruction on pulmonary
function. Pediatr Res 1980;14: 34-38.
22. Ting P, Brady JP. Tracheal suction in meconium aspiration. Am J Obstet Gynecol 1975;122:767-771.
23. Hofmeyr GJ. Amnioinfusion for meconium-stained liquor in labour. Cochrane Database Syst Rev 2002;1:
CD000014.
24. Halliday HL. Endotracheal intubation at birth for preventing morbidity and mortality in vigorous,
meconium-stained infants born at term. Cochrane Database Syst Rev 2001;1:CD000500.
25. Soll RF, Dargaville P. Surfactant for meconium aspiration syndrome in full term infants. Cochrane
Database Syst. Rev 2000;2: CD002054.
26. Findlay RD, Taeusch HW, Walther FJ. Surfactant replacement therapy for meconium aspiration syndrome.
27. Kinsella JP, Truog WE, Walsh WF, et al. Randomized, multicenter trial of inhaled nitric oxide and highfrequency oscillatory ventilation in severe, persistent pulmonary hypertension of the newborn. J Pediatr
1997;131:55-62.
28. Extracorporeal Life Support Registry Report, International Summary. July, 2003.
29. Gersony W, Duc G, Sinclair J. PFC syndrome. Circulation 1969; 40[Suppl III]:87.
30. Walsh MC, Stork EK. Persistent pulmonary hypertension of the newborn. Rational therapy based on
pathophysiology. Clin Perinatol 2001;28:609-627.
31. Rabinowitz M. Structure and function of the pulmonary vascular bed: an update. Cardiol Clin 1989;7:227.
32. Weinberger B, Weiss K, Heck DE, et al. Pharmacologic therapy of persistent pulmonary hypertension of
the newborn. Pharmacol Ther 2001;89:67-79.
33. Sadiq HF, Mantych G, Benawra RS, et al. Inhaled nitric oxide in the treatment of moderate persistent
pulmonary hypertension of the newborn: A randomized controlled, multicenter trial. J Perinatol 2003;23:98103.
34. Finer NN, Barrington KJ. Nitric oxide for respiratory failure in infants born at or near term. Cochrane
Database Syst Rev 2001;4: CD000399.
35. Bartlett RH, Gazzaniga AB, Huxtable RF, et al. Extracorporeal circulation (ECMO) in neonatal respiratory
failure. J Thorac Cardiovasc Surg 1977;74:826-823.
36. Rais-Bahrami K, Short BL. The current status of neonatal extracorporeal membrane oxygenation. Semin
Perinatol 2000;24:406-417.
37. Ballard RA, Leonard CH. Developmental follow-up of infants with persistent pulmonary hypertension of the
newborn. Clin Perinatol 1984;11:737-744.
38. Lipkin PH, Davidson D, Spivak L, et al. Neurodevelopmental and medical outcomes of persistent
pulmonary hypertension in term newborns treated with nitric oxide. J Pediatr 2002;140:306-310.
39. Rosenberg AA, Kennaugh JM, Moreland SG, et al. Longitudinal follow-up of a cohort of newborn infants
treated with inhaled nitric oxide in persistent pulmonary hypertension. J Pediatr 1997;131:70-75.
40. Gaynes RP, Edwards JR, Jarvis WR, et al. Nosocomial infections among neonates in high risk nurseries in
the United States. National Nosocomial Infection's Surveillance System. Pediatrics 1996;98:357-361.
41. Schrag S, Gorwitz R, Fultz-Butts K, et al. Prevention of perinatal group B streptococcal disease. Revised
guidelines from CDC. MMWR Recomm Rep 2002;51:1-22.
42. Hyde TB, Hilger TM, Reingold A, et al. Trends in incidence and antimicrobial resistance of early-onset
sepsis: population-based surveillance in San Francisco and Atlanta. Pediatrics 2002;110: 690-695.
43. Stoll BJ, Hansen N, Fanaroff AA, et al. Pathogens causing early-onset sepsis in very low birth weight
infants. N Engl J Med 2002;347:240-247.
P.577
44. Schuchat A, Zywicki SS, Dinsmoor MJ, et al. Risk factors and opportunities for prevention of early-onset
neonatal sepsis: a multicenter case-control study. Pediatrics 2000;105:21-26.
45. Hamprecht K, Maschmann J, Vochem M, et al. Epidemiology of transmission of cytomegalovirus from

mother to preterm infant by breastfeeding. Lancet 2001;357:513-518.
46. Kimberlin DW, Lin CY, Jacobs RF, et al. Natural history of neonatal herpes simplex virus infection in the
acyclovir era. Pediatrics 2001;108:223-229.
47. Hickman ME, Rench MA, Ferrieri P, et al. Changing epidemiology of group B streptococcal colonization.
Pediatrics 1999;104: 203-209.
48. Schrag SJ, Zywicki S, Farley MM, et al. Group B streptococcal disease in the era of intrapartum antibiotic
prophylaxis. N Engl J Med 2000;342:15-20.
49. Bromberger P, Lawrence JM, Braun D, et al. The influence of intrapartum antibiotics on the clinical
spectrum of early-onset group B streptococcal infection in term infants. Pediatrics 2000; 106:244-250.
50. Rojas J, Stahlman M. The effect of group B Streptococcus and other organisms on the pulmonary
vasculature. Clin Perinatol 1984;11: 591-599.
51. Whitley R, Arvin A, Prober C, et al. Predictors of morbidity and mortality in neonates with herpes simplex
virus infections. The National Institute of Allergy and Infectious Diseases Collaborative Antiviral Study Group.
N Engl J Med 1991;324:450-454.
52. Kimberlin DW, Lin CY, Jacobs RF, et al. Safety and efficiency of high dose intravenous acyclovir in the
management of neonatal herpes simplex virus infection. Pediatrics 2001;108:230-238.
53. Avery ME, Gatewood OB, Brumley G. Transient tachypnea of newborn. Possible delayed resorption at
birth. Am J Dis Child 1966;111:380-385.
54. Levine EM, Ghai V, Barton JJ, et al. Mode of delivery and risk of respiratory diseases in newborns. Obstet
Gynecol 2001;97:439-442.
55. Schaubel D, Johansen H, Dutta M, et al. Neonatal characteristics as risk factors for preschool asthma. J
Asthma 1996;33:255-264.
56. Lewis V, Whitelaw A. Furosemide for transient tachypnea of the newborn. Cochrane Database Syst Rev
2002;1:CD003064.
57. Kluckow M, Evans N. Ductal shunting, high pulmonary blood flow, and pulmonary hemorrhage. J Pediatr
2000;137:68-72.
58. Madansky DL, Lawson EE, Chernick V, et al. Pneumothorax and other forms of pulmonary air leak in
newborns. Am Rev Respir Dis 1979;120:729-737.
Chapter 30
Bronchopulmonary Dysplasia
Jonathan M. Davis
Warren N. Rosenfeld
Bronchopulmonary dysplasia (BPD) is a chronic lung disease that develops in newborn infants treated with
oxygen and positive-pressure mechanical ventilation for a primary lung disorder. The introduction of new
treatment modalities (e.g., surfactant replacement therapy, high-frequency ventilation, extracorporeal
membrane oxygenation, inhaled nitric oxide [INO]) has significantly improved the outcome for many critically ill
premature and term infants. As a result, more infants are surviving the newborn period and developing BPD.
Approximately 10,000 new cases of BPD occur each year in the United States and there is significant associated
morbidity and mortality. There has been significant debate about the exact definition of BPD, further complicated
because the specific nature of the disease has changed with the widespread use of surfactant and other
therapeutic interventions. Although the current form of BPD appears to be much less severe than in the past,
BPD is still an extremely important complication of neonatal intensive care (NICU) and the most common form
of chronic lung disease in infants.
The modern history of BPD began with Northway's observations in 1967 (1). This study documented the clinical
course, radiographic findings, and histopathologic lung changes in a group of infants who had received oxygen
and ventilatory support for treatment of respiratory distress syndrome (RDS) and established the term BPD.
Although Northway originally postulated that oxygen toxicity caused BPD, the exact mechanisms causing the
lung injury are complex and are currently the subject of intense investigation. Treatment with positive-pressure
ventilation appears to be a critical factor in the development of BPD, although factors such as oxygen toxicity,
prematurity, genetic predisposition, infection and inflammation may also play important roles. Therapies for
infants with BPD are directed toward improving the pathophysiologic abnormalities after they occur and include
oxygen, mechanical ventilation, fluid restriction, and a variety of medications. Many different therapies can be
used in these infants, often concurrently. The optimal treatment and prevention strategies have not been
definitively established.
This chapter reviews the definition and incidence of BPD (also known as chronic lung disease in newborns), its
pathogenesis, pathophysiologic changes, treatment strategies, and long-term outcome. Newly developed
approaches for the prevention of BPD in high-risk infants are presented.
DEFINITION AND INCIDENCE

In the original description of BPD, Northway defined chronic lung changes in a group of premature neonates who
survived mechanical ventilation for treatment of RDS (1). Northway's definition of BPD included radiologic,
pathologic, and clinical criteria. BPD was divided into four developmental stages. The acute stages, I and II were
indistinguishable from RDS and were seen in the first 10 days of life. Stages III and IV marked the transition to
the chronic stages of this disease, with stage IV forming the basis for the definition of BPD. This stage described
abnormalities that persisted beyond 1 month of age in patients who continued to require respiratory support (i.
e., ventilation or oxygen supplementation). Chest radiographs demonstrated cyst formation and hyperexpansion
alternating with areas of atelectasis.
As the care of neonates has become more sophisticated and smaller, sicker infants survived, the clinical and
radiographic findings that define BPD have changed. Northway's original criteria depended heavily on a
progression of radiographic changes, and clinical criteria were considered secondary. Further refinement was
offered by Bancalari, whose criteria included ventilation for at least the first 3 days of life and respiratory
symptoms (e.g., tachypnea, auscultatory rales, retractions) at 28 days of life, a need for supplemental oxygen
to maintain a partial pressure (PaO2) greater than 50 mm Hg and an abnormal chest radiograph at 28 days of
life (2).
TABLE 30-1 DEFINITION OF BRONCHOPULMONARY DYSPLASIA: DIAGNOSTIC CRITERIAa
Gestational Age
Time point of assessment
<32 Wkr
36 wk PMA or discharge to home,

whichever comes first
32 Wk
>28 d but <56 d postnatal age or

discharge home, whichever comes
first
Treatment with oxygen >21% for at least 28 d plus

Mild BPD
Moderate BPD
Severe BPD
Breathing room air at 36 wk PMA or

discharge, whichever comes first
Need for <30% oxygen at 36 wk
PMA or discharge, whichever comes
first
56 d postnatal age or discharge,

Need for <30% oxygen at 56 d
postnatal age or discharge,
Need for >30% oxygen and/or
Need for >30% oxygen and/or
positive pressure, (PPV or NCPAP) at positive pressure (PPV or NCPAP) at
36 wk PMA or discharge, whichever
comes first
56 d postnatal age or discharge,

Definition of abbreviations: BPD = bronchopulmonary dysplasia; NCPAP = nasal continuous positive

airway pressure; PMA = postmenstrual age; PPV = positive-pressure ventilation.
a
From Jobe AH, Bancalari E. Bronchopulmonary dysplasia. Am J Respir Crit Care Med 2001;163:1723,
with permission.
P.579
Many infants who eventually develop chronic lung disease do not require prolonged mechanical ventilation, nor
have significant radiographic abnormalities. Several studies have used less stringent criteria to define BPD such
as simply the requirement for oxygen supplementation at 28 days of life (3,4). Shennan and associates
questioned the large number of normal neonates who would be included by this criterion and suggested that the
need for supplemental oxygen at 36 weeks postconceptual gestational age may be a more accurate predictor of
ultimate pulmonary outcome (5). Other investigators have found that the most reliable method to definitively
establish the diagnosis of BPD is the presence of chronic respiratory symptoms (e.g., asthma, repeated
pulmonary infections) requiring treatment with bronchodilators or corticosteroids in the first year or two of life
(6,7). Recently, a consensus conference convened by the National Institutes of Health suggested a new
definition of BPD that incorporates many elements of previous definitions and attempts to categorize the severity
of BPD (Table 30-1) (7). Ultimately, the definition of BPD will need to be validated with clinically important longterm endpoints in a prospective fashion.
The incidence of BPD depends on the definition used and the patient population studied. Several surfactant
replacement trials have reported the incidence of BPD, defined as oxygen dependency at 28 days with
appropriate radiographic findings, to be in the range of 11% to 63% and varying significantly by center
(8,9,10,11). It appeared from these studies that the incidence of BPD decreased only slightly after surfactant
therapy. However, the prevalence, or total number of infants with BPD, increased because of improved survival.
Fenton and associates subsequently reported that exogenous surfactant had significantly improved the survival
of many low-birth-weight infants in their geographically defined population and was actually associated with a
marked increase in the incidence of BPD (12). BPD will continue to be an important problem for the
neonatologist and pediatric pulmonologist in the future. Further study of the mechanisms involved in the lung
injury process and the development of possible prevention strategies are urgently needed.
PATHOGENESIS
No single factor has been identified as the cause of BPD. Its origin is multifactorial and may depend on the
nature of the injury, mechanisms of response, or the infant's inability to respond appropriately to the injury
process (Fig. 30-1). Although Northway attributed the occurrence of BPD primarily to prolonged hyperoxia in
infants with RDS, numerous other causes have been proposed.
Barotrauma/Volutrauma
Although the initial phases of lung injury in BPD are the result of the primary disease process (e.g., RDS),
superimposed positive-pressure mechanical ventilation appears to
P.580
add to the lung injury and provoke a complex inflammatory cascade that ultimately leads to chronic lung
disease. Barotrauma is the term generally used to describe the lung injury that occurs secondary to positivepressure mechanical ventilation, although volutrauma from excessive tidal volume ventilation may be a more
appropriate term to describe this lung injury process.
Figure 30-1 An overview of the pathogenesis of BPD showing prenatal and postnatal factors. Adapted from
Jobe AH. The new BPD: an arrest of lung development. Pediatr Res 1999;46:641, with permission.
The role of barotrauma in BPD depends on several factors, including the structure of the tracheobronchial tree
and the physiologic effects of surfactant deficiency. With surfactant deficiency, surface tension forces are
elevated, aeration is unequal, and most terminal alveoli are largely collapsed. The pressure needed to distend
these poorly compliant saccules is high and is transmitted to the terminal bronchioles and alveolar ducts. In the
premature infant, these airways are highly compliant and subject to rupture. Gas then dissects into the
interstitium and pleural space, resulting in the development of pulmonary interstitial emphysema (PIE) and
pneumothorax. These complications are strongly associated with the development of BPD, suggesting that
ventilator-induced lung injury is important in the pathogenesis (13,14).
Nilsson and colleagues have shown that even brief periods of positive-pressure ventilation cause bronchiolar
epithelial damage in the lung, with the severity of the injury correlating with the amount of peak pressure used
(15). Even with normal tidal volumes, ventilation of the immature or injured newborn lung results in nonuniform
inflation and relative overdistension of ventilated segments, especially in the presence of low functional residual
capacity (FRC) as a result of surfactant deficiency. Stretching of capillary endothelium and distal lung epithelium
increases permeability to serum proteins that may further inhibit surfactant function, creating a vicious cycle
that promotes lung injury.
The adverse effects of tidal volume breaths in lungs with low FRC can be ameliorated with better lung
recruitment by application of higher positive end-expiratory pressure (PEEP). Experimental studies have shown
that overdistension of the lung (not increased pressure) is responsible for lung injury in the surfactant-deficient
lung (14). For example, strapping the chest wall to restrict overexpansion increases lung pressure but limits lung
stretch, and ameliorates lung injury in animal models (16). Tremblay and associates have observed that
mechanical ventilation with high tidal volumes and low PEEP markedly increases edema and cytokine production
in the lung of rats (17). In this same model, lung injury increased cytokine release from the lungs into the
systemic circulation, suggesting a potential mechanism linking multi-organ dysfunction and sepsis syndrome
with lung injury. Improved survival of adults with acute respiratory distress syndrome (ARDS) has been found
when employing a ventilator strategy that used low tidal volumes with high PEEP to recruit atelectatic lung
regions and improve FRC (18).
Strategies to prevent lung injury have included changes in methods of ventilation. Even brief exposure to large
tidal volume breaths during resuscitation shortly after birth can initiate early lung injury, which decreases the
subsequent response to surfactant therapy (19). With the increased use of nasal continuous positive airway
pressure (NCPAP), many preterm infants may be able to be managed without the need for endotracheal
intubation and mechanical ventilation, which may lessen the risk for lung injury. Synchronized intermittent
mechanical ventilation (SIMV) and high-frequency devices have also been extensively studied. Bernstein and
associates reported that infants (<1,000 g) treated with SIMV developed less BPD than those receiving
conventional ventilation (20). Several multicenter trials have demonstrated that high-frequency oscillatory
ventilation (HFOV) or high frequency jet ventilation (HFJV) using high volume, alveolar recruitment strategies in
conjunction with surfactant replacement therapy reduced complications of mechanical ventilation (PIE,
pneumothorax) and lowered the incidence of BPD compared to conventional ventilation (21,22). However, these
findings have not been consistently found in other trials. Regardless of the type of ventilation strategy used, it is
important to avoid even brief periods of overdistension, because high tidal volumes and hypocarbia are
associated with a greater risk for developing both BPD and neurologic injury (23,24). A retrospective study of
1,105 newborns born with birth weights about 2,000 g between 1984 to 1987 demonstrated a strong
relationship between prolonged hypocapnia (PaCO2 <35 torr) and the development of cerebral palsy (CP) (24).
The use of surfactant replacement therapy has reduced some complications of mechanical ventilation. Surfactant
permits more equal distribution of pressures and ventilation to all alveoli, prevents overdistension of distal air
spaces and bronchioles, and stabilizes lung volume. A major benefit of surfactant therapy has been the ability to
reduce airway pressures and the incidence of air leak, while maintaining a greater uniformity of lung
recruitment. However, BPD continues to be an important problem despite the widespread use of exogenous
surfactant and strategies to minimize lung injury from mechanical ventilation (e.g., permissive hypercapnia).
Van Marter and associates recently studied 452 premature infants weighing 500 to 1,500 g at birth who were
born at specific neonatal centers in either Boston or New York. They found that the incidence of BPD was
significantly higher in Boston (22%) compared to New York (4%), even after adjusting for baseline risk factors,
such as severity of prematurity (25). Using multivariate analyses to examine differences in specific respiratory
care practices during the first week of life, most of the increased risk of BPD was associated with the early
initiation of mechanical ventilation. Interestingly, the use of both exogenous surfactant and indomethacin
increased the risk for BPD. This suggests that attempts to minimize the use of mechanical ventilation using nasal
continuous positive airway pressure (NCPAP) (with or without exogenous surfactant) may actually lower the
incidence and severity of BPD in high-risk infants. Ongoing multicenter trials of early NCPAP in premature infants
are currently underway. BPD continues to be a serious problem, suggesting that
P.581
barotrauma is only one of many factors involved in the pathogenesis of BPD.
Oxygen and Antioxidants

Under normal conditions, a delicate balance exists between the production of oxygen free radicals (also called
reactive oxygen species) and the antioxidant defenses that protect cells in vivo. Free radicals are molecules with
extra electrons in their outer ring and they are toxic to living tissues (Table 30-2). Oxygen has a unique
molecular structure and is abundant within cells. It readily accepts free electrons generated by oxidative
metabolism within the cell, producing free radicals. Increased free-radical production can occur under conditions
of hyperoxia, reperfusion, or inflammation. Alternatively, free radicals can increase because of an inability to
quench production because of inadequate antioxidant defenses. Damage caused by oxygen free radicals includes
lipid peroxidation, mitochondrial injury, protein nitration and unraveling of nucleic acids.
The premature neonate may be more susceptible to free-radical damage because adequate concentrations of
antioxidants may be absent at birth. Frank and associates documented the development of the antioxidant
enzymes superoxide dismutase (SOD), catalase, and glutathione peroxidase in the lungs of rabbits during late
gestation (Fig. 30-2) (26). The 150% increase in these enzymes during the last 15% of gestation parallels the
maturation pattern of pulmonary surfactant. These developmental changes in the fetal lung allow proper
ventilation by reducing surface tension and provide for the transition from the relative hypoxia of intrauterine
development to the oxygen-rich extrauterine environment. Premature birth can occur before the normal
upregulation of these antioxidant systems and other free radical scavengers (e.g. vitamin E, ascorbic acid,
glutathione, and ceruloplasmin) and may result in an imbalance between oxidants and antioxidants and an
increased risk for the development of BPD. Experimental animal models have shown that exposure to chronic
hyperoxia can induce inflammation and lung injury that has many features seen in infants developing BPD. Both
epithelial and endothelial cells are extremely susceptible to oxidant injury, leading to increased permeability
edema and cell dysfunction. Hyperoxia also impairs mucociliary function, promotes inflammation and inactivates
antiproteases, further complicating clinical outcome.
TABLE 30-2 FREE RADICALS
Radical
Symbol
Antioxidant
Superoxide anion
O2-
Superoxide dismutase, uric acid, vitamin E
Singlet oxygen
1O
-carotene, uric acid, vitamin E
Hydrogen peroxide
H2O2
Catalase, glutathione peroxidase, glutathione
Hydroxyl radical
OH
Vitamins C and E
Peroxide radical
LOO
Vitamins C and E
Hydroperoxyl radical
LOOH
Glutathione transferase, glutathione peroxidase
aL,
lipid.
Figure 30-2 Developmental changes in antioxidant levels and activity during gestation. The increases in
superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GP) late in gestation are similar to
those seen for pulmonary surfactant (dark, thick line). From Frank L, Groseclose EE. Preparation for birth into
an O2-rich environment: the antioxidant enzymes in the developing rabbit lung. Pediatr Res 1984;18:240, with
permission.
Clinical studies suggest that free radicals are involved in the pathogenesis of BPD. Plasma concentrations of
allantoin, an oxidation byproduct of uric acid, has been shown to be significantly elevated in the first 48 hours of
life in infants developing BPD compared to controls (27). Expired pentane and ethane have also been measured
as indirect evidence of free radical-induced lipid peroxidation in the first week of life and found to be significantly
elevated in neonates subsequently developing BPD (28). Varsila and colleagues analyzed proteins in tracheal
aspirates in the first week of life and found evidence of protein oxidation (carbonyl formation) in infants
developing BPD (29). Banks and associates found a fourfold increase in plasma 3-nitrotyrosine concentration in
infants developing BPD compared to controls, indicating increased evidence of reactive nitrogen species
(endogenous nitric oxide [NO] reacting with superoxide) modifying serum proteins (30). Of note, the multicenter
STOP-retinopathy of prematurity (ROP) trial examined whether exposing premature infants to higher inspired
oxygen concentrations would prevent the development of severe ROP. Although the effects of the increased
oxygen were minimal on the eyes, exposed infants had a marked increase (55%) in the incidence of BPD and
pulmonary infections (31).
Further evidence for the role of oxygen free radicals in lung injury comes from animal studies, which have shown
that antioxidant supplementation with SOD and catalase reduces cell damage, increases survival, and prevents
lung
P.582
injury from prolonged hyperoxia and mechanical ventilation (32,33). Genetically engineered mice
overexpressing SOD survive longer, whereas mice with disrupted SOD genes die more quickly in a hyperoxic
environment compared to normal diploid controls (34,35). These studies all demonstrate that oxygen free
radicals are intimately involved in the development of acute lung and chronic lung disease in newborn infants.
Inflammation
Marked inflammation in the lung appears to begin a cascade of destruction and abnormal repair that develops
into BPD. The initial stimuli activating the inflammatory process in the lung may be oxygen free radicals,
pulmonary barotrauma, infectious agents, or other stimuli that result in the attraction and activation of
leukocytes. Inflammation appears to play an important role in the pathogenesis of BPD and allows many factors
to be unified into a single cause of this disease. Inflammatory mediators and cellular responses have been found
to be prominent in animal models of lung injury and in infants who develop BPD.
A sustained increase in the number of neutrophils that generate excessive oxygen free radicals in tracheal fluid
samples distinguishes infants who develop BPD from those who recover spontaneously (36,37). Additionally, the
presence of activated macrophages, high concentrations of lipid products, inactivated -1-antitrypsin activity,
and other markers of active inflammation are strongly linked with the development of BPD (38). Release of early
response cytokines, such as tumor necrosis factor- (TNF-), interleukin-1 (IL-1), interleukin-8 (IL-8), and
transforming growth factor (TGF)- by macrophages and the presence of soluble adhesion molecules (e.g.,
selectins) may impact other cells to release chemoattractants and recruit neutrophils which further amplify the
inflammatory response (38,39,40,41,42). Elevated concentrations of pro-inflammatory cytokines (IL-6, IL-8) in
conjunction with reduced antiinflammatory products, (IL-10) usually appear in tracheal aspirates within a few
hours of life in infants subsequently developing BPD (40,41,42). All these agents recruit and activate leukocytes
resulting in significant pulmonary damage, including breakdown of capillary endothelial integrity and leakage of
macromolecules (e.g., albumin) into alveolar spaces. Albumin leakage and pulmonary edema are known to
inhibit surfactant function and have been postulated to be important factors in the development of BPD (43).
The release of elastase and collagenase from activated neutrophils directly destroys the elastin and collagen
framework of the lung. The breakdown products of collagen (hydroxyproline) and elastin (desmosine) have been
recovered in the urine of infants who develop BPD (44). The major defense against the action of elastase activity
is 1-proteinase inhibitor, which may be inactivated by oxygen radicals. Increased elastase activity accompanied
by compromised antiproteinase function may enhance lung injury. This combination has been demonstrated in
tracheal aspirates and serum of neonates who develop BPD (45,46,47).
As the acute cycle of injury continues with further production and accumulation of inflammatory mediators,
significant injury to the lung can occur during a particularly critical period of rapid growth; six divisions from 24
to 40 weeks of gestation. It appears likely that this abnormal inflammatory process is primarily responsible for
the acute and the chronic changes that occur in the lungs of infants with BPD. Watterberg and associates have
demonstrated that extremely premature infants have evidence of adrenal insufficiency at birth with the lowest
serum cortisol concentrations during the first week of life correlating with increased lung inflammation and
adverse respiratory outcome (48). Pilot studies examining early treatment with low-dose hydrocortisone in
extremely low birth weight (LBW) infants increased the likelihood of survival without BPD, a benefit that was
particularly apparent in infants born to mothers with chorioamnionitis (49). The mechanisms of action may
include suppression of lung inflammation with less inhibition of lung growth observed with higher doses of
steroids or even beneficial effects of low dose hydrocortisone on lung development. These studies all clearly
support the notion that this abnormal inflammatory process is primarily responsible for the acute and the chronic
changes that occur in the lungs of infants with BPD.
Infection
Subclinical intrauterine infection and the ensuing inflammatory response have been clearly implicated in the
etiology of preterm labor and premature rupture of membranes (50). Growing evidence indicates that prenatal
infection and inflammation are major risk factors for the subsequent development of BPD. Although several
investigators have found a lower incidence of RDS in preterm infants born to mothers with chorioamnionitis and
funisitis (possibly as a result of an adaptive response to in utero stress), they also observed a significantly
higher incidence of BPD in these same infants (51). This suggests that although intrauterine infection may
accelerate lung maturation, the ensuing inflammatory response may also prime the lung, causing lung injury,
progressive inflammation and subsequent inhibition of lung growth. Subsequent exposure to hyperoxia and
mechanical ventilation may further exacerbate the injury process in an already infected and/or inflamed lung.
Ureaplasma urealyticum has been recovered from cervical cultures of pregnant women and implicated as a
possible cause of chorioamnionitis, prematurity, and BPD (52). Cassell and colleagues cultured tracheal aspirates
and blood and found that BPD developed in 82% of infants (<1,000 g) colonized with Ureaplasma, compared
with 41% of those with negative cultures (52). In contrast, other studies have found that although Ureaplasma
was frequently detected (by culture or polymerase chain reaction) in many LBW infants, its presence was not
associated
P.583
with the development of BPD (53,54). It is possible that infection simply acts as a stimulus for the inflammatory
response, with recruitment of leukocytes and the production of cytokines and oxygen free radicals, ultimately
leading to the development of BPD. Normal defense mechanisms against infection can be compromised in the
lungs of these chronically ventilated, premature infants. This makes them more susceptible to colonization and
subsequent infection with a variety of infectious agents (e.g., virus, bacteria, fungi) that may affect the severity
of BPD (55). Several large clinical studies have found a strong correlation between the presence of BPD and the
development of late-onset sepsis, usually with organisms such as staphylococcus epidermidis (56,57). These
infections are associated with increased morbidity, mortality, and length of hospital stay and appear to
contribute to the severity of BPD.
Nutrition
The nutritional status of the critically ill premature infant may also be important in the development of BPD.
Adequate calories and essential nutrients for growth may be lacking during a period of stress and growth; vital
components for immunologic and antioxidant defenses may be inadequate; and the nutritional supplements
provided may actually contribute to ongoing damage.
Premature infants have increased nutritional requirements because of increased metabolic needs and rapid
growth requirements. If these increased energy needs are not met, the infant will develop a catabolic state,
which most likely contributes to the pathogenesis of BPD. Inadequate nutrition could interfere with normal
growth and maturation of the lung and may potentiate the deleterious effects of oxygen and barotrauma.
Newborn rats with inadequate caloric intake have decreased lung weights, protein levels, and DNA content (58).
These abnormalities were even greater in pups that were nutritionally deprived at birth and exposed to
hyperoxia.
Antioxidant enzymes may play a vital role in the protection of the lung and the prevention of BPD. Many of these
enzymes have trace elements (e.g., copper, zinc, selenium) that are an integral part of their structure.
Deficiencies in these elements may compromise the premature infant's defenses and predispose the lung to
further injury. The repair of elastin and collagen is limited in animals that are undernourished, and copper and
zinc may be necessary for this repair (59). Although supplementation with these elements may provide
protection to the lung and prevent hyperoxic lung injury, clinical trials using limited dosing strategies have not
been able to demonstrate a beneficial effect in preventing BPD (60). Vitamin deficiency has been postulated to
be important in the development of BPD. Although current nursery feeding and hyperalimentation regimens
appear to provide adequate amounts of vitamin E for preterm infants, a relative decrease in the concentrations
of other vitamins may be important in the pathogenesis of BPD. For example, Vyas and colleagues reported
significant decreases in levels of ascorbate (which can function as an antioxidant) in tracheal aspirate fluid from
infants developing BPD compared to control infants, suggesting that supplementation with vitamin C might be
beneficial (61). However, early supplementation with high dose antioxidant vitamins has been studied in
preterm baboons and was not efficacious in preventing lung damage from prolonged hyperoxia (62).
Concentrations of vitamin A (retinol) may also be deficient in very LBW infants (63,64,65). This vitamin appears
to be important in maintaining cell integrity and in tissue repair. Its deficiency has been associated with changes
in the ciliated epithelium of the tracheobronchial tree (66). Hustead and associates demonstrated lower serum
retinol levels in cord blood and at day 21 of life in infants who developed BPD (65). Similarly, Shenai and
colleagues demonstrated lower plasma retinol concentrations in the first month of life in infants who
subsequently developed BPD (67). Despite adequate supplementation, some infants remain vitamin A deficient,
presumably from increased absorption of parenteral vitamin A into the tubing of the intravenous administration
set or from higher nutritional requirements (68). A multicenter trial of vitamin A supplementation in premature
infants at risk for developing BPD recently demonstrated that large doses of intramuscular vitamin A given three
times per week was associated with a small (7%), but significant reduction in the incidence of BPD, suggesting
that vitamin A deficiency is an important contributor to lung injury (69).
Large volumes of intravenous fluids are often administered to premature infants to provide adequate fluid
requirements (from increased insensible water losses) and sufficient calories. Excessive fluid administration can
be associated with the development of a patent ductus arteriosus and pulmonary edema, which can lead to an
increase in oxygen and ventilator requirements and the subsequent risk of BPD (55,70). Early closure of the
ductus, using indomethacin or surgical ligation, has been associated with improvements in pulmonary function,
but these approaches have not affected the incidence of BPD (71).
Genetics
Numerous investigators have observed that neonates were more likely to develop BPD if there was a strong
family history of atopy and asthma. Nickerson and Taussig found a positive family history of asthma in 77% of
infants with RDS who subsequently developed BPD, compared with only 33% who did not (72). Bertrand
evaluated the relationship of prematurity, RDS, and need for mechanical ventilation to a family history of airway
hyperactivity (73). The severity of lung disease was directly related to the degree of prematurity and the
duration of oxygen exposure. However, siblings and mothers of infants with the most significant lung disease
had evidence of airway reactivity, suggesting that all three factors are involved in
P.584
determining long-term outcome. When histocompatibility loci (HLA) were examined, Clark and associates found
that only infants with HLA-A2 developed BPD, again suggesting that other underlying factors that are poorly
understood may be important in the pathogenesis of BPD (74). More recently, Hagan reported that a family
history of asthma is associated with an increase in the overall severity of BPD in premature infants but does not
appear to be a causative factor (75).
PATHOPHYSIOLOGIC CHANGES
Infants with BPD demonstrate abnormal findings on clinical examination, chest radiograph, pulmonary function
testing, echocardiogram, and morphologic examination of the lung. The severity of BPD is directly proportional
to the degree of the pathophysiologic insult and can be assessed through all of these techniques. Determining
the severity of BPD is complex and has been the subject of several workshops sponsored by the National
Institutes of Health and many publications. Several scoring systems have been developed to address this
important issue.
Clinical Assessment
A clinical scoring system to help evaluate the severity of BPD was developed by Toce and colleagues (76).
Infants with BPD are tachypneic and may have intercostal and subcostal retractions. Accessory muscles may be
used to assist with respiration. Infants can be hypoxic and hypercarbic and may grow poorly despite adequate
caloric intake. The Toce system attempts to standardize clinical assessment, and a severity score can be
assigned to each infant at 28 days of postnatal age and at 36 weeks postconceptual age. The clinical assessment
should be adjusted if infants are receiving multiple medications for their BPD (e.g., diuretics, methylxanthines),
NCPAP, or positive-pressure mechanical ventilation.
Radiographic Abnormalities
Northway and associates first described radiographic abnormalities characteristic of BPD in 1967 (1). A staging
system was employed that documented the progression of the disease process through four distinct stages. The
first stage was similar to uncomplicated RDS and occurred in the first few days of life. The BPD often progressed
to pulmonary parenchymal opacities (stage II), a bubbly appearance (stage III), and then to an inhomogeneous
appearance with marked hyperinflation, bleb formation, irregular fibrous streaks, and cardiomegaly (stage IV).
The radiographic progression of BPD is now seldom categorized by these four stages, so Edwards and colleagues
redefined the radiologic changes (77). Their system is based on the four most prominent radiographic findings in
BPD, including lung expansion, emphysema (including bleb formation), interstitial densities, and cardiovascular
(CV) abnormalities (Table 30-3). More severe changes are associated with higher scores (maximum of 10). The
occurrence of hyperinflation or interstitial abnormalities on chest radiograph appears to correlate with the
development of airway obstruction later in life (78). Because the severity of BPD has continued to change so
significantly over the past decade, Weinstein developed a new scoring system incorporating some of the more
subtle radiographic signs that are often seen today in infants with BPD (Fig. 30-3) (79). The utility of these
scoring systems remains to be demonstrated.
Computed tomography (CT) and magnetic resonance imaging (MRI) of the lung may provide more details of the
structural disease in BPD, and can reveal significant abnormalities that are not readily apparent on chest
radiographs (Fig. 30-4) (80). These findings can be important in determining ultimate pulmonary morbidity. CT
scans often show regional heterogeneity, with regions of hyperinflation or emphysema and sparse arterial
density alternating with relatively normal appearing regions. MRI of ventilated premature newborns may
demonstrate striking regional variations in lung disease, with marked gravity-dependent
P.585
atelectasis and edema. More studies of the role of CT and MRI in BPD are needed to correlate structural and
functional changes.
TABLE 30-3 ROENTGENOGRAPHIC SCORING SYSTEM FOR SEVERITY OF BRONCHOPULMONARY
DYSPLASIAb
a
Score
Variable
Cardiovascular abnormalities
None
Cardiomegaly
Gross cardiomegaly or RVH or

enlarged MPA
Hyperexpansion (anterior plus 14

posterior rib count)
14.5-16
Emphysema
No focal
areas
Scattered, small,
abnormal lucencies
16.5, or flattened
hemidiaphragms
At least one large bleb or bulla
Fibrosis or interstitial
abnormalities
None
Subjective
Mild
Interstitial prominence;
few abnormal, streaky
densities
Moderate
Dense fibrotic bands, many

abnormal strands
Severe
Rib counts intersecting level of the dome of the right hemidiaphragm. MPA, main pulmonary artery;
RVH, right ventricular hypertrophy.
b
From Edwards DK. Radiographic aspects of bronchopulmonary dysplasia. J Pediatr 1979;95:823, with
permission.
Figure 30-3 Typical chest radiograph of a 1-month-old infant with evolving bronchopulmonary dysplasia. The
bilateral hazy appearance represents inflammatory exudate, edema, and atelectasis.
Cardiovascular Changes
In addition to adverse effects on airways and alveoli, acute lung injury also impairs growth, structure and
function of the developing pulmonary circulation after premature birth (81). Endothelial cells have been shown
to be particularly susceptible to oxidant injury and the media of small pulmonary arteries may also undergo
smooth muscle cell proliferation (82). Structural changes in the lung vasculature contribute to high pulmonary
vascular resistance (PVR) as a result of narrowing of the vessel diameter and decreased angiogenesis. In
addition to these structural changes, the pulmonary circulation is further characterized by abnormal
vasoreactivity, which also increases PVR (83). Overall, injury to the pulmonary circulation can lead to the
development of pulmonary hypertension and cor pulmonale, which contributes significantly to the morbidity and
mortality of severe BPD (84). Echocardiographic assessment is an extremely valuable tool in confirming these
diagnoses.
In addition to pulmonary vascular disease and right ventricular hypertrophy, other CV abnormalities that are
associated with BPD include left ventricular hypertrophy (LVH) and systemic hypertension. Steroid therapy can
cause LVH, which tends to be transient and resolves when the drug is stopped. A high incidence of systemic
hypertension can be seen in BPD, but its etiology remains obscure (85). Systemic hypertension may be mild,
transient or severe and usually responds to pharmacologic therapy. On occasion, further evaluation of such
infants may reveal significant renal vascular or urinary tract disease. Whether the high incidence of systemic
hypertension in BPD reflects altered neurohumoral regulation, or increased catecholamines, angiotensin or
antidiuretic hormone levels are still unknown.
Changes in Pulmonary Mechanics

The development of computerized pulmonary function systems has enabled more accurate measurements of
pulmonary mechanics in newborn infants with BPD. Increased respiratory system resistance and bronchial
hyperactivity can be demonstrated even during the first week of life in preterm neonates who subsequently
develop BPD (86). These abnormalities are common in older infants with BPD and can cause dynamic airway
collapse and expiratory flow limitation (Fig. 30-5) (87). Other abnormalities of lung function include increased
dead
P.586
space ventilation, decreased lung compliance, maldistribution of ventilation and abnormal ventilation-perfusion
matching (88). Functional residual capacity (FRC) is often reduced early in the course as a result of atelectasis,
but increases during the later stages of BPD as a result of gas trapping and hyperinflation. Increased respiratory
rate with shallow breathing increases dead space ventilation that results from marked regional variability and
decreased lung surface area. Nonuniform damage to the airways and distal lungs results in variable time
constants for different areas of the lungs, altering the distribution of inspired gas to relatively poorly perfused
lung segments and worsening ventilation-perfusion matching.
Figure 30-4 A: Chest radiograph of a 2-month-old infant with bronchopulmonary dysplasia, showing rightsided atelectasis and a shift of the mediastinum. The lung fields have a hazy appearance. B: Computed
tomography scan on the same infant. The major bronchi and areas of atelectasis are apparent. Fibrotic changes
and a bleb are seen on the left (arrow).
Figure 30-5 A: Normal flow-volume loop. B: Expiratory flow limitation as a result of dynamic collapse of small
airways during expiration. INSP, inspiration; EXP, expiration.
Two methods used to measure lung compliance include dynamic measurement and passive expiratory
techniques after airway occlusion (89). Both methods show a reduction of lung compliance in infants with BPD.
The decrease in lung compliance appears to correlate well with morphologic changes in the lung (e.g.,
atelectasis, edema, fibrosis). Compliance can be reduced because of increased resistance, with frequency
dependence of compliance, when infants are breathing rapidly (90). The use of pulmonary function testing to
follow the progression of BPD and the response of the lung to various therapeutic interventions has become
widespread, but care must be used in the interpretation of results because of inherent variability in the
measurement and possible error from excessive chest wall distortion (91,92).
Pathologic Changes
Detailed morphometric studies have extensively characterized the lung pathology of infants dying with BPD
(7,93-95). The pathology of BPD provides insights into the effects of acute lung injury and repair processes in
the developing lung, and the impact of the timing of this injury. Pathologically, the original reports of BPD
described a continuous process through distinct stages of the disease, originating with an acute exudative phase
and progressing to a chronic proliferative phase of the disease (93). Older reports describe a gross cobblestone
appearance of the lungs, representing alternating areas of atelectasis, marked scarring and regional
hyperinflation. Typical histologic features of this BPD include marked airway changes, such as squamous
metaplasia of large and small airways, increased peribronchial smooth muscle and fibrosis, chronic inflammation
and airway edema, and hyperplasia of submucosal glands. Parenchymal disease is characterized by volume loss
from atelectasis and alveolar septal fibrosis alternating with overdistension or emphysematous regions (Fig. 306). Mesenchymal thickening with increased cellularity and destruction of alveolar septae with alveolar hypoplasia
is present, suggesting a marked reduction in surface area available for gas exchange. Growth of capillary beds is
reduced and small pulmonary arteries have hypertensive structural
P.587
remodeling, which includes smooth muscle hyperplasia and distal extension of smooth muscle growth into
vessels that are normally nonmuscular.
Figure 30-6 Light micrograph from a 1-year-old infant with bronchopulmonary dysplasia shows areas of
atelectasis alternating with areas of hyperinflation (Original magnification times 4.).
Figure 30-7 Lung histology from an infant dying in the postsurfactant era with the typical changes of the new
BPD, showing alveolar simplification and reduced septation (original magnification times four; provided by Dr.
Steven Abman).
Although reductions in alveolar number were described in older infants dying with BPD, a pattern of alveolar
simplification has become the most striking pathologic feature of the new BPD (Fig. 30-7) (7,94). In contrast
with past reports, recent studies of infants dying with BPD describe fewer signs of airway injury and interstitial
fibrosis, but emphasize persistent reductions of distal airspace and vascular growth (95). Decreased
alveolarization and impaired growth of small pulmonary arteries results in decreased lung surface area for gas
exchange, which has important functional implications regarding late cardiopulmonary sequelae (see above). In
addition to changes in the distal lung, the pathology of BPD is further characterized by abnormal airway
structure. The trachea and main bronchi of infants with BPD often reveal significant lesions, depending on the
frequency and duration of endotracheal intubation. Grossly, mucosal edema or necrosis can be focal or diffuse.
The earliest histologic changes include patchy loss of cilia from columnar epithelial cells, which can then become
dysplastic or necrotic, resulting in breakdown of the epithelial lining and decreased pulmonary clearance of
mucous and other material. Ulcerated areas may involve the mucosa or extend into the submucosa. Infiltration
of inflammatory cells (neutrophils and lymphocytes) into these areas may be prominent. Goblet cells appear
hyperplastic, suggesting greater capacity for increased mucous production that can mix with cellular debris.
Granulation tissue often develops in the subglottis as a result of damaged from the endotracheal tube or more
distally throughout the airway as a result of trauma from repeated suctioning. Significant narrowing of the
trachea and main bronchi secondary to injury can lead to subglottic stenosis, tracheal cysts and polyps, and
related lesions. Tracheomalacia may often complicate the course of severe BPD and can appear as marked
redundancy of the posterior wall of the trachea, as a result of chronic ventilation of the compliant premature
airway.
MANAGEMENT
The current approach to infants with BPD is multidisciplinary and directed toward improving the complex
pathophysiologic abnormalities that have been previously described. In addition to chronic respiratory disease,
infants with BPD have significant growth, nutritional, neurodevelopmental, and CV problems. The severity of
respiratory disease varies widely among infants with BPD. In its most severe presentation, children with BPD
may require chronic ventilation or prolonged home oxygen therapy. Even infants who have successfully weaned
off supplemental oxygen therapy have recurrent hospitalizations for lower respiratory exacerbations as a result
of viral infections (respiratory syncytial virus and influenza), reactive airways disease, pulmonary hypertension
or congestive heart failure. Additionally, persistent or recurrent respiratory exacerbations may also be as the
result of structural lesions (e.g., tracheomalacia, subglottic stenosis, bronchomalacia), chronic aspiration
(gastroesophageal reflux or swallowing dysfunction), or others. In addition to supplemental oxygen therapy,
chronic treatment with bronchodilators, diuretics, corticosteroids, and nutritional supplements are commonly
used as supportive therapy. Often, many of these therapies are used concurrently with the potential for
significant side effects.
P.588
The major treatment approaches are described in the following sections.
Mechanical Ventilation
With established disease, patients with BPD who require prolonged ventilatory assistance generally require
larger tidal volume breaths and a slower rate to enhance the distribution of gas as a result of heterogeneity of
lung units. Ventilation of infants with severe BPD with fast rates and smaller tidal volumes increases dead space
ventilation and may worsen gas exchange and increase gas trapping. Arterial blood gases should be optimally
maintained with a pH of 7.25 to 7.40, PCO2 of 45 to 55 mm Hg, and PO2 of 55 to 70 mm Hg. The best method
of assessing ade-quate oxygenation and ventilation is an arterial blood gas obtained from an indwelling arterial
catheter when the infant is quiet. Intermittent arterial puncture may be accurate if obtained quickly or if local
anesthesia is used, but may not be representative if the infant is awake and agitated. Capillary blood gases
should not be used to make significant therapeutic decisions because of wide variability and poor correlation
with arterial blood gases (especially in older infants who have already had multiple heel sticks for phlebotomy)
(96). However, if samples are obtained properly (i.e., after adequate warming and without squeezing the heel or
finger), the pH and PCO2 may correlate with arterial values. The use of pulse oximetry can assist in ventilatory
management and may reflect arterial values. End-tidal CO2 measurements may not correlate well with arterial
values in infants with significant ventilation-perfusion mismatch and chronic lung disease. As mentioned
previously, hypocarbia and hyperventilation may increase the risk for BPD and should definitively be avoided
(23,24).
Because oxygen appears to cause more significant lung damage than mechanical ventilation in animal models,
sufficient mean airway pressure should be used to avoid atelectasis and maintain the fraction of inspired oxygen
(FiO2) at less than 0.5 (97). Inspired gas temperature should be maintained from 36.5C to 37.5C to ensure
adequate humidity and minimize core temperature fluctuations and the progression of chronic lung disease (98).
Weaning infants with BPD from mechanical ventilation is often difficult and should be done slowly. When peak
inspiratory pressures have been reduced to approximately 15 to 20 cm H2O and FiO2 to less than 0.4 with
acceptable blood gases, the ventilator rate (if using a conventional, time cycled ventilator) should be reduced
slowly to allow the infant to gradually breathe more independently. This is essential because prolonged
ventilation may be associated with atrophy of the muscle fibers of the diaphragm and increased diaphragmatic
fatigue (99). When infants have been weaned to a ventilator rate of 10 to 20 breaths/min, elective extubation
should be attempted. With synchronized mechanical ventilation, the infant is allowed to set his/her own
inspiratory time and respiratory rate. Weaning infants from this type of ventilation involves reducing the
inspiratory pressure as previously described. Analysis of pulmonary mechanics before extubation may not be
useful in determining optimal time of extubation because multiple factors (e.g., central inspiratory drive,
diaphragmatic endurance, chest wall stability) may be important (100). Infants should not be weaned to
continuous endotracheal tube positive airway pressure (CPAP) because increased airway resistance and work of
breathing can cause fatigue, apnea, and carbon dioxide (CO2) retention (101). The use of methylxanthines
before extubation or NCPAP just after extubation may facilitate successful extubation (102,103). Higgins and
colleagues demonstrated that infants weighing less than 1,000 g were almost four times as likely to be
successfully extubated on the first attempt when NCPAP was used compared with an oxyhood (76% vs. 21%)
(103). Chest physiotherapy and suctioning should be performed frequently to maintain a patent airway and
prevent atelectasis.
Prolonged intubation and ventilation may be associated with the development of airway abnormalities (e.g.,
subglottic stenosis, tracheomalacia) (104). These should be considered in infants who rapidly and repeatedly fail
attempts at extubation. A bronchoscopic evaluation should be performed in these infants or in any infant
intubated for more than 2 to 3 months who continues to require prolonged ventilation. Surgical intervention (e.
g., cricoid split, tracheostomy) should be performed if necessary.
Oxygen
In infants with BPD, chronic hypoxia results in pulmonary vasoconstriction, pulmonary hypertension, and the
development of cor pulmonale. This contributes significantly to the morbidity and mortality of BPD. Elevated
pulmonary artery pressures and pulmonary vascular resistance have been described in infants with BPD
undergoing cardiac catheterization (83-85,105). Significant reductions in pulmonary pressures were found when
oxygen was administered. Oxygen is now known to act as a pulmonary vasodilator by stimulating the production
of endogenous NO (106). The NO diffuses into the vascular smooth muscle and stimulates the production of
cyclic GMP (cGMP), which causes vasodilation by sequestering calcium. Ideally, PaO2 should be maintained
between 55 and 70 mm Hg in infants with BPD.
Pulse oximetry has become the most popular noninvasive method of neonatal oxygen monitoring. Pulse
oximeters are most accurate when operating along the steep portion of the oxygen-hemoglobin dissociation
curve and may be more reliable than measuring PaO2 (107). Keeping the oxygen saturation (SaO2) generally
between 90% to 95% should exclude values of PaO2 that are less than 45 mm Hg or greater than 100 mm Hg.
Careful monitoring of oxygen status is essential because hypoxia can cause pulmonary hypertension, increased
airway constriction, and growth failure. Repeated episodes of oxygen desaturation often develop in premature
neonates with BPD during mechanical ventilation (108). This appears to be related to
P.589
a decrease in central respiratory drive and alterations in pulmonary mechanics. Prone positioning (with careful
monitoring) may increase SaO2 and decrease the frequency of hypoxemic episodes (109).
Hyperoxia may worsen BPD or increase the risk of retinopathy of prematurity (110). The STOP-ROP trial
examined the role of oxygen therapy in retinopathy of prematurity, and suggested that infants who were treated
with supplemental oxygen to maintain oxygen saturations more than 95% had a higher risk for BPD than infants
treated with supplemental oxygen that targeted lower saturations (31). However, interpretation of these data in
the setting of preventing or treating infants with pulmonary hypertension are limited, because the targeted
ranges for the low and high oxygen saturation groups fell between typical recommendations for oxygen therapy.
Current recommendations for treatment of patients with BPD and pulmonary hypertension are to maintain
oxygen saturations at 92% to 96%. Whether higher oxygen saturations increase lung injury remains unproven,
but the adverse effects of hypoxemia are clear.
Oxygen can be administered through an endotracheal tube, hood, tent, or nasal cannula. Increased FiO2 may be
needed during periods of increased stress (e.g., during feedings). Oxygen should be withdrawn gradually and
may be required for months. If oxygen-dependent infants can maintain an SaO2 of more than or equal to 90%
for at least 40 minutes in room air, then it appears that they can be successfully weaned from supplemental
oxygen (111). Infants with BPD can be discharged from the neonatal intensive care unit and receive oxygen at
home. The use of booster transfusions to increase oxygen-carrying capacity in oxygen-dependent infants with
BPD is controversial. Alverson and associates demonstrated significant increases in oxygen content and systemic
oxygen transport and decreases in oxygen consumption and oxygen use in infants with BPD after booster blood
transfusion (112). However, hemoglobin levels did not appear to correlate well with systemic oxygen transport
and did not predict which infants would benefit physiologically from transfusion. The need for multiple
transfusions has been significantly reduced by minimizing phlebotomy and by using human recombinant
erythropoietin therapy (113).
Nutrition
Because infants with BPD have increased metabolic demands, calories must be maximized to support tissue
repair and growth. Enteral feedings with fortified breast milk or with premature formulas provide the best source
of calories. Feedings should be given by intermittent or continuous gavage until the infant can be fed orally.
Feedings can be concentrated to increase caloric density, or glucose polymers (e.g., polycose), protein
supplements (ProMod) and medium-chain triglycerides (MCT) can be added to provide optimal calories although
minimizing fluid intake. Infants may require 120 to 140 calories/kg of body weight each day to gain weight (10
to 30 g/day). If fluid restriction interferes with the administration of adequate calories, a diuretic can be used to
prevent fluid overload.
Intravenous nutrition should be started early (day 1-2 of life) to provide adequate sources of protein, fat, and
carbohydrates. This may affect the ultimate outcome and severity of BPD, especially in the very-low-birthweight infant. The early use of percutaneous silastic central catheters has greatly enhanced the ability to provide
more optimal calories to LBW infants (114). Progressive increases in intravenous amino acid concentrations
should optimally provide up to 3.5 g of protein per kilogram each day. The acid-base status of the infant should
be monitored because acid loads may not be well tolerated. Intravenous lipids (20% suspension) should be
administered as a continuous infusion over 20 to 24 hours. Up to 3 g of lipids per kilogram each day can be
safely infused if serum triglyceride levels are closely followed. However, early administration of lipids may be
associated with more severe BPD (115,116). Intravenous glucose is a good source of calories, but excessive
loads (<4 mg/kg/min) may result in increased oxygen consumption, CO2 production, and resting energy
expenditures in infants with BPD (117). Adequate calcium and phosphorus intake is necessary, especially in
infants receiving furosemide, to promote bone mineralization and prevent secondary hyperparathyroidism and
rickets. Vitamins and trace metals should also be supplemented. Many premature infants are deficient in vitamin
A, and adequate supplementation may promote tissue regeneration and growth in the lung and decrease the
incidence and severity of BPD (66). Trace metals such as copper, zinc, and selenium are essential for the
structure and function of antioxidant enzymes and should be provided.
Medications
Many different types of drug therapies are used, often concurrently, to improve the clinical status of infants with
BPD. The exact dosages, efficacy, mechanisms of action, pharmacokinetics, and side effects have not been well
established. National Institutes of Health (NIH)/FDA consensus panels are currently working to establish optimal
treatment regimens for infants with BPD.
Diuretics
Furosemide (Lasix) is the treatment of choice for fluid overload in infants with BPD. It acts on the ascending loop
of Henle and blocks chloride transport. Furosemide increases plasma oncotic pressure and lymphatic flow and
decreases interstitial edema and pulmonary vascular resistance (118,119). Several studies have demonstrated
that daily, alternate-day or even aerosolized furosemide improves clinical respiratory status and pulmonary
mechanics and facilitates weaning from mechanical ventilation in infants with BPD (120,121,122). Side effects of
furosemide are numerous and include volume depletion, contraction alkalosis, hyponatremia, hypokalemia,
P.590
chloride depletion, renal calculi secondary to hypercalcinuria, nephrocalcinosis, cholelithiasis, osteopenia, and
ototoxicity (118). Supplemental potassium chloride is usually needed to prevent electrolyte depletion and
alkalosis, but sodium chloride supplements should be avoided if possible.
The thiazides affect renal tubular excretion of electrolytes, but they are less potent than furosemide (123).
Potassium and bicarbonate excretion accompany the sodium and chloride excretion produced by the thiazides.
For this reason, the thiazides are usually given in conjunction with spironolactone (Aldactone), which is a
competitive inhibitor of aldosterone. Spironolactone is a relatively weak diuretic that causes increased sodium,
chloride, and water excretion although sparing potassium. Although a few controlled trials examining the use of
a thiazide diuretic and spironolactone in infants with moderate BPD demonstrated increased urine output and
improvements in pulmonary mechanics, others could find no effect on gas exchange or pulmonary mechanics
(124,125,126). Side effects of the combination of a thiazide with spironolactone include azotemia,
hyperuricemia, hyponatremia, hyperkalemia or hypokalemia, hyperglycemia, hypercalcinuria, and
hypomagnesemia (118). Overall, although it appears that short-term administration of diuretics may improve
pulmonary mechanics in premature infants, there is limited data regarding the long-term benefits of these
agents in reducing the need for ventilatory support, reducing the length of hospitalization or improving longterm clinical outcome (127). Longer-term studies establishing optimal treatment regimens in infants with
evolving or established BPD are definitely needed. Diuretic dosing is shown in Table 30-4.
Bronchodilators
Albuterol is a specific 2-agonist that has become the inhaled agent of choice in the treatment of reversible
bronchospasm in infants with BPD. Albuterol aerosolization has been associated with acute improvements in
pulmonary resistance and lung compliance secondary to bronchial smooth muscle relaxation (128). These
changes in pulmonary mechanics returned to baseline by 4 hours after administration. Side effects are
infrequent but can include tachycardia and hypertension. Tolerance may develop with prolonged usage.
Atropine is a competitive inhibitor of acetylcholine. In the lung, atropine decreases mucus secretion and
transport in large airways and causes significant bronchial smooth muscle relaxation (129). Ipratropium bromide
is a related muscarinic antagonist that is a much more potent bronchodilator than atropine and has significantly
fewer side effects. In infants with BPD, ipratropium causes a significant improvement in pulmonary mechanics
that is similar to that seen after treatment with albuterol (130). The combination of ipratropium and albuterol
may be more effective than either agent alone (129,130). A selective 2 agent should initially be used in older
infants with BPD; ipratropium can be added if clinical improvement is not seen. Ipratropium can be used alone if
significant side effects from albuterol occur.
The methylxanthines (e.g., caffeine, theophylline) are routinely used to increase respiratory drive and reduce
the frequency of apnea in infants with apnea of prematurity (99). They are also used in the treatment of infants
with BPD. Measurements of pulmonary mechanics in infants with BPD have shown that caffeine and theophylline
can reduce pulmonary resistance and increase lung compliance, presumably through a direct bronchodilator
action (131,132). These agents act as mild diuretics and improve skeletal muscle and diaphragmatic
contractility. This is particularly important in chronically ventilated infants who may develop diaphragmatic
atrophy and fatigue. Improved skeletal muscle contractility may stabilize the chest wall and improve functional
residual capacity (133). These actions may facilitate successful weaning from mechanical ventilation. There may
be a synergistic effect if theophylline and a diuretic are used concurrently (124).
TABLE 30-4 COMMONLY USED MEDICATIONS FOR BRONCHOPULMONARY DYSPLASIA
Medication
Diuretics
Furosemide
Dosage
Chlorthiazide
0.5-2.0 mg/kg/dose i.v or p.o. bid (qd in infants <31

weeks postconceptual age)
5-20 mg/kg/dose i.v. or p.o. bid
Hydrochlorthiazide
Spironolactone
1-2 mg/kg/dose p.o. bid

1.5 mg/kg/dose p.o. bid
Inhaled agents
Albuterol
Ipratropium bromide
Systemic agents
Aminophylline (i.v.), theophylline (p.o.)
Caffeine citrate
Dexamethasone
aLD,
0.02-0.04 mL/kg/dose of a 0.5% solution diluted to 1-2

mL with half-normal or normal saline q4-6h
0.025-0.08 mg/kg diluted to 1.5-2.5 mL in half-normal or
normal saline q6h; doses up to 0.176 mg may be used
LD 5 mg/kg; MD 2 mg/kg/dose q8-12h; serum levels of 515 mg/L
LD 20 mg/kg; MD 5 mg/kg i.v. or p.o. q24h
0.1-0.2 mg/kg/day i.v. or p.o. divided q12h for 2-3 days,
decrease by 50% for 2-3 days, then taper another 50%
for 2-3 days and discontinue
loading dose; MD, maintenance dose.

P.591
The half-life of theophylline is 30 to 40 hours in newborns, and theophylline is metabolized primarily to caffeine
in the liver and excreted in the urine. Adverse reactions include gastrointestinal (GI) (e.g., gastroesophageal
reflux, diarrhea), central nervous system (e.g., agitation, seizures), CV (e.g., tachycardia, hypertension), and
endocrine (e.g., hyperglycemia) disturbances (118). The half-life of caffeine may be as long as 100 hours. It is
excreted unchanged in the urine. Side effects of caffeine are similar to those of theophylline, but are rarely
encountered. Caffeine is a safer drug with a wider therapeutic index and fewer side effects than theophylline and
may be a more appropriate adjunct in the treatment of apnea and BPD in preterm infants. Although all of these
agents appear to improve lung mechanics in infants with BPD short-term, it is unclear whether prolonged
therapy causes sustained clinical benefits. Older patients with BPD often have asthma, characterized by
intermittent and reversible airways obstruction during acute respiratory exacerbations, and seem to be highly
responsive to bronchodilator therapy.
Corticosteroids
Corticosteroids are synthesized by the adrenal cortex and are composed of mineralocorticoids, which affect fluid
and electrolyte balance, and glucocorticoids, which affect the metabolism of many tissues and possess potent
antiinflammatory properties (134). Dexamethasone is a synthetic corticosteroid that has been used in the
prevention and treatment of BPD. Dexamethasone has multiple pharmacologic effects, although the downregulation of the inflammatory cascade is thought to be primarily responsible for the improvements in
pulmonary function in infants with severe BPD. Clinical studies have consistently shown that steroids acutely
improve lung mechanics, gas exchange and reduce inflammatory cells and their products in tracheal samples of
patients with BPD (135,136,137). Despite these studies, there have been multiple experimental and clinical
studies that have raised concerns that excessive doses and prolonged use of corticosteroids can impair head
growth, neurodevelopmental outcome, lung structure and long-term survival (138,139). Additionally, a recent
multicenter study was halted prior to completion as a result of a high incidence of GI perforation (140). Steroids
(e.g., dexamethasone) should be selectively used at lower doses and tapered appropriately over shorter
durations (5-7 days) in ventilator dependent infants with severe, persistent lung disease. Discussions with the
family regarding potential risks and benefits should occur prior to the initiation of treatment (139). The use of
steroids should be delayed if possible until approximately one month of age. Other side effects include systemic
hypertension, hyperglycemia, cardiac hypertrophy, poor somatic growth, sepsis, intestinal bleeding, and
myocardial hypertrophy (141,142). Inhaled steroids have been studied, but the major effect was to decrease the
perceived need for the use of systemic steroids (143). Recognition that some premature newborns may have
adrenal insufficiency that could increase the risk for BPD prompted interest in the use of low dose cortisol
replacement therapy (48). A randomized multicenter study was initiated to examine this question, but even this
trial was recently halted as a result of excessive side effects in infants receiving low-dose hydrocortisone
supplementation.
Physical Therapy
Physical therapy may help overcome various types of motor deficits (144). Infants with RDS and BPD are at
increased risk for subsequent gross motor, fine motor, or cognitive developmental delays. To optimize
ventilation, these infants use neck extension and accessory muscles. This produces abnormal posture of the
neck, scapula, shoulder, and trunk. Efforts to reduce this abnormal posture and normalize tone should be
provided in conjunction with a physical therapist. Infants are first positioned in a more neutral alignment.
Strengthening of neck and trunk muscles follows this, although independent movements and exploration of the
environment through infant stimulation techniques can be performed. A pacifier is used to facilitate and
strengthen the suck reflex, especially when the infant is able to tolerate gavage feedings. When infants are fed
orally, coordination of breathing, sucking, and swallowing may be difficult. Positioning the infants in natural
flexion and using mandibular compression and cheek and upper palate stimulation may be helpful. Nasal oxygen
is often necessary to assist the infant in feeding without tiring.
At the time of discharge, a comprehensive home therapy program is implemented. Nursing needs and home
physical and occupational or speech therapy are ordered as necessary. Reevaluation at appropriate intervals is
scheduled in a neonatal high-risk follow-up program, with emphasis on the possible need for an early
intervention program. These treatment programs emphasize teaching parents specific handling, positioning, and
stimulation techniques. Normalizing muscle tone and posture and stimulating desired patterns of movements are
the goals of these therapies.
OUTCOME
Most neonates who develop BPD ultimately achieve normal lung function and thrive. However, this group of
neonates is at higher risk of dying in the first year of life

P.592
and developing significant long-term complications. During infancy, continued normal lung growth should result
in slow improvement of pulmonary function and weaning from oxygen therapy. Later in childhood, other
respiratory problems (e.g., reactive airway disease) and abnormal neurologic development are additional
complications that may occur and require careful follow-up. Interpreting outcome studies has become more
difficult as a result of the frequent introduction of new treatment modalities such as exogenous surfactant which
has significantly influenced neonatal outcome.
Mortality
In Northway's original group of 32 infants, only 13 survived the first month of life (1). Nine (69%) of these
survivors developed BPD and five died within the first year of life primarily of pulmonary hypertension and cor
pulmonale. The remaining four infants with BPD had persistent respiratory abnormalities that resolved slowly
over time. In sharp contrast to this 66% mortality rate, more recent studies have shown a significant
improvement in mortality which has continued to fall with the changing epidemiology of BPD. Several surfactant
replacement studies have demonstrated significantly improved survival, even in infants who develop BPD.
Although Phibbs found a 32% incidence of BPD in his surfactant treatment and control groups, there were no
deaths from BPD in the surfactant-treated group and only three deaths in the control group (145). Palta followed
a cohort of 533 infants weighing less than 1500 grams who were born between 1988 and 1990 and survived
until discharge (6). The incidence of BPD was 25% in this group. Only 9 patients (1.7%) died following hospital
discharge, but it is unclear whether these deaths were related to the development of BPD. Twenty-one percent
of all infants in this cohort had respiratory problems at 2 years corrected age, which correlated only weakly with
a diagnosis of BPD at 36 weeks postconceptual age. In a more recent study from Finland, 211 survivors from the
NICU weighing less than 1,000 grams were followed (146). Although 39% of neonates met criteria for a
diagnosis of BPD, only one patient died as a result of causes associated with BPD (mortality 1.2%) and none
died of sudden infant death syndrome (SIDS). More recent studies from the postsurfactant era suggest that BPD
has generally become less severe than seen in the past and is associated with a better outcome if it does
develop.
Cardiopulmonary Function
Although pulmonary function in most survivors with BPD improves over time with continued lung growth and
permits normal activity, abnormalities detected by pulmonary function testing may remain. Follow-up studies of
children with BPD have shown increased airway resistance and reactivity, decreased lung compliance, ventilationperfusion mismatch, and blood gas abnormalities (e.g., increased PCO2) that may continue into later years
(147,148). These abnormalities appear to correlate extremely well with the presence of clinical respiratory
symptoms such as wheezing. Although some authors have suggested that the pulmonary dysfunction is directly
related to the development of BPD, others have demonstrated that many mechanically ventilated premature
babies (with/without BPD) can have abnormal pulmonary function in later life. Blayney investigated patients with
BPD at 7 and 10 years of age and found that although lung growth had occurred normally, residual volumes
were increased and forced expiratory volumes and flow rates were reduced (149). Fifty percent of these children
had a history of wheezing, suggesting airway hyperactivity. Significant improvement in pulmonary function
occurred from years 7 through 10, indicating that pulmonary abnormalities from BPD persist well into childhood
and continue to improve slowly over time. Filippone and colleagues reported similar results from a cohort of
patients (<1250 grams at birth) with BPD who were followed up to 8-9 years of age (150). At 2 years of age,
66% had abnormalities on pulmonary function testing (diminished maximal flow at FRC). At school age only
16% reported any clinical difficulties, but abnormalities of pulmonary mechanics persisted with forced expiratory
volume (FEV1) reduced to 76% of predicted and forced mid-expiratory flow (FEF) (25-75) to 63%. Hakulinen
also found lower airway conductance and increased residual volumes in children who had BPD (151). The most
significant abnormalities were found in children who had clinical respiratory symptoms, especially early in their
childhood (<2 years of age). Although there was an increased need for hospitalization because of pulmonary
problems for the first 2 years of life in children with BPD, by 6 to 9 years of age, none had evidence of wheezing
or respiratory distress.
As the overall severity of BPD has decreased in recent years, so have the long-term clinical and cardiopulmonary
sequelae. Baraldi and colleagues demonstrated progressive improvements in pulmonary function tests over the
first 2 years of life, although evidence of airway dysfunction persisted (152). Mitchell and Teague studied
premature infants with and without BPD at 6 to 9 years of age (153). They found reduced soluble gas transfer at
rest and with exercise in children with BPD, explained by abnormal pulmonary architecture and/or right
ventricular dysfunction affecting cardiac output. It appears that abnormal cardiopulmonary function is greatest
in the first 2 years of life in infants with BPD. Survival beyond that age permits children to function at near
normal capacity, although persistent abnormalities can be detected into adolescence.
Infection
Increased susceptibility to infection has been documented in infants with BPD. Respiratory syncytial virus (RSV)
is a major pathogen that causes illness and the need for rehospitalization and mechanical ventilation in children
with BPD. Infants are more susceptible to RSV infection because of impaired lung defenses secondary to
damaged lung tissue (154,155). Monthly intramuscular injections of
P.593
monoclonal antibodies (palivuzumab) have been shown to decrease the incidence and severity of RSV infection
in high-risk premature infants by 55% (156). Indications for use include significantly premature infants who are
home and less than 6 months of age during RSV season (November-December to March-April), infants under 2
years of age with BPD, or any preterm infant with BPD who has received supplemental oxygen in the previous 6
months. Other viruses such as rhinovirus and influenza, although not as common as RSV, may also be an
important cause of lower respiratory tract disease in infants with BPD (157).
Growth and Neurologic Development

It is not surprising that the most critically ill and premature infants who develop BPD have an increased risk for
growth failure and abnormal neurodevelopmental outcome. Infants with BPD have increased metabolic demands
and caloric requirements and may grow poorly during infancy and childhood (158,159,160). Markestad observed
that children with improving respiratory function exhibited faster catch-up growth, and those with continued
respiratory problems failed to do so (161). Other studies have shown that a significant proportion of infants with
BPD are consistently in the lower percentiles for height, weight, and head circumference (162). This is frequently
found during the first 2 years of life, when respiratory symptoms and illness may be prominent (151). Huysman
found that although infants with BPD increased in length, total body fat and free fat mass, growth was
significantly less than term controls (163). Gregoire compared growth rates in extremely premature infants with
and without BPD and found that significant differences existed at birth, but had resolved by 18 months corrected
age (164).
Giacoia studied a group of preterm infants with BPD over the first 2 years of life and compared them to a group
of age-matched premature infants and term controls (165). Both groups of premature infants had comparable
growth rates and average mean standard scores for Wechsler IQ. However, all of the premature infants were
significantly smaller than the term controls and had lower scores for performance and full-scale IQ. Singer found
no difference in intelligence between infants with BPD and age-matched premature controls at 3 years of life,
but motor function was significantly impaired (166). Hack and associates also followed a group of extremely
premature infants who received exogenous surfactant at birth and found that BPD was among the most
significant predictors of an abnormal motor developmental index (MDI) (odds ratio (OR 2.18) and neurologic
abnormalities (OR 2.46) (167). Several more recent follow-up studies have confirmed that a diagnosis of BPD
continues to be associated with a higher risk for abnormal neurodevelopmental outcome in premature infants
(168,169). The relative contribution of BPD to poor outcome continues to be defined. However, infants
developing BPD are often the smallest, sickest patients in the neonatal intensive care unit and consequently can
develop other conditions such as periventricular leukomalacia, intraventricular hemorrhage, retinopathy of
prematurity and sepsis which also appear to be important variables in determining developmental outcome
(168).
The introduction of surfactant replacement therapy has resulted in significantly improved survival and the
development of less severe BPD. It is reassuring that follow-up studies have found that the number of children
with significant neurodevelopmental handicaps has not increased despite the survival of an increased number of
smaller, sicker infants who frequently develop BPD.
PREVENTION STRATEGIES
A multidisciplinary approach to the prevention of BPD in infants is needed. The use of antenatal steroids in
mothers at high risk of delivering a premature infant may reduce the severity and incidence of BPD (170). The
early use of NCPAP (with or without prior surfactant treatment) may eliminate the need for mechanical
ventilation in some premature infants or facilitate successful extubation in others (103). Aggressive treatment of
symptomatic patent ductus arteriosus with fluid restriction, diuretics, indomethacin (or possibly ibuprofen), or
surgical closure may reduce the severity of BPD. Tidal volumes and inspired oxygen concentrations should be
reduced as low as necessary to reduce hypocarbia, volutrauma, and oxygen toxicity. Using more aggressive lung
recruitment strategies, such as more liberal PEEP, HFV or perhaps, prone positioning may also be beneficial. The
early use of synchronized mechanical ventilation or high frequency ventilation (HFV) in newborn infants with
significant RDS may reduce the severity and incidence of BPD (20,21,22). The combined use of HFV and
surfactant replacement may prevent significant lung damage in premature and term infants with significant lung
disease unresponsive to surfactant replacement and conventional mechanical ventilation (171,172). Further
conclusive evidence of the beneficial effects of both synchronized ventilation and HFV in premature infants is
needed.
Aggressive nutritional support is critical in helping to promote normal lung growth, maturation, and repair. It
also protects the lung from the damaging effects of hyperoxia, infection, and barotraumas (173). Systemic
supplementation with vitamin A in sufficient quantities to establish normal serum retinol concentrations has been
reported to reduce the incidence of BPD, although the need to administer repeated doses through the
intramuscular route over a prolonged period of time has limited the widespread use of this therapy.
Numerous studies have demonstrated that the early use of dexamethasone may reduce the incidence of BPD,
presumably by treating cortisol deficiency and minimiz-ing inflammation (138). However, early dexamethasone
should not be used to prevent BPD because of significant concerns regarding increased mortality, side effects (e.
g., GI perforation) and long-term sequelae (CP) (139,140). A
P.594
promising method for preventing the development of BPD appears to be prophylactic supplementation of human
recombinant antioxidant enzymes. This seems to be a logical strategy in preventing BPD, because oxygen free
radicals appear to play a major role in the pathogenesis of lung injury and premature infants are known to be
relatively deficient in these enzymes at birth. Several animal studies have shown that prolonged exposure to
high oxygen concentrations can cause severe lung damage and death, and administration of antioxidants can
prevent many of these complications (174,175,176,177,178,179,180,181). Recombinant human CuZnSOD
(rhSOD) has been administered prophylactically to the lung of premature infants at high risk for developing BPD.
In preliminary studies in premature infants, the prophylactic use of both single and multiple intratracheal doses
of rhSOD appeared to mitigate inflammatory changes and severe lung injury from oxygen and mechanical
ventilation with no apparent associated toxicity (179,180). In animal studies, the rhSOD appeared to localize
both in intracellular and extracellular compartments following intratracheal instillation with significant quantities
of active protein present 48 hours after the dose is given (182). Multicenter trials using prophylactic,
intratracheal rhSOD in premature infants at high-risk for developing BPD have recently been completed (181).
Premature infants (birth weight 600-1,200 g) receiving intratracheal instillation of rhSOD at birth had
significantly (44%) less episodes of respiratory illness (wheezing, asthma, pulmonary infections) severe enough
to require treatment with bronchodilators or corticosteroids at a median of one year corrected age compared to
placebo controls (Fig. 30-8). The largest effects were seen in infants less than 27 weeks gestation, with
decreased episodes of respiratory illness accompanied by significant reductions (<50%) in emergency room
visits and hospital readmissions (Fig. 30-9). This suggests that rhSOD did prevent long-term pulmonary injury
from RDS in high-risk premature infants. Further therapeutic intervention trials are needed to ultimately develop
a therapy that can prevent
P.595
or significantly ameliorate this important chronic lung disease.
Figure 30-8 The use of asthma medications to treat significant respiratory illness in infants at 1-2 years of age
from the multicenter trial of rhSOD in preterm infnats. The entire group is presented and a subset of infants
<27 weeks gestation at birth. The open bars represent the percent of treated infants in the placebo group and
the shaded bars the r-h CuZnSOD group (*p = 0.05; **p < 0.01). From Davis JM, Parad RB, Michele T, et al.
Pulmonary outcome at 1 year corrected age in premature infants treated at birth with recombinant human
CuZn superoxide dismutase. Pediatrics 2003;111: 469, with permission.
Figure 30-9 The number of emergency room visits and hospital admissions (all causes) in a subset of infants
<27 weeks gestation at birth who had received placebo (open bars) or rhSOD (shaded bars) (*p = 0.05; **p <
0.01). From Davis JM, Parad RB, Michele T, et al. Pulmonary outcome at 1 year corrected age in premature
infants treated at birth with recombinant human CuZn superoxide dismutase. Pediatrics 2003;111:469, with
permission.
Another area that is currently being investigated involves protecting the pulmonary vasculature from injury with
inhaled NO (INO). INO may not only lower PVR and improve gas exchange, but may also enhance distal lung
growth and improve long-term outcome. Extensive laboratory and clinical studies suggest that INO lowers PVR
and improves oxygenation in patients with pulmonary hypertension, including premature infants with severe
RDS and BPD (183,184). However, there are persistent concerns about potential toxicity and adverse effects of
INO therapy in premature infants (185). Experimental data have suggested that INO therapy may be lung
protective in several animal models, including premature lambs with RDS (186). Whether INO therapy has a
potential role in the prevention of pulmonary vascular injury in premature newborns at risk for BPD is unknown.
A multicenter clinical trial of low-dose INO therapy (5 ppm) was recently performed in severely ill premature
newborns with RDS who had marked hypoxemia despite surfactant therapy (183). In this study, INO acutely
improved PaO2, but did not improve survival. Notably, there was no increase in the frequency or severity of
intracranial hemorrhage or BPD, and the duration of mechanical ventilation was reduced. Based on these
findings, a multicenter trial is now underway to determine whether early treatment with low dose INO therapy
will prevent the early inflammatory changes that contribute to BPD, and protect the pulmonary circulation from
injury during this critical time period.
REFERENCES
1. Northway WH Jr, Rosan RC, Porter DY. Pulmonary disease following respirator therapy of hyalinemembrane disease. N Engl J Med 1967;276:357.
2. Bancalari E, Abdenour GE, Feller R, et al. Bronchopulmonary dysplasia: clinical presentation. J Pediatr
1979;95:819.
3. Avery ME, Tooley WH, Keller JB, et al. Is chronic lung disease in low birth weight infants preventable? A
survey of eight centers. Pediatrics 1987;79:26.
4. Sinkin RA, Cox C, Phelps DL. Predicting risk for bronchopulmonary dysplasia: selection criteria for clinical
trials. Pediatrics 1990;86:728.
5. Shennan AT, Dunn MS, Ohlsson A, et al. Abnormal pulmonary outcomes in premature infants: prediction
from oxygen requirement in the neonatal period. Pediatrics 1988;82:527.
6. Palta M, Sadek M, Barnet JH, et al. Evaluation of criteria for chronic lung disease in surviving very infants.
Newborn Lung Project. J Pediatr 1998;132:57.
7. Jobe AH, Bancalari E. Bronchopulmonary Dysplasia. Am J Resp Crit Care Med 2001;163:1723.
8. Soll RF, Hoekstra RE, Fangman JJ, et al. Multicenter trial of single-dose modified bovine surfactant extract
(Survanta) for prevention of respiratory distress syndrome. Pediatrics 1990;85: 1092.
9. Bose C, Corbet A, Bose G, et al. Improved outcome at 28 days of age for very low birth weight infants
treated with a single dose of a synthetic surfactant. J Pediatr 1990;117:947.
10. Liechty EA, Donovan E, Purohit D, et al. Reduction of neonatal mortality after multiple doses of bovine
surfactant in low birth weight neonates with respiratory distress syndrome. Pediatrics 1991;88:19.
11. Long W, Thompson T, Sundell H, et al. Effects of two rescue doses of a synthetic surfactant on mortality
rate and survival without bronchopulmonary dysplasia in 700- to 1350-gram infants with respiratory distress
syndrome. The American Exosurf Neonatal Study Group I. J Pediatr 1991;118:595.
12. Fenton AC, Mason E, Clarke M, et al. Chronic lung disease following neonatal ventilation. Changing
incidence in a geographically defined population. Pediatr Pulmonol 1996;21:24.
13. Dreyfus D, Saumon G. Should the lung be rested or recruited? Am J Respir Crit Care Med 1994;149:1066.
14. Dreyfuss DD, Saumon G. Ventilator induced lung injury: lessons from experimental studies. Am J Respir
Crit Care Med 1998;157: 294.
15. Nilsson R, Grossman G, Robertson B. Lung surfactant and the pathogenesis of neonatal bronchiolar lesions
induced by artificial ventilation. Pediatr Res 1978;12:249.
16. Hernandez LA, Peevy KJ, Moise AA, et al. Chest wall restriction limits high airway pressure induced lung
injury in young rabbits. J Appl Physiol 1989;66:2364.
17. Tremblay L, Valenza F, Ribeiro SP, et al. Injurious ventilatory strategies increase cytokines and c-fos mRNA expression in an isolated rat lung model. J Clin Invest 1997;99:944.
18. Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with
traditional tidal volumes for acute lung injury and the ARDS. N Engl J Med 2000; 342:1301.
19. Bjorklund LJ, Ingimarsson J, Curstedt T, et al. Manual ventilation with a few large breaths at birth
compromises the therapeutic effect of subsequent surfactant replacement in immature lungs. Pediatr Res
1997;42:348.
20. Bernstein G, Mannino FL, Heldt GP, et al. Randomized multicenter trial comparing synchronized and
conventional intermittent mandatory ventilation in neonates. J Pediatr 1996;128:453.
21. Clark RH, Gerstmann DR, Null DM, et al. Prospective, randomized comparison of high-frequency oscillatory
and conventional ventilation in respiratory distress syndrome. Pediatrics 1992;89:5.
22. Keszler M, Modanlou HD, Brudno DS, et al. Multicenter controlled clinical trial of high-frequency jet
ventilation in preterm infants with uncomplicated respiratory distress syndrome. Pediatrics 1997;100:593.
23. Garland JS, Buck RK, Allred EN, et al. Hypocarbia before surfactant therapy appears to increase
bronchopulmonary dysplasia risk in infants with respiratory distress syndrome. Arch Pediatr Adolesc Med
1995;149:617.
24. Collins MP, Lorenz JM, Jetton JR, et al. Hypocapnia and other ventilation-related risk factors for cerebral
palsy in low birth weight infants. Pediatr Res 2001;50:712.
25. Van Marter LJ, Allred EN, Pagano M, et al. Do clinical markers of barotrauma and oxygen toxicity explain
interhospital variation in rates of chronic lung disease? The Neonatology Committee for the Developmental
Network. Pediatrics 2000;105:1194.
26. Frank L, Groseclose EE. Preparation for birth into an O2-rich environment: the antioxidant enzymes in the
developing rabbit lung. Pediatr Res 1984;18:240.
27. Ogihara T, Okamoto R, Kim HS, et al. New evidence for the involvement of oxygen radicals in triggering
neonatal chronic lung disease. Pediatr Res 1996;39:117.
28. Pitkanen OM, Hallman M, Andersson SM. Correlation of free oxygen radical-induced lipid peroxidation with
outcome in very low birth weight infants. J Pediatr 1990;116:760.
29. Varsila E, Pesonen E, Andersson S. Early protein oxidation in the neonatal lung is related to the
development of chronic lung disease. Acta Paediatr 1995;84:1296.
30. Banks BA, Ischiropoulos H, McClelland M, et al. Plasma 3-nitrotyrosine is elevated in premature infants
who develop bronchopulmonary dysplasia. Pediatrics 1998;101:870.
31. The STOP-ROP Multicenter Study Group. Supplemental therapeutic oxygen for prethreshold retinopathy of
prematurity, a randomized, controlled trial. Pediatrics 2000;105:295.
32. Davis JM, Rosenfeld WN, Sanders RJ, et al. Prophylactic effects of recombinant human superoxide
dismutase in neonatal lung injury. J Appl Physiol 1993;74:2234.
33. Padmanabhan RV, Gudapaty R, Liener IE, et al. Protection against pulmonary oxygen toxicity in rats by
the intratracheal administration of liposome-encapsulated superoxide dismutase or catalase. Am Rev Respir
Dis 1985;132:164.
34. Carlsson LM, Jonsson J, Edlun T, et al. Mice lacking extracellular superoxide dismutase are more sensitive
to hyperoxia. Proc Natl Acad Sci U S A 1995;92:6264.
P.596
35. Wispe JR, Warner BB, Clark JC, et al. Human Mn-superoxide dismutase in pulmonary epithelial cells of
transgenic mice confers protection from oxygen injury. J Biol Chem 1992;267:23937.
36. Brus F, van Oeveren W, Okken A, et al. Activation of circulating polymorphonuclear leukocytes in preterm
infants with severe idiopathic respiratory distress syndrome. Pediatr Res 1996; 39:456.
37. Buss IH, Senthilmohan R, Darlow BA, et al. 3-Chlorotyrosine as a marker of protein damage by
myeloperoxidase in tracheal aspirates from preterm infants: association with adverse respiratory outcome.
38. Groneck P, Speer CP. Inflammatory mediators and bronchopulmonary dysplasia. Arch Dis Child Fetal
39. Jones CA, Cayabyab RG, Kwong KY, et al. Undetectable interleukin (IL)-10 and persistent IL-8 expression
early in hyaline membrane disease: a possible developmental basis for the predisposition to chronic lung
inflammation in preterm newborns. Pediatr Res 1996;39:966.
40. Pierce MR, Bancalari E. The role of inflammation in the pathogenesis of bronchopulmonary dysplasia.
Pediatr Pulmonol 1995; 19:371.
41. Munshi UK, Niu JO, Siddiq MM, et al. Elevation of interleukin-8 and interleukin-6 precedes the influx of
neutrophils in tracheal aspirates from preterm infants who develop bronchopulmonary dysplasia. Pediatr
Pulmonol 1997;24:331.
42. Ramsay PL, O'Brian SE, Hegemier S, et al. Early clinical markers for the development of
bronchopulmonary dysplasia: soluble E-Selectin and ICAM-1. Pediatrics 1998;102:927.
43. Groneck P, Gotz-Speer B, Opperman M, et al. Association of pulmonary inflammation and increased
microvascular permeability during the development of bronchopulmonary dysplasia: a sequential analysis of
inflammatory mediators in respiratory fluids of high-risk neonates. Pediatrics 1994;93:712.
44. Alnahhas MH, Karathanasis P, Kriss VM, et al. Elevated laminin concentrations in lung secretions of
preterm infants supported by mechanical ventilation are correlated with radiographic abnormalities. J Pediatr
1997;131:555.
45. Ossanna PJ, Test ST, Matheson NR, et al. Oxidative regulation of neutrophil elastase-alpha-1-proteinase
inhibitor interactions. J Clin Invest 1986;77:1939.
46. Walti H, Tordet C, Gerbaut L, et al. Persistent elastase/proteinase inhibitor imbalance during prolonged
ventilation of infants with bronchopulmonary dysplasia: evidence for the role of nosocomial infections. Pediatr
Res 1989;26:351.
47. Stiskal JA, Dunn MS, Shennan AT, et al. alpha1-Proteinase inhibitor therapy for the prevention of chronic
lung disease of prematurity: a randomized, controlled trial. Pediatrics 1998; 101:89.
48. Watterberg KL, Scott SM, Backstrom C, et al. Links between early adrenal function and respiratory
outcome in preterm infants: airway inflammation and patent ductus arteriosus. Pediatrics 2000;105:320.
49. Watterberg KL, Gerdes JS, Gifford KL, et al. Prophylaxis against early adrenal insufficiency to prevent
chronic lung disease in premature infants. Pediatrics 1999;104:1258.
50. Yoon BH, Romero R, Jun JK, et al. Amniotic fluid cytokines (interleukin-6, tumor necrosis factor-alpha,
interleukin-1 beta, and interleukin-8) and the risk for the development of bronchopulmonary dysplasia. Am J
51. Watterberg KL, Demers LM, Scott SM, et al. Chorioamnionitis and early lung inflammation in infants in
whom bronchopulmonary dysplasia develops. Pediatrics 1996;97:210.
52. Cassell GH, Waites KB, Crouse DT, et al. Association of Ureaplasma urealyticum infection of the lower
respiratory tract with chronic lung disease and death in very-low-birth-weight infants. Lancet 1988;2:240.
53. Da Silva O, Gregson D, Hammerberg O. Role of Ureaplasma urealyticum and Chlamydia trachomatis in
development of bronchopulmonary dysplasia in very low birth weight infants. Pediatr Infect Dis J 1997;16:364.
54. Heggie AD, Jacobs MR, Butler VT, et al. Frequency and significance of isolation of Ureaplasma urealyticum
and Mycoplasma hominis from cerebrospinal fluid and tracheal aspirate specimens from low birth weight
infants. J Pediatr 1994;124:956.
55. Gonzalez A, Sosenko IRS, Chandar J, et al. Influence of infection or patent ductus arteriosus and chronic
lung disease in premature infants weighing 1000 grams or less. J Pediatr 1996;128:470.
56. Stoll BJ, Fanaroff AA, Wright LL, et al. Late-onset sepsis in very low birth weight neonates: the experience
of the NICHD network. Pediatrics 2002;110:285.
57. Makhoul IR, Sujov P, Smolkin T, et al. Epidemiological, clinical and microbiological characteristics of lateonset sepsis among very low birth weight infants in Israel: a national survey. Pediatrics 2002;109:134.
58. Frank L, Groseclose E. Oxygen toxicity in newborn rats: the adverse effects of undernutrition. J Appl
Physiol 1982;53:1248.
59. O'Dell BL, Kilburn KH, McKenzie WN, et al. The lung of the copper-deficient rat: a model for developmental
pulmonary emphysema. Am J Pathol 1978;91:413.
60. Darlow BA, Winterbourne CC, Inder TE, et al. The effect of selenium supplementation on outcome in very
low birth weight infants: A randomized controlled trial. J Pediatr 2000;136:473.
61. Vyas JR, Currie A, Dunster C, et al. Ascorbate acid concentration in airways lining fluid from infants who
develop chronic lung disease of prematurity. Eur J Pediatr 2001;160:177.
62. Berger TM, Frei B, Rifai N, et al. Early high dose antioxidant vitamins do not prevent bronchopulmonary
dysplasia in premature baboons exposed to prolonged hyperoxia: a pilot study. Pediatr Res 1998;43:719.
63. Shenai JP, Rush MG, Stahlman MT, et al. Plasma retinol-binding protein response to vitamin A
administration in infants susceptible to bronchopulmonary dysplasia. J Pediatr 1990;116:607.
64. Darlow BA, Graham PJ. Vitamin A supplementation for preventing morbidity and mortality in very low birth
weight infants. Cochrane Database Syst Rev 2002;4:CD00051.
65. Hustead VA, Gutcher GR, Anderson SA, et al. Relationship of vitamin A (retinol) status to lung disease in
the preterm infants. J Pediatr 1984;105:610.
66. Anzano MA, Olson JA, Lamb AJ. Morphologic alterations in the trachea and the salivary gland following the
induction of rapid synchronous vitamin A deficiency in rats. Am J Pathol 1980; 98:717.
67. Shenai JP. Vitamin A supplementation in very low birth weight neonates: rationale and evidence.
68. Hartline JV, Zachman RD. Vitamin A delivery in total parenteral nutrition solution. Pediatrics 1976;58:448.
69. Tyson JE, Wright LL, Oh W, et al. Vitamin A supplementation for extremely-low-birth-weight infants.
National Institute of Child Health and Human Development Neonatal Research Network. New Engl J Med
1999;340:1962.
70. Van Marter LJ, Leviton A, Allred EN, et al. Hydration during the first days of life and the risk of
bronchopulmonary dysplasia in low birth weight infants. J Pediatr 1990;116:942.
71. Bancalari E, Sosenko I. Pathogenesis and prevention of neonatal chronic lung disease: recent
developments. Pediatr Pulmonol 1990;8:109.
72. Nickerson BG, Taussig LM. Family history of asthma in infants with bronchopulmonary dysplasia. Pediatrics
1980;65:1140.
73. Bertrand JM, Riley SP, Popkin J, et al. The long-term pulmonary sequelae of prematurity: the role of
familial airway hyperreactivity and the respiratory distress syndrome. N Engl J Med 1985; 312:742.
74. Clark DA, Pincus LG, Oliphant M, et al. HLA-A2 and chronic lung disease in neonates. JAMA
1982;248:1868.
75. Hagan R, Minutillo C, French N, et al. Neonatal chronic lung disease, oxygen dependency, and a family
history of asthma. Pediatr Pulmonol 1995;20:277.
76. Toce SS, Farrell PM, Leavitt LA, et al. Clinical and radiographic scoring systems for assessing
bronchopulmonary dysplasia. Am J Dis Child 1984;138:581.
77. Edwards DK. Radiographic aspects of bronchopulmonary dysplasia. J Pediatr 1979;95:823.
78. Mortensson W, Andreasson B, Lindroth M, et al. Potential of early chest roentgen examination in ventilator
treated newborn infants to predict future lung function and disease. Pediatr Radiol 1989;20:41.
79. Weinstein MR, Peters ME, Sadek M, et al. A new radiographic scoring system for bronchopulmonary
dysplasia. Pediatr Pulmonol 1994;18:284.
80. Oppenheim C, Mamou-Mani T, Sayegh N, et al. Bronchopulmonary dysplasia: value of CT in identifying

pulmonary sequelae. Am J Roentgenol 1994;163:169.
81. Abman SH. Pulmonary hypertension in chronic lung disease of infancy. Pathogenesis, pathophysiology and
treatment. In: Bland RD, Coalson JJ, eds. Chronic lung disease of infancy. New York: Marcel Dekker,
2000;619.
82. Jones R, Zapol WM, Reid LM. Oxygen toxicity and restructuring of pulmonary arteries: a morphometric
study. Am J Pathol 1985; 121:212.
P.597
83. Abman SH, Wolfe RR, Accurso FJ, et al. Pulmonary vascular response to oxygen in infants with severe
BPD. Pediatrics 1985; 75:80.
84. Abman SH, Sondheimer HS. Pulmonary circulation and cardiovascular sequelae of BPD. In: Weir EK,
Archer SL, Reeves JT, eds. Diagnosis and treatment of pulmonary hypertension. Mt. Kisco: Futura Publishing,
1992:155.
85. Abman SH. Monitoring cardiovascular function in infants with chronic lung disease of prematurity. Arch Dis
Child Fetal Neonatal Ed 2002;87:F15.
86. Lui K, Lloyd J, Ang E, et al. Early changes in respiratory compliance and resistance during the
development of bronchopulmonary dysplasia in the era of surfactant therapy. Pediatr Pulmonol 2000;30:282.
87. Tepper RS, Morgan WJ, Cota K, et al. Expiratory flow limitation in infants with bronchopulmonary
dysplasia. J Pediatr 1986; 109:1040.
88. Greenspan JS, DeGiulio PA, Bhutani VK. Airway reactivity as determined by a cold air challenge in infants
with bronchopulmonary dysplasia. J Pediatr 1989;114:452.
89. McCann EM, Goldman SL, Brady JP. Pulmonary function testing in the sick newborn infant. Pediatr Res
1987;21:313.
90. Gerhardt T, Bancalari E. Lung function in bronchopulmonary dysplasia. In: Bancalari E, Stocker JT, eds.
Bronchopulmonary dysplasia. Washington, DC: Hemisphere Publishing, 1988:182.
91. Nickerson BG, Durand DJ, Kao LC. Short-term variability of pulmonary function tests in infants with
bronchopulmonary dysplasia. Pediatr Pulmonol 1989;6:36.
92. Hanrahan JP, Tager IB, Castile RG, et al. Pulmonary function measures in healthy infants. Variability and
size correction. Am Rev Respir Dis 1990;141:1127.
93. Cherukupalli K, Larson JE, Rotschild A, et al. Biochemical, clinical, and morphologic studies on lungs of
infants with bronchopulmonary dysplasia. Pediatr Pulmonol 1996;22:215.
94. Jobe AH. The new BPD: an arrest of lung development. Pediatr Res 1999;46:641.
95. Bhatt AJ, Pryhuber GS, Huyck H, et al. Disrupted pulmonary vasculature and decreased vascular
endothelial growth factor, Flt-1, and TIE-2 in human infants dying with bronchopulmonary dysplasia. Am J
Respir Crit Care Med 2001;164:1971.
96. Courtney SE, Weber KR, Breakie LA, et al. Capillary blood gases in the neonate. Am J Dis Child
1990;144:168.
97. Davis JM, Dickerson B, Metlay L, et al. Differential effects of oxygen and barotrauma on lung injury in the
neonatal piglet. Pediatr Pulmonol 1991;10:157.
98. Tarnow-Mordi WO, Reid E, Griffiths P, et al. Low inspired gas temperature and respiratory complications in
very low birth weight infants. J Pediatr 1989;114:438.
99. Aranda JV, Turmen T. Methylxanthines in apnea of prematurity. Clin Perinatol 1979;6:87.
100. Veness-Meehan KA, Richter S, Davis JM. Pulmonary function testing prior to extubation in infants with
respiratory distress syndrome. Pediatr Pulmonol 1990;9:2.
101. Kim EH. Successful extubation of newborn infants without preextubation trial of continuous positive
airway pressure. J Perinatol 1989;9:72.
102. Viscardi RM, Faix RG, Nicks JJ, et al. Efficacy of theophylline for prevention of post-extubation respiratory
failure in very low birth weight infants. J Pediatr 1985;107:469.
103. Higgins RD, Richter SE, Davis JM. Nasal continuous positive airway pressure facilitates extubation of very
low birth weight neonates. Pediatrics 1991;88:999.
104. Miller RW, Woo P, Kelman RK, et al. Tracheobronchial abnormalities in infants with bronchopulmonary
dysplasia. J Pediatr 1987;111:779.
105. Goodman G, Perkin RM, Anas NG, et al. Pulmonary hypertension in infants with bronchopulmonary
106. Morin FC, Davis JM. Persistent pulmonary hypertension. In: Spitzer AR, ed. Intensive care of the fetus
and newborn. St Louis: CV Mosby, 1996:506.
107. Ramanathan R, Durand M, Larrazabal C. Pulse oximetry in very low birth weight infants with acute and
chronic lung disease. Pediatrics 1987;79:612.
108. Dimaguila MA, Di Fiore JM, Martin RJ, et al. Characteristics of hypoxemic episodes in very low birth
weight infants on ventilatory support. J Pediatr 1997;130:577.
109. McEvoy C, Mendoza ME, Bowling S, et al. Prone positioning decreases episodes of hypoxemia in
extremely low birth weight infants with chronic lung disease. J Pediatr 1997;130:305.
110. Higgins RD, Phelps DL. Oxygen-induced retinopathy: lack of adverse heparin effect. Pediatr Res
1990;27:580.
111. Simoes EAF, Rosenberg AA, King SJ, et al. Room air challenge: prediction for successful weaning of
oxygen-dependent infants. J Perinatol 1997;17:125
112. Alverson DC, Isken VH, Cohen RS. Effect of booster transfusion on oxygen utilization in infants with
113. Messer J, Haddad J, Donato L, et al. Early treatment of premature infants with recombinant human
erythropoietin. Pediatrics 1993;92:519.
114. Gilhooly J, Lindenberg J, Reynolds JW. Central venous silicone elastomer catheter placement by basilic
vein cutdown in neonates. Pediatrics 1986;78:636.
115. Hammerman C, Aramburo MJ. Decreased lipid intake reduces morbidity in sick premature infants. J
Pediatr 1988;113:1083.
116. Sosenko IR, Rodriguez-Pierce M, Bancalari E. Effect of early initiation of intravenous lipid administration
on the incidence and severity of chronic lung disease in premature infants. J Pediatr 1993;123:975.
117. Yunis KA, Oh W. Effects of intravenous glucose loading on oxygen consumption, carbon dioxide
production, and resting energy expenditure in infants with bronchopulmonary dysplasia. J Pediatr
1989;115:127.
118. Davis JM, Sinkin RA, Aranda JV. Drug therapy for bronchopulmonary dysplasia. Pediatr Pulmonol
1990;8:117.
119. Bland RD, McMillan DD, Bressack MA. Decreased pulmonary transvascular fluid filtration in awake
newborn lambs after intravenous furosemide. J Clin Invest 1978;62:601.
120. Engelhardt B, Elliott S, Hazinski TA. Short- and long-term effects of furosemide on lung function in
infants with bronchopulmonary dysplasia. J Pediatr 1986;109:1034.
121. Rush MG, Engelhardt B, Parker RA, et al. Double-blind, placebo-controlled trial of alternate-day
furosemide therapy in infants with chronic bronchopulmonary dysplasia. J Pediatr 1990;117:112.
122. Brion LP, Primhak RA, Yong W. Aerosolized diuretics for preterm infants with (or developing) chronic lung
disease. Cochrane Database Syst Rev 2001;2:CD001694.
123. Weiner IM, Mudge GH. Diuretics and other agents employed in the mobilization of edema fluid. In:
Gilman AG, Goodman LS, Rall TW, et al, eds. The pharmacological basis of therapeutics. New York: Macmillan,
1985:887.
124. Kao LC, Durand DJ, McCrea RC, et al. Randomized trial of long-term diuretic therapy for infants with
oxygen-dependent bronchopulmonary dysplasia. J Pediatr 1994;124:772.
125. Kao LC, Durand DJ, Phillips BL, et al. Oral theophylline and diuretics improve pulmonary mechanics in
infants with bronchopulmonary dysplasia. J Pediatr 1987;111:439.
126. Engelhardt B, Blalock WA, DonLevy S, et al. Effect of spironolactone-hydrochlorothiazide on lung function
in infants with chronic bronchopulmonary dysplasia. J Pediatr 1989;114:619.
127. Brion LP, Primhak RA, Ambrosio-Perez I. Diuretics acting on the distal renal tubule for preterm infants
with (or developing) chronic lung disease. Cochrane Database Syst Rev 2002;1: CD001817.
128. Wilkie RA, Bryan MH. Effect of bronchodilators on airway resistance in ventilator-dependent neonates
with chronic lung disease. J Pediatr 1987;111:278.
129. Weiner N. Atropine, scopolamine, and related antimuscarinic drugs. In: Gilman AG, Goodman LS, Rall
TW, et al, eds. The pharmacologic basis of therapeutics. New York: Macmillan, 1985:130.
130. Brundage KL, Mohsini KG, Froese AB, et al. Bronchodilator response to ipratropium bromide in infants
with bronchopulmonary dysplasia. Am Rev Respir Dis 1990;142:1137.
131. Davis JM, Bhutani VK, Stefano JL, et al. Changes in pulmonary mechanics following caffeine
administration in infants with bronchopulmonary dysplasia. Pediatr Pulmonol 1989;6:49.
132. Rooklin AR, Moomjian AS, Shutack JG, et al. Theophylline therapy in bronchopulmonary dysplasia. J
Pediatr 1979;95:882.
133. Polgar G. Mechanical properties of the lung and chest wall. In: Thibeault DW, Gregory GA, eds. Neonatal
pulmonary care. Norwalk: Appleton-Century-Crofts, 1986:49.
P.598
134. Haynes RC, Murad F. Adrenocorticotropic hormone; adrenocortical steroids and their synthetic analogs:
inhibitors of adrenocortical steroid biosynthesis. In: Gilman AG, Goodman LS, Rall TW, et al, eds. The
pharmacological basis of therapeutics. New York: Macmillan, 1985:1459.
135. Cummings JJ, D'Eugenio DB, Gross SJ. A controlled trial of dexamethasone in preterm infants at high risk
for bronchopulmonary dysplasia. N Engl J Med 1989;320:1505.
136. Wang JY, Yeh TF, Lin YJ, et al. Early postnatal dexamethasone therapy may lessen lung inflammation in
premature infants respiratory distress syndrome on mechanical ventilation. Pediatr Pulmonol 1997;23:193.
137. Yoder MC Jr, Chua R, Tepper R. Effect of dexamethasone on pulmonary inflammation and pulmonary
function of ventilator-dependent infants with bronchopulmonary dysplasia. Am Rev Respir Dis 1991;143:1044.
138. Halliday HL, Ehrenkranz RA, Doyle LW. Early post-natal steroids for preventing chronic lung disease in
preterm infants. Cochrane Database Syst Rev 2003;1:CD001146.
139. Postnatal corticosteroids to treat or prevent chronic lung disease in preterm infants. Pediatrics
2002;109:330.
140. Stark AR, Carlo WA, Tyson JE, et al. Adverse effects of early dexamethasone in extremely-low-birthweight infants. NICHD neonatal research network. N Engl J Med 2001;344:95.
141. Marinelli KA, Burke GS, Herson VC. Effects of dexamethasone on blood pressure in premature infants
with bronchopulmonary dysplasia. J Pediatr 1997;130:594.
142. Rizvi ZB, Aniol HS, Myers TF, et al. Effects of dexamethasone on the hypothalamic-pituitary-adrenal axis
in preterm infants. J Pediatr 1992;120:961.
143. Cole CH, Colton T, Shah BL, et al. Early inhaled glucocorticoid therapy to prevent bronchopulmonary
dysplasia. N Engl J Med 1999;340:1005.
144. Parker A. Expert handling. Nurs Times 1990;86:35.
145. Phibbs RH, Ballard RA, Clements JA, et al. Initial clinical trial of Exosurf, a protein-free synthetic
surfactant, for the prophylaxis and early treatment of hyaline membrane disease. Pediatrics 1991;88:1.
146. Tommiska V, Heinonen K, Kero P, et al. A national two year follow up study of extremely low birthweight
infants born in 1996- 1997. Arch Dis Child Fetal Neonatal Ed 2003;88:F29.
147. Bader D, Ramos AD, Lew CD, et al. Childhood sequelae of infant lung disease: exercise and pulmonary
function abnormalities after bronchopulmonary dysplasia. J Pediatr 1987;110:693.
148. Andreasson B, Lindroth M, Mortensson W, et al. Lung function eight years after neonatal ventilation. Arch
Dis Child 1989;64:108.
149. Blayney M, Kerem E, Whyte H, et al. Bronchopulmonary dysplasia: improvement in lung function
between 7 and 10 years of age. J Pediatr 1991;118:201.
150. Filippone M, Sartor M, Zacchello F, et al. Flow limitation in infants with bronchopulmonary dysplasia and
respiratory function at school age. Lancet 2003;361:753.
151. Hakulinen AL, Heinonen K, Lansimies E, et al. Pulmonary function and respiratory morbidity in school-age
children born prematurely and ventilated for neonatal respiratory insufficiency. Pediatr Pulmonol 1990;8:226.
152. Baraldi E, Filippone M, Trevisanuto D, et al. Pulmonary function until two years of life in infants with
bronchopulmonary dysplasia. Am J Respir Crit Care Med 1997;155:149.
153. Mitchell SH, Teague WG. Reduced gas transfer at rest and during exercise in school-age survivors of
bronchopulmonary dysplasia. Am J Respir Crit Care Med 1998;157:1406.
154. Weisman LE. Populations at risk for developing respiratory syncytial virus and risk factors for respiratory
syncytial virus severity: infants with predisposing conditions. Pediatr Infect Dis J 2003; 22:S33-S37.
155. Groothuis JR, Gutierrez KM, Lauer BA. Respiratory syncytial virus infection in children with
156. The IMpact-RSV Study Group. Palivizumab, a humanized respiratory syncytial virus monoclonal antibody,
reduces hospitalization from respiratory syncytial virus infection in high-risk infants. Pediatrics 1998;102:531.
157. Chidekel AS, Rosen CL, Bazzy AR. Rhinovirus infection associated with serious lower respiratory illness in
patients with bronchopulmonary dysplasia. Pediatr Infect Dis J 1997;16:43.
158. Kalhan SC, Denne SC. Energy consumption in infants with bronchopulmonary dysplasia. J Pediatr
1990;116:662.
159. Kao LC, Durand DJ, Nickerson BG. Improving pulmonary function does not decrease oxygen consumption
in infants with bronchopulmonary dysplasia. J Pediatr 1988;112:616.
160. Kurzner SI, Garg M, Bautista DB, et al. Growth failure in infants with bronchopulmonary dysplasia:
nutrition and elevated resting metabolic expenditure. Pediatrics 1988;81:379.
161. Markestad T, Fitzhardinge PM. Growth and development in children recovering from bronchopulmonary
162. Yu VYH, Orgill AA, Lim SB, et al. Growth and development of very low birth weight infants recovering
from bronchopulmonary dysplasia. Arch Dis Child 1983;58:791.
163. Huysman WA, de Ridder M, de Bruin NC, et al. Growth and body composition in preterm infants with
bronchopulmonary dysplasia. Arch Dis Child Fetal Neonatal Ed 2003;88:F46.
164. Gregiore MC, Lefebvre F, Glorieux J. Health and developmental outcomes at 18 months in very preterm
infants with bronchopulmonary dysplasia. Pediatrics 1998;101:856.
165. Giacoia GP, Venkataraman PS, West-Wilson KI, et al. Follow-up of school-age children with
166. Singer L, Yamashita T, Lilien L, et al. A longitudinal study of developmental outcome of infants with
bronchopulmonary dysplasia and very low birth weight. Pediatrics 1997;100:987.
167. Hack M, Wilson-Costello D, Friedman H, et al. Neurodevelopment and predictors of outcomes of children
with birth weights of less than 1,000 grams; 1992-1995. Arch Pediatr Adolesc Med 2000;154:725.
168. Schmidt B, Asztalos EV, Roberts RS, et al. Trial of Indomethacin Prophylaxis in Preterms (TIPP)
Investigators. Impact of bronchopulmonary dysplasia, brain injury, and severe retinopathy on the outcome of
extremely low-birth-weight infants at 18 months: results from the trial of indomethacin prophylaxis in
preterms. JAMA 2003;289:1124.
169. Hanke C, Lohaus A, Gawrilow C, et al. Preschool development of very low birth weight children born
1994-1995. Eur J Pediatr 2003;162:159.
170. Van Marter LJ, Leviton A, Kuban KC, et al. Maternal glucocorticoid therapy and reduced risk of
171. Soll RF, Dargaville P. Surfactant for meconium aspiration syndrome in full term infants. Cochrane
Database Syst Rev 2000;2: CD002054.
172. Davis JM, Richter SE, Kendig JW, et al. High frequency jet ventilation and surfactant treatment of
newborns with severe respiratory failure. Pediatr Pulmonol 1992;13:108.
173. Frank L, Sosenko IR. Undernutrition as a major contributing factor in the pathogenesis of
bronchopulmonary dysplasia. Am Rev Respir Dis 1988;138:725.
174. de Los Santos R, Seidenfeld JJ, Anzueto A, et al. One hundred percent oxygen lung injury in adult
baboons. Am Rev Respir Dis 1987;136:657.
175. Jacobson JM, Michael JR, Jafri MH Jr, et al. Antioxidants and antioxidant enzymes protect against
pulmonary oxygen toxicity in the rabbit. J Appl Physiol 1990;68:1252.
176. Tanswell AK, Freeman BA. Liposome-entrapped antioxidant enzymes prevent lethal O2 toxicity in the
newborn rat. J Appl Physiol 1987;63:347.
177. Walther FJ, Gidding CE, Kuipers IM, et al. Prevention of oxygen toxicity with superoxide dismutase and
catalase in premature lambs. J Free Radic Biol Med 1986;2:289.
178. Rosenfeld W, Evans H, Concepcion L, et al. Prevention of bronchopulmonary dysplasia by administration

of bovine superoxide dismutase in preterm infants with respiratory distress syndrome. J Pediatr 1984;105:781.
179. Rosenfeld WN, Davis JM, Parton L, et al. Safety and pharmacokinetics of recombinant human superoxide
dismutase administered intratracheally to premature neonates with respiratory distress syndrome. Pediatrics
1996;97:811.
180. Davis JM, Rosenfeld WN, Richter SE, et al. Safety and pharmacokinetics of multiple doses of recombinant
human CuZn superoxide dismutase administered intratracheally to premature neonates with respiratory
distress syndrome. Pediatrics 1997; 100:24.
P.599
181. Davis JM, Parad RB, Michele T, et al. Pulmonary outcome at 1 year corrected age in premature infants
treated at birth with recombinant human CuZn superoxide dismutase. Pediatrics 2003;111:469.
182. Sahgal N, Davis JM, Robbins C, et al. Localization and activity of recombinant human CuZn superoxide
dismutase after intratracheal administration. Am J Physiol 1996;271:L230
183. Kinsella JP, Walsh WF, Bose C, et al. Randomized controlled trial of inhaled nitric oxide in premature
neonates with severe hypoxemic respiratory failure. Lancet 1999;354:1061.
184. Banks BA, Seri I, Ischiropoulos H, et al. Changes in oxygenation with inhaled nitric oxide in severe
185. Robbins CG, Davis JM, Merritt TA, et al. Combined effects of nitric oxide and hyperoxia on surfactant
function and pulmonary inflammation. Am J Physiol 1995;269:L545.
185. Kinsella JP, Parker TA, Galan H, et al. Effects of inhaled NO on pulmonary edema and lung neutrophil
accumulation in severe experimental HMD. Pediatr Res 1997;41:457.
Chapter 31
Principles of Management of Respiratory Problems
William E. Truog
Sergio G. Golombek
Although the principles of management of respiratory disorders are constant, the techniques of treatment have
become more complex recently because of increasing and increasingly sophisticated technology. This complexity
results from the incorporation into routine care of highly technical equipment and the necessity for skilled
personneldedicated to the smooth functioning of life-sustaining devices. Knowledge of the pathophysiology of
lung disorders and the maturational status of the lung is essential to applying the imperfect evidence derived
from clinical trials and from experience to everyday use of assisted ventilation.
The goal of respiratory treatment is to provide tissue oxygenation and carbon dioxide (CO2) removal in a safe
and effective manner. Excessive oxygen delivered to the airways or to organs can be harmful. Excessive lung
distention can result in stretching and tearing of the lung. Excessive alveolar ventilation with resultant
respiratory alkalosis and hypocarbia can in a detrimental manner alter blood flow distribution and oxygen
unloading. In contrast, respiratory and/or metabolic acidosis may constrict pulmonary blood vessels, thus
affecting pulmonary blood flow and oxygen uptake. Attention to organ system dysfunction other than the lung
must occur in parallel with respiratory management or the efforts directed to optimizing delivery of assisted
ventilation will be useless. One of the most challenging aspects of ventilatory treatment is the dynamic way in
which the ventilatory needs can change because of either the treatment applied or the progression of the
underlying disease. This chapter reviews the indications, methods, and complications of respiratory management
and available methods of assessment.
OXYGEN THERAPY
Optimal Level Of Oxygenation
Considerable debate has occurred about what constitutes an acceptable level of oxygen tension or hemoglobin
saturation, especially in extremely preterm infants (1). The debate has focused on the contribution of excessive
or of insufficient supplemental oxygen levels and hence levels of arterial pO2 in the etiology of multiple disorders
of prematurity. These disorders include central nervous system injuries, especially leukomalacia or hemorrhagic
infarction; development or progression of retinopathy of prematurity (ROP); chronic lung disease of prematurity;
and perhaps other disorders associated with excess level of reactive oxygen species (ROS).
Because of the direct relationship between elevated fraction of inspired oxygen (FiO2) and arterial oxygen partial
pressure (PaO2) and the generation of ROS, it is reasonable that the arterial partial pressure of oxygen (PO2)
and/or oxyhemoglobin saturation be minimized to a level sufficient to allow adequate oxygen tissue delivery with
satisfactory reserves. There are several practical limitations to this seemingly innocent idea. Cardiopulmonary
status of extremely preterm babies is inherently unstable. There is intrinsic cyclicity to such physiologic events
as cardiac output and to spontaneous respiration rate and depth, and the variable contribution of spontaneously
generated respirations in addition to those associated with assisted respiration.
Hemoglobin concentration is subject to significant changes and the relative quantity of the various forms of
hemoglobin can change rapidly over the first days and weeks (e.g., decreasing fetal hemoglobin and increasing
adult hemoglobin). As this occurs, the relationship between partial pressure of oxygen tension and oxygen
saturation can change in somewhat unpredictable ways in short periods of time. Relying on only one
measurement exclusively (e.g., PaO2 or pulse oximeter saturation [SpO2]) either may mask or may exaggerate
changes in the tissue oxygen delivery. Hemoglobin concentration is measured per mL or 100 mL of blood, but
the actual blood volume, and hence total body hemoglobin concentration, can vary in sometimes unpredictable
ways also.
P.601
The overall limitation in assessing tissue gas exchange is that all the elements of tissue oxygen delivery are not
routinely assessed at bedside. For example, arterial blood pressure measurements, either by sphygmometry or
by direct arterial measurement, do not measure cardiac output or oxygen tissue delivery. Clinicians have
available only summary information, at best, of total body oxygen delivery and consumption. In fact the body
might be considered as interdependent organs with individualized oxygen consumption. Satisfactory levels of
oxygen delivery in one or most of these may not be satisfactory for a particularly high-consuming area, and
interorgan changes and distribution of blood flow can have a marked impact on local oxygen tissue delivery and
consumption and hence on the risk of organ-specific ischemic injury.
For babies treated with assisted ventilation, it is possible to manipulate arterial oxygen tension, perhaps at the
cost of other kinds of injury, especially to the lung. The tradeoff between mean airway pressure and FiO2 in the
treatment of conditions associated with low end expiratory volume (e.g., respiratory distress syndrome or RDS)
is one example of a difficult balancing act in clinical medicine. Central venous sampling or selective sampling of
organ specific venous drainage would be crucial to optimizing the pressure FiO2-dichotomy, but obtaining that
information is impractical.
Two randomized studies (2,3) and one case series (4) demonstrated that maintaining pulse Doppler oximetry
levels in a higher end of normal range in a group of babies with progressive ROP did not result in reduced
progression of ROP, but was associated with borderline statistically and clinically significant increases in
pulmonary problems. However, Schultz and associates (5) demonstrated that with oxygen saturation and FiO2
maintained in a lower range of normal, there appeared to be an elevated pulmonary vascular resistance (PVR)
and subsequent VA/Q mismatch (Fig. 31-1). Should episodes of lower SpO2 and higher PVR be exacerbated,
then altered gas exchange with acute elevations in CO2 and subsequent ill effects on cerebral circulation could
develop. The interdependency of these variables has meant that at present it is difficult to recommend, based on
clinically relevant outcomes, a proper oxygen level.
Figure 31-1 Relationship between venous admixture and PaO2. From Schulze A, Whyte RK, Way RC, Sinclair
JC. Effect of the arterial oxygenation level on cardiac output, oxygen extraction, and oxygen consumption in
low birth weight infants receiving mechanical ventilation. J Pediatr 1995;126 (5):777, with permission.
An arterial PaO2 of 45 mm Hg results in a saturation of fetal hemoglobin (HbF) of approximately 90%.

Maintaining the PaO2 above 50 mm Hg should be sufficient for tissue oxygen needs but excessive levels are
quickly achieved. Mitochondrial PO2 is about 2 mm Hg.
The gap between the alveolar partial pressure of oxygen (PaO2) and the PaO2 indicates the magnitude of the
arterial O2 gradient across the lungs and provides an indication of the magnitude of right-to-left shunting of
blood. A simplification of the alveolar air equation provides an estimate of PaO2 i.e., PaO2 = PiO2-PaCO2.
Because at sea level the barometric pressure, minus water vapor pressure, is approximately 700 mm Hg, the
percent of inspired oxygen multiplied by 7 equals PiO2 in mm Hg (e.g., 21% [approximate, equals] 147 mm Hg,
50% [approximate, equals] 350 mm Hg). Because PaCO2 approximates PaCO2 because of a usually small
arterial-alveolar CO2 gradient (aADCO2), PaCO2 can be substituted for PACO2, and PAO2 can be derived. For
example, if an infant is breathing gas with FIO2 of 0.6, with the measured PaO2 of 70 mm Hg and the PaCO2 of
40 mm Hg, the PAO2 = 420 - 40 = 380 mm Hg, and the alveolar-arterial gradient for O2 (AaDO2) approximates
310 mm Hg. In an infant without either lung disease or a significant right-to-left cardiac shunt, the AaDO2
should not exceed 25 mm Hg while breathing ambient air. Infants with severe RDS may have an AaDO2 in
excess of 500 mm Hg while breathing 100% oxygen.
Oxygen Delivery
Each gram of HbF binds 1.37 dL of oxygen. The full-term newborn with a hemoglobin (Hb) of 17 g/dL binds and
transports 23 dL of oxygen per 100 dL of blood. Less than 2% of transported O2 is carried as oxygen dissolved
in plasma. Normal tissue oxygen consumption extracts 4 mL O2/100 mL if oxygen consumption and cardiac
output are normal. HbF binds oxygen with a greater affinity than adult HbA. The oxyhemoglobin saturation curve
is nonlinear, and the P50, the PaO2 at which Hb is 50% saturated, increases with gestational age (Fig. 31-2). The
higher the P50, the greater the driving pressure for oxygen unloading. The curve gradually shifts to the right as
hemoglobin A increases after birth. Several factors can adversely affect tissue oxygen delivery, including
decreased cardiac output, maldistribution of cardiac output, arterial vasoconstriction, and shifts in the O2
dissociation curve. Oxygen unloading in the tissues is increased with a shift to the right of the O2 dissociation
curve (i.e., decreased O2 affinity of Hb) facilitated by a local decrease in pH, increase in PaCO2, and increase in
temperature. A shift to the right of the O2 dissociation curve can result from transfusion of adult red blood cells.
Oxygen uptake depends on adequate
P.602
alveolar ventilation (VA), an appropriate ventilation-perfusion match in the lungs, and absence of right-to-left
shunting. Oxygen uptake or increased O2 affinity of Hb (associated with a shift of the curve to the left) is
enhanced by alkalosis, decreased temperature, decreased 2,3-diphosphoglycerate, and increased HbF.
Figure 31-2 Oxygen equilibrium curves of hemoglobin at birth, 2 months of age, and adulthood. Note the
increase in P50 with age.
Oxygen Administration
There are two methods by which to deliver supplemental oxygen to neonates: an oxygen hood with sufficient
gas flow to prevent CO2 retention and a nasal cannula or prongs. The concentration and the rate of flow are
varied, and the precise amount of oxygen delivered to the lungs by nasal prongs is difficult to determine. This is
because there may be dilution of inspired air through ill-fitting prongs or through an open mouth. Estimates of
the effective FiO2 have been based on patient's weight, gas flow rate, and concentration of oxygen blended in
the blender (6). The O2-air mixture should be warmed to the same temperature as the incubator air, which
should be in the range of thermal neutrality (see Chapter 24).
ASSESSMENT OF GAS EXCHANGE

Clinical Assessment
The infant with respiratory problems may present with a wide spectrum of clinical findings. The infant's response
depends on the degree of prematurity, lung and chest wall development, and maturation of respiratory control.
The full-term infant may be able to increase the work of breathing to accomplish adequate gas exchange without
treatment, including oxygen administration. The extremely premature infant will have a much weaker
respiratory drive and inadequate muscular development and, thus, is less able to compensate for lung
abnormalities. The classic clinical signs of respiratory distress are helpful in the assessment of the mature
newborn infant. Nasal flaring, grunting respirations, and tachypnea are almost always present. With progression
of lung disease and decreased lung compliance, chest wall retractions become more marked. With increased
work of breathing, retractions progress from sternal to subcostal, to intercostal, and then to a seesaw pattern of
chest and abdominal wall movement. The full-term infant may increase respiratory rate above 100 per minute,
with shallow respirations. This pattern is the most efficient way to increase gas exchange, with the least costly
work of breathing. Expiratory grunting represents an effort to retard expiratory flow to increase end-expiratory
pressure and maintain alveolar patency. It is unsafe to rely on color changes as an indication of oxygenation, as
abnormalities in peripheral perfusion as a result of poor cardiac output, hypotension, or hypovolemia may be
misleading. Similarly, infants with recurrent apnea will have intermittent deficiency in gas exchange.
Auscultation assists in determining the quality of air entry in various parts of the lung and the presence of
airway secretions or obstructions.
Measurement and Techniques

The measurement of blood gases and pH helps to confirm clinical impressions, and verifies the values obtained
by pulse oximetry, transcutaneous electrodes, and end-tidal CO2 monitoring. Most importantly, it helps to
minimize the risks of hypoxia, hyperoxia, hypocapnia, hypercapnia, and metabolic acidosis.
Blood Gas Tension Measurements

Blood may be obtained by puncture of a peripheral artery, cannulation of a peripheral or umbilical artery, venous
blood sampling, or capillary puncture. Venous and/or capillary blood values provide an approximation of arterial
partial pressure of CO2 (PCO2) and pH but limited information about arterial oxygenation. The PCO2 in venous
blood is approximately 6 mm Hg higher and the pH is 0.03 lower than in arterial blood.
Peripheral arterial blood most often is sampled from the radial or posterior tibial arteries using a 23- or 25gauge needle. Peripheral puncture is indicated when the anticipated number of samples will be small, or if
arterial cannulation is unsuccessful. The painful nature of the procedure may result in altered breathing and a
spurious result.
Cannulation
The choice of peripheral artery or umbilical artery catheterization usually is determined by the size of the infant,
and the anticipated duration of the cannulation. The major advantage of a peripheral arterial catheter is the
avoidance of umbilical cannulation and risk of thrombosis. A right radial arterial catheter has the unique
advantage
P.603
of sampling preductal PaO2, which more accurately reflects retinal artery PO2.
The need to insert a catheter into the umbilical artery should be based on the infant's maturation, postnatal age,
and the type, severity, and expected duration of the illness. The high incidence of thrombus formation on the
catheter tip, with its attendant risk of mural thrombus or emboli, must be weighed against the potential benefits
to the infant. The tip of the umbilical catheter should rest in a low or high position to avoid the risk of occlusion,
thrombosis, or direct infusion into a major branching artery. The distance to various levels within the aorta is
estimated from the infant's shoulder to umbilicus distance and the chart of Dunn (Fig. 31-3) (7). The high
location range is below the ductus arteriosus and above the takeoff of the celiac artery, with the tip resting
between T5 and T10. At the lower site, the tip should be above the bifurcation of the aorta (L3 to L5) and below
the takeoff of the renal and inferior mesenteric arteries. There appears to be a higher complication rate with
catheters in the low position. The higher position allows greater leeway for catheter migration, although there is
a greater risk of downstream embolization. A thoracoabdominal radiograph should be obtained immediately after
catheter placement and before medications are infused. Once the catheter is in place, there is a risk of
intraluminal clotting. Therefore, a continuous infusion is required, as is flushing of the line following blood
sampling. Heparinization of fluid infusions is performed in many special care nurseries routinely. The umbilical
arterial catheter fluids serve as a site of low grade hemolysis. Amino acid containing fluids can reduce this risk
and supply some necessary protein constituents (8). A pressure strain gauge should be connected to the
catheter, with alarm limits to indicate changes in blood pressure. Dampening of the pressure waveform indicates
intraluminal or catheter tip narrowing or obstruction. The catheter should be removed once the infant has
improved, the frequency of blood gas determinations has decreased to once or twice per day, and systemic
arterial pressure monitoring no longer is essential.
The complications of umbilical artery catheterization are listed in (Table 31-1). The most frequent problem is
peripheral vasospasm associated with blanching or patchy cyanosis (See Color Plate) of the distal leg, foot, or
toes. If this does not improve with reflex vasodilation by warming of the contralateral leg, or worsens over the
next 15 to 20 minutes, the catheter should be removed. Some thrombosis at the catheter tip is inevitable.
Fortunately, the rate of complications from thrombus formation is relatively low.
Transcutaneous PO2 and PCO2

Transcutaneous PO2 monitoring can be used selectively in conjunction with pulse oximetry. Most devices contain
oxygen (O2) and CO2 electrodes. Noninvasive PO2 measurements continue to be helpful in certain situations,
particularly if transcutaneous PCO2 also is obtained. Calibration before use and correlation with an arterial
sample are necessary, but the need for subsequent blood samples should be reduced.
Figure 31-3 Relation between the shoulder-to-umbilicus measurement and the length of umbilical artery
catheter needed to reach the aortic bifurcation, diaphragm, and aortic valve. From Dunn PM. Localization of the
umbilical catheter by post-mortem measurement. Arch Dis Child 1966;41:69, with permission.
Arterial PO2 and transcutaneous PO2 are not identical. Differences can arise from local O2 consumption by the
skin or by the electrode itself, heating of the skin, O2 diffusion time, and response time of the electrode (9).
Skin blood flow may be affected by vasopressor medications, hypotension, and shock (10). Use of
transcutaneous PO2 can reduce the number of blood sampling procedures, particularly during a period when
rapid changes in O2 administration or mechanical ventilatory settings are taking place. Continuous monitoring
for several hours also allows assessment of changes as a result of position, handling, suctioning, and feeding
and for comparison with arterial oxygen saturation (SaO2) monitoring. The risk of partial thickness burns
precludes its use for longer than 5 hours at a single site on the body.
TABLE 31-1 COMPLICATIONS OF UMBILICAL CATHETERIZATION
Complication
Incidence
Comment
Limb ischemia
<20%
Thrombosis
<90%
Assess Doppler arterial and venous flow from femoral region

to dorsalis pedis; if reflex and unilateral, apply warmth to
contralateral extremity; if persistent (>15 min), remove
catheter
Assess by aortic ultrasonography
Infection
Colonization
57%
Remove in presence of gram-positive blood culture if catheter

was inserted more than 24 h previously; no relation to
Sepsis
Blood loss
5%
Rare
Vascular perforation
Extremely rare
duration of catheter
Connection to blood pressure transducer with alarm should
prevent significant loss
On removal, clamp for 5-10 min
P.604
Pulse Oximetry
Pulse oximetry provides a safe, accurate, and noninvasive adjunct to the assessment of tissue oxygenation
11,12). Oxygen saturation is determined by infrared spectrometry, utilizing two electrodes and a small cuff that
can be placed around a hand, foot, or toe without requiring heating or calibration. One electrode contains two
diodes that emit light at two wavelengths: red at 660 nm and infrared at 940 nm. The other electrode senses
the light from both of these diodes that has not been absorbed by blood or tissue. The relative concentration of
hemoglobin-oxygen (HbO2) and deoxyhemoglobin determines the amount of transmitted light, because different
forms of Hb have markedly different absorption characteristics. The ratio of the amount of light absorbed at each
wavelength is used to calculate a SaO2 value. The pulsed element of the apparatus allows the instrument to
differentiate added arterial blood oxygenation and absorption from tissue, and it subtracts the amount
contributed by nonpulsatile venous blood flow. With PO2 values greater than 40 mm Hg, the saturation
accurately reflects measurements of PO2 obtained by catheter sample or by transcutaneous PO2 (12).
A PaO2 of 60 to 90 mm Hg results in a saturation value of 94% to 98% (see Fig. 31-2), and changes of 1% to
2% usually reflect a PaO2 change of 6 to 12 mm Hg. The point of inflection at which the HbO2 dissociation curve
grows steep has considerable variability and depends on proportions of hemoglobin A, HbF, PCO2, pH, and
temperature. Generally, these variables are not so critical to the interpretation of the percent SaO2 in arterial
blood as they are to PaO2. Below 40 mm Hg, the SaO2 falls below 90%. Poor correlation with PaO2 exists when
the SaO2 is above 96%, in which case the PaO2 may be well above 100 mm Hg. In very low birth weight (VLBW)
infants who require chronic oxygen administration and who are at risk for developing ROP, the upper limit of
saturation should be reduced to 95% or less depending on local nursery policies. Inaccuracies may reflect
improper placement, movement, or peripheral ischemia. Motion artifacts also produce invalid readings. However,
these problems are less significant with newer pulse-Doppler devices.
Near-Infrared Spectroscopy
Utilization of the unique light-absorbing properties of Hb and HbO2, as used in pulsed oximetry, has led to a
sophisticated method of appraising tissue oxygenation by means of near-infrared spectroscopy. Near-infrared
light penetrates the skin, bone, and various tissues and can be detected by electrodes placed on opposite sides
of an infant's skull. This permits assessment of cerebral tissue O2 and alterations in cerebral blood volume. Hb
and cyto-chrome a and a3 (cyt a, a3) change their absorption characteristics according to the degree of
oxygenation. The wavelength at which maximal absorption occurs is different for HbO2, deoxygenated Hb, total
Hb, and reduced and oxygenated cyt a, a3. The small cranial size of the infant weighing less than 1500 g makes
cross-temple spectroscopy feasible. However, the technique has not gained widespread use in clinical
neonatology.
End-Tidal CO2 Monitoring

The concentration of CO2 at the mouth or nose rises to reach a plateau at the end of each breath. This plateau
reflects the alveolar CO2 concentration under normal conditions, but may be inaccurate if there is marked
ventilation-perfusion inequality or inhomogeneity of lung disease. Recent refinements to end-tidal CO2 detection
equipment permit in-line or mainstream infrared monitoring just proximal to the endotracheal tube
(capnography), with a continuous display of the PCO2 waveform. There is minimal dead space of the apparatus,
and sampling accuracy has improved to compensate for the low expiratory flow rates characteristic of small
premature infants. End-tidal CO2 is less reliable in extremely small premature infants. Capnography is as
accurate as capillary PCO2 but is less precise than transcutaneous monitoring (1314,15).
Hazards of High or Low Arterial CO2

Identifying a safe range of PCO2 has proven as difficult as for PaO2. There are substantial risks to hypocarbia
(Fig. 31-4)
P.605
(16) (Table 31-2). Limiting the exposure time to hypocarbia may be important in presenting the development of
cystic periventricular leukomalacia or decreasing the risk of CLD (17,18 and 19). There has been an increase in
use of permissive hypercarbia, by which PaCO2 levels of 50 to 60 mm Hg are sought and maintained. The
rationale is to avoid high peak inspiratory pressure or to delay the onset of or to avoid assisted ventilation
(20,21).
Figure 31-4 Effects of Hypocapnia on the Brain in Premature Infants. Hypocapnia has been implicated in the
pathogenesis of neonatal white-matter injuries, including periventricular leukomalacia, resulting in
intraventricular hemorrhage. At normal carbon dioxide levels (left-hand side of figure), cerebral blood flow is
determined by local metabolic demand. Prolonged or severe hypocapnia includes severe cerebral
vasoconstriction, resulting in brain ischemia, particularly in poorly perfused areas of the brain such as
watershed areas (right-hand side of figure). This ischemia may initiate white-matter destruction in the brain of
premature infants. Additionally, antioxidant depletion (caused by excitatory amino acids), lipopolysaccharide
(LPS), and cytokines produced in response to sepsis, such as interleukin-1 and tumor necrosis factor (TNF), potentiate the process. Finally, restoration of the normal partial pressure of arterial carbon dioxide can
result in cerebral vasodilation, which may precipitate or contribute to intraventricular hemorrhage. From Laffey
JG, Kavanagh BP. Hypocapnia. N Engl J Med 2002;347 (1):43, with permission.
STRATEGIES FOR RESPIRATORY SUPPORT

Continuous Positive Airway Pressure
The application of end-expiratory pressure is intended to prevent alveoli and/or terminal airways from collapsing
to airlessness. Continuous positive airway pressure (CPAP) may be applied during spontaneous breathing or as
positive end-expiratory pressure (PEEP) during mechanical ventilation. This usually requires pressures between
4 to 6 cm H2O for CPAP and 3 to 8 cm H2O for PEEP.
TABLE 31-2 RISKS OF HYPOCARBIA AND HYPERCARBIA
Hypocarbia
Overventilation increases risk for pulmonary injury

Decreases cerebral perfusion, associated with ischemic whitematter injury
Increase in pH, interfering with tissue O2 unloading
Hypercarbia
Underventilation may increase areas of lung collapse and increase VA/Q mismatching necessitating higher
FIO2
May decrease pH elevating pulmonary vascular resistance andincrease VA/Q mismatching
Increases cerebral perfusion with associated risk of hemorrhagic infarctions
The physiologic effects of CPAP/PEEP may vary depending on the underlying pulmonary pathology, although the
primary goal is to prevent alveolar collapse. Grunting respirations in infants with respiratory distress suggest
laryngeal narrowing and increased resistance to expiratory flow to increase end-expiratory alveolar pressure. In
the surfactant-deficient state, alveoli will collapse at end-expiration unless a minimum distending pressure is
maintained. CPAP of 3 to 4 cm H2O will prevent alveolar collapse but will not recruit atelectatic alveoli. Opening
pressures of 12 to 15 cm H2O are required to inflate collapsed alveoli. The infant will need to create a large
distending airway pressure in the absence of CPAP. The shear forces from opening and closing of small airways
may contribute to alveolar epithelial damage. Additionally, resultant abnormal distending forces on terminal or
respiratory bronchioles will contribute to small airway injury. CPAP above physiologic levels of 3 to 4 cm H2O
may cause overinflation of some alveoli. Therefore, inflation and deflation may occur on the flatter portion of the
pressure-volume curve and increase the work of breathing. CPAP theoretically could stimulate surfactant
secretion. Maintenance of alveolar volume will reduce right-to-left shunting of blood through atelectatic alveoli,
hence reducing oxygen needs.
Indications
The clinical indications for CPAP are varied. Initial use was directed at infants with RDS with the goal of avoiding
or at least delaying intubation and mechanical ventilation. The gestational age, birth weight, and stage and
severity of respiratory disease should be factored into the decision to initiate CPAP. If infants have severe lung
disease, e.g., meconium aspiration, severe respiratory distress syndrome with FiO2 needs <6, or idiopathic
pulmonary hypertension, more appropriate therapy would include assisted ventilation.
Infants <1,000 g birth weight are at considerable risk for developing chronic lung disease (CLD) and recurrent
apneic episodes; therefore, they commonly are intubated and supported with mechanical ventilation as a result
of
P.606
their inability to sustain an adequate respiratory effort. They often develop an increasing oxygen need during
the second week of life, associated with radiographic signs of CLD and attributed, in part, to assisted ventilation.
CPAP may be beneficial in maintaining patency of extremely small terminal airways and prealveolar gas
exchange units in these very immature infants. CPAP appears to be well tolerated by such infants over a period
of many days. The increased enthusiasm for the use of CPAP is stimulated by the desire to minimize or prevent
CLD. Because assisted ventilation is one putative factor in the etiology of CLD, the use of CPAP shortly after birth
to avoid or minimize barotrauma has much appeal. Avoidance of endotracheal intubation should decrease the
chances for tracheal injury, airway infection, abnormal mucociliary function, and overinflation from excess
ventilator pressure or volume.
Brief intubation and the administration of a single dose of surfactant followed by nasal CPAP has been advocated
as another method of reducing the need for mechanical ventilation in infants with moderate RDS (22). Improvement in gas exchange has been demonstrated; however, more evidence of its efficacy in reducing the incidence
or severity of CLD is required.
Recurrent Apnea
CPAP helps some infants with recurrent apnea of prematurity to sustain a more regular respiratory rate. The
mechanism of its action is not well understood, although an increase in functional residual capacity (FRC) may
alter the Hering-Breuer reflex or stabilize the thoracic cage, minimizing chest wall distortion and possibly altering
inhibitory spinal cord reflexes (23). CPAP also helps to overcome obstructive apnea and decreases total
respiratory system resistance (24,25).
Earlier methods of applying CPAP utilized an enclosed head box, face masks, and nasopharyngeal tubes. More
recently, nasal prongs have been adapted to fit most infants. One device (the ALADDIN Infant Flow System,
Hamilton Medical Inc., Reno, NV) appears to be well tolerated by both large and small infants. This apparatus
maintains a constant flow of air by incorporating a double fluidic jet system within the apparatus. During
inspiration, one jet maintains the flow to match the infant's inspiratory effort; during expiration, gas flow is
reversed by a second jet to assist outflow while maintaining a constant minimum pressure. This system
presumably does not add to the work of breathing and reduces the need to use high flow rates to compensate
for air leak around the nasal prongs. A second system utilizes a variable depth water seal for the expiratory
circuit to sustain continuous airway pressure (26). CPAP can also be applied utilizing constant flow ventilators.
Complications
CPAP may have adverse effects. Overinflation can result in increased work of breathing and decreased efficiency
of gas exchange. Pneumothorax and pneumomediastinum may result from lung overdistension. Carbon dioxide
retention may occur as a result of either increased dead space or ineffective ventilation of some alveoli. If the
mean thoracic pressure is elevated with high levels of PEEP or CPAP, e.g., above 7 to 8 cm H20 in the absence
of lung disease, cardiac output may be decreased because of impaired systemic and pulmonary venous return.
The nasal prongs may cause irritation if the fit is not appropriate or the infant is active. Gastric distention may
occur, making gastric feedings difficult, and often an indwelling or gastric tube is required for decompression.
Effectiveness
Does the use of CPAP decrease the need for assisted ventilation and does it prevent or ameliorate CLD in VLBW
infants? Clinical trials conducted before the era of routine surfactant use and contemporary neonatal ventilation
devices are now of limited relevance. Recent studies comparing the efficacy of conventional mechanical
ventilation (CMV) to CPAP following the administration of surfactant found some differences in need for
ventilation as an outcome although the number of infants studied was small (22,28).
CPAP is used to facilitate weaning from mechanical ventilation. Some infants, experiencing recurrent apneic
episodes, appear to benefit (29), whereas infants evaluated in other studies have shown no benefit (30).
Additional information is needed to confirm whether CPAP is an effective adjunct to successful extubation. The
technique of nasal intermittent positive pressure ventilation (NIPPV) may gain widespread use if it can be shown
to improve long-term outcome (31).
Assisted Mechanical Ventilation

Treatment with assisted ventilation is applied commonly to newborns across the neonatal birth weight spectrum
(Fig. 31-5) (32). Illustrated is the rate of use of assisted ventilation, adjusted by birth weight, and the percent of
all infants treated for more than or equal to 24 hours. Thirty-eight percent of treated infants were over 2,500 g
birth weight in this survey, covering all California and New York State births for 1994-1995. Neonatologists must
be expert in providing assisted ventilation across a 10-fold range of patient weight (0.5 kg-5.0 kg) and providing
safe and effective assisted ventilation across a range of lung development from premature airways with
preacinar gas exchange spaces to a virtually completely alveolarized organ. Thus, both size (patient and lung
volume) and range of development are immense.
The Immature Lung

The immature lung presents a special hazard for the application of assisted ventilation. The application of
positive pressure for the purpose of increasing ventilation and optimizing ventilation-perfusion (V/Q) matching
may injure
P.607
epithelial and endothelial tissues. In the incompletely developed lungs, structures are less elastic and more
vulnerable to barotrauma. Injury to the mesenchymal and epithelial tissues that later give rise to alveolar
septation and vascular formation may be irreversible. Studies in adult animals have demonstrated that
otherwise healthy lungs can suffer injury, which is reflected by increased airway fluid and deterioration of gas
exchange, if inappropriate distending lung pressures are applied (33). The particular problems of providing
assisted ventilation are illustrated in Fig. 31-6 (34). Relative immaturity of distal bronchioles and respiratory
ducts, coupled with fluid-filled and collapsed alveoli, create a set of conditions leading to overdistention of some
areas and underventilation of other areas, with resultant ineffective gas exchange. This uneven ventilation,
coupled with injury produced by reactive oxygen species, contributes to the common problem of chronic lung
disease of prematurity. The risk of its development is inversely correlated with birth weight.
Figure 31-5 Distribution of total births and neonatal mechanical ventilation by birth weight. The primary y axis
is the mechanical ventilation rate per 1000 live births within a given weight category. The secondary y axis is
the percentage of all births and of all mechanical ventilation hospitalizations. Data are from New York and
California for 1994. Higher mechanical ventilation rates are seen in smaller babies. However, because most
babies are >2500 g, a large proportion of ventilated babies are of normal weight. From Angus DC, LindeZwirble WT, Clermont G, Griffin MF, Clark RH. Epidemiology of neonatal respiratory failure in the United States.
Projections from California and New York. Am J Respir Crit Care Med 2001;164:1154, with permission.
Great strides have been made in understanding how the immature lung differs from a mature lung in
phospholipid and surfactant-associated protein biosynthesis. However, there are factors other than surfactant
biosynthesis that are unique to the immature lung and that increase the susceptibility to injury. These factors
include, but are not limited to, incomplete development of the supportive net of collagen and elastin (35,36),
incomplete development of the capillary bed in the gas exchange areas (37), relative instability of the chest wall
with reduced capacity to maintain expiratory lung volume at FRC, immaturity of the neural control producing
sustained spontaneous respiratory effort, and probable immaturity of the metabolic functions of the pulmonary
endothelium.
Respiratory failure ensues when spontaneous breathing efforts fail to produce adequate alveolar ventilation. In
newborn infants, this may occur because of failure of adequate output from central nervous system respiratory
centers, an overly compliant chest wall that increases the work of breathing, metabolic problems as a result of
limited energy stores, or profoundly noncompliant lungs requiring more work and depleting available energy
stores. Each of these may be an indication for assisted ventilation. In most neonatal respiratory disorders, these
problems occur in combination, and the diagnosis of respiratory failure cannot be ascribed to any single cause.
Establishment of an Artificial Airway

PHYSIOLOGIC AND ANATOMIC AIRWAY PECULIARITIES
The newborn infant has distinct anatomic and physiologic characteristics of the airways and a strong preference
for nasal breathing for the first few months of life (38). Nasal or nasopharyngeal obstruction because of
secretions, mucosal
P.608
injury, or congenital abnormalities may produce respiratory distress. Approximately one-half of the infant's
airway resistance occurs in the nose, although the narrowness of the lower respiratory tract results in a total
airway resistance approximately 15 times greater than that of an adult (39). Edema and inflammation can
produce extremely high resistance to air flow in these narrow airways. During expiration, the airways become
narrower, and resistance increases.
Figure 31-6 A: A mature alveolar duct and alveoli. Dotted line, surfactant; PALV, alveolar pressure; PAW,
airway pressure; PPL, pleural pressure; Fi, tissue force (stretched springs) acting inward; Fo, tissue force
directed outward; Tw, wall tension or recoil pressure. B: The end-expiratory airway pressure (PAW) equals zero
in an immature distal airway (left). The saccules (SAC) and airways contain fluid (shaded area). The axial
airway is concave at the air-liquid interface as a result of the surface tension forces. The peripheral SACs are
collapsed or fluid filled. The lax tissues are represented by relaxed springs. The inspiratory airway pressure
(PAW) is equal to 26 cm H2O (right). The distended distal airway has a high wall tension (TW). The liquid front
has been pushed peripherally, but the SACs are still not inflated. From Thibeault DW, Lang MJ. Mechanisms and
pathobiologic effects of barotrauma. In: Merritt TA, Northway WH Jr, Boynton BR, eds. Bronchopulmonary
dysplasia. Contemporary issues in fetal neonatal medicine. Boston: Blackwell Scientific Publishers, 1988:82,
with permission.
Endotracheal Intubation
Route
Orotracheal and nasotracheal intubation may be used for prolonged mechanical ventilation of term and
premature infants. The principal advantage of the nasal route is the stabilization of the tube afforded by the
close fit within the naris, but the nasal passages may limit the size of tube that can be used. Necrosis of the
nasal septum or the alae nasi can occur if circulation is impaired because the tube is too large. Orotracheal
intubation is more easily and quickly accomplished and is indicated for delivery room and emergency situations.
It is the preferred route for prolonged mechanical ventilation.
The endotracheal tube should allow a small air leak between the tube and the glottis. A tube that fits too snugly
within the trachea is likely to cause pressure necrosis of the mucosa. If too large a leak is allowed, it may be
difficult to achieve sufficient pressure for ventilation of noncompliant lungs. A tube with a 2.5-mm inner
diameter usually fits infants weighing less than 1,000 g; a 3-mm tube fits those from 1,000 to 2,000 g; a 3.5mm tube fits those from 2,000 to 3,000 g; and 3.5- to 4.0-mm tube fits larger infants.
Technique
Orotracheal intubation is a simple procedure that can be accomplished atraumatically within a few seconds. The
necessary equipment consists of a straight-bladed laryngoscope, a suction catheter connected to a suction
apparatus, an endotracheal tube of the appropriate size with an adapter for the bag or respirator, and an optical
flexible Teflon introducer, bent to prevent its tip from protruding beyond the end of the endotracheal tube. The
infant is ventilated with 100% oxygen by mask for a few breaths. A catheter to deliver oxygen can be taped to
the laryngoscope blade to enhance oxygen delivery during intubation (40). The infant's neck is straightened
without hyperextension by placing a small towel under the shoulders, and the head is steadied by an assistant.
The laryngoscope is
P.609
held in the left hand between the thumb and first two fingers. The heel of the hand is placed against the infant's
left cheek to provide stability. The blade is introduced into the right side of the mouth, and the tongue is
deflected to the left as the blade is advanced into the vallecula, anterior to the epiglottis. The laryngoscope is
lifted rather than rotated so that the larynx is elevated and the glottis is brought into view (Fig. 31-7). The
pharynx is suctioned if necessary. The endotracheal tube is introduced into the mouth to the right of the
laryngoscope and gently guided into the glottis under direct vision.
Figure 31-7 Laryngoscopy for endotracheal intubation.
The nasotracheal tube is inserted without an introducer through the naris and gently guided along the floor of
the nose. The laryngoscope is placed in the mouth to the right of the orotracheal tube, and the tip of the
nasotracheal tube is seen in the posterior pharynx. A Magill forceps is held in the right hand and introduced to
the right of the laryngoscope. The nasotracheal tube is grasped a few millimeters back from its tip with the
forceps, and the tip of the tube is elevated until it is almost at the glottis. It is helpful to have an assistant grasp
the exterior end of the nasotracheal tube to assist in advancing it. The orotracheal tube is left in place until just
before insertion of the nasotracheal tube in the glottis.
Pretreatment with pancuronium bromide and atropine may minimize heart rate and intracranial pressure
changes associated with endotracheal intubation (41,42). Morphine and/or midazolam are commonly provided
prior to intubation attempts to reduce agitation. Size and stability of the patient dictate the exact choice of
adjunctive mediation use.
Positioning
The length of the trachea from the vocal cords to the carina varies from about 3.6 cm in the smallest premature
infants to 6 cm in large, term infants. Optimal positioning for the tip of an endotracheal tube is in the middle of
the trachea, in which it is least subject to dislodgment into the pharynx or displacement into a bronchus. The
proper depth of insertion of an endotracheal tube, as determined by postmortem and radiographic
measurements, is related to body weight (43). Suggested depths of insertion for orotracheal intubation are
given in (Table 31-3) (43,44).
TABLE 31-3 DEPTH OF INSERTION OF AN OROTRACHEAL TUBE FROM THE LIPS OF A PREMATURE
INFANT
Infant Weight (kg)
Depth of Insertion (cm)
1.0
2.0
3.0
4.0
7
8
9
10
Immediately after intubation, the position of the tube should be confirmed by inspection and auscultation.
Identification of humidity in the tube or of CO2by a sensitive color detectorhelps confirm endotracheal tube
placement. Two common errors of tube placement are intubation of the esophagus and intubation of the right
main-stem bronchus. The former should be suspected if insufflation through the tube produces abdominal
distention with little chest expansion and if air movement is heard better over the stomach than over the chest.
Breath sounds that are louder over the right chest than the left suggest that the tube is in the right main-stem
bronchus. Auscultation, although helpful, is not reliable because breath sounds are well transmitted in a small
chest. A chest radiograph should be obtained to confirm tube placement.
Care of Endotracheal Tubes

A tube in the trachea interferes with the physiologic mechanisms for clearance of respiratory secretions and may
itself stimulate secretions. Meticulous care is needed to prevent accumulation and inspissation of secretions
which can obstruct the tube. Routine changing of the tube is unnecessary and subjects the infant to the
repeated risk of trauma to the larynx and interruption of ventilation. However, the most common reason for
sudden deterioration of an otherwise stable ventilated infant is dislodgment or blockage of the tracheal tube. The
volume and quality of secretions vary with the type of pulmonary disease and with the individual patient.
Suction frequency may vary from once each hour to fewer than two or three times per day. Both in-line
techniques and discontinuation of positive-pressure ventilation are used for suctioning of airway secretions. Inline techniques offer the advantage of no loss of distending pressure during the suctioning process. However,
no formal comparisons of the safety of the two techniques have been conducted.
Another important aspect of the care of artificial airways is the provision of adequate humidification. The
endotracheal tube bypasses the nasal and pharyngeal mucosa, which normally warms and humidifies inspired
gases. If heat and humidity are not provided from an external
P.610
source, drying of the lower airway mucosa, thickened secretions, and hypothermia may result. Inspired gases
should be passed through a heated nebulizer with a constant readout temperature probe so that they are
delivered to the airway already warmed to a range of 34C to 36C, and saturated. Inadequate humidification
can contribute to airway injury (45).
Indications for Assisted Ventilation

The decision for initiation of assisted ventilation should be individualized for each baby. Factors to consider
include underlying disease and its expected natural history, birth weight, gestational age, postnatal age, chest
radiographic appearance, progression of clinical signs, serial arterial blood gas tension measurements, and pH
measurements. The criteria indicating a need for mechanical ventilation are difficult to define, and there is lack
of unanimity about a particular threshold for PaO2, partial pressure of CO2 (PaCO2), or FiO2. In general, the
PaO2 should be maintained at or above 50 mm Hg because of reasonable oxyhemoglobin saturation at this level,
but the maximal level of inspired O2 dictating intubation and application of assisted ventilation remains
controversial. A trend of rising PaCO2 with concomitant decrease in pH or onset of apnea indicates a need for
mechanical assistance. After assisted ventilation is initiated, the generally accepted goals are to maintain the
PaO2 between 45 and 70 mm Hg, PaCO2 at 45 and 60 mm Hg or less, and pH at 7.25 or more although
minimizing PIP and FiO2, and optimizing mean airway pressure (PAW[gas]) and PEEP. These general guidelines
must be interpreted and often must be modified to provide optimal support for the individual patient.
Because acute lung disease is usually more severe and protracted in more immature infants, criteria for
intervention for infants weighing less than 1,000 g differ from those for larger or older infants. For example, a
750 g infant with RDS has a high probability of developing apnea, fatigue, or both, and most of these infants
require assisted ventilation even if the FiO2 need is less than 40%. A 2,500 g, 36-week-old infant with RDS has
greater muscular and caloric reserve and is able to sustain rapid ventilatory rates and higher respiratory work
for several days without assistance. With a normal PaCO2, inspired O2 may be increased to between 80% and
90% before intubation in some infants. Infants of gestational age 35 to 39 weeks and older than 24 hours who
develop respiratory failure with RDS may benefit from surfactant treatment (46).
Physiologic Considerations
An understanding of the effects of mechanical ventilation on the lungs requires knowledge of the interplay
among thoracic mechanics, including pulmonary compliance and airway resistance, lung volumes, respiratory
control mechanisms, and alveolar gas exchange.
Lung compliance change in lung volume per unit pressure change, in units of mL/cm H2O depends on the elastic
properties of the tissue, which are influenced by the lung volume and abnormalities such as tissue inflammation
and edema. Compliance is low if there is alveolar collapse or overdistention. Expansion from alveolar collapse
requires inflation pressures of 12 to 20 cm H2O in preterm infants with RDS to achieve tidal volumes of 3 to 5
ml/kg. The lungs of infants with RDS have areas of collapse and overexpansion, and there is nonuniformity of
compliance. Other conditions, such as pneumothorax, lobar atelectasis or consolidation, and pulmonary edema,
decrease compliance. The most relevant measure of compliance, specific compliance, is calculated by
normalizing compliance by end-expiratory volume (EEV), the functional equivalent of FRC measured during
positive pressure ventilation. Very low or high values for EEV will reduce compliance. Changes in compliance,
EEV, and gas exchange are not concordant, at least not during RDS. This has limited the value of bedside
measurements of compliance, particularly without concomitant measurements in EEV. Chest wall compliance
usually is high and does not present a problem to mechanical ventilation.
Airway resistance (cm H2O/L/s) is inversely related to the fourth power of the radius during laminar air flow.
Airway resistance is high in infants, increasing with low lung volumes and with obstruction of the airway. High
rates of air flow increase resistance by producing turbulence in the airways.
The rate at which lung areas inflate and deflate is determined by resistance and compliance. An increase in
airway resistance increases the time required for air to reach the alveoli; a decrease in compliance results in less
time required to reach equilibrium. The product of resistance and compliance is the pulmonary time constant.
Changes in resistance or compliance can alter the pattern or distribution of ventilation, and recognition of the
variations in the time constant (e.g., short with poor compliance, prolonged with increased airway resistance)
helps determine respirator settings. Unfortunately, a single time constant does not exist for all lung areas during
complex pulmonary disorders. Thus, all conventional positive-pressure ventilators produce areas of overinflation
and underinflation of gas exchanging areas, each contributing to suboptimal gas exchange.
Because RDS should result in a short time constant, rapid inspiratory and expiratory respirator times are
permissible, and mean airway pressure should be increased to improve oxygenation. With meconium aspiration
or airway edema, the time constant is slower, and sufficient time for expiration is important to avoid gas
trapping, overdistention of the lungs, and possible air leak. If the expiratory time (TE) is shorter than the time
constant of the lung for expiration, overdistention results. If the overall time constant for the lung is longer than
the imposed ventilator inspiratory time (TI), inadequate ventilation could result. Unequal time constants
coexisting in different parts of the lung are most likely to occur if pulmonary abnormalities are unevenly
distributed, as in pneumonia, meconium aspiration, pulmonary interstitial emphysema, pneumothorax, or CLD,
in which case the optimal TI or TE becomes difficult to determine.
The circulatory effects of mechanically applied pressure to the alveoli are important. Normal breathing results in
P.611
negative intrapleural pressure that enhances venous return and cardiac output. Positive-pressure breathing can
impede venous return and may diminish cardiac output. Pressure during inspiration decreases the pulmonary
capillary circulation as long as alveolar pressure exceeds capillary pressure and can affect total pulmonary blood
flow and hence gas exchange.
Lung Volume Measurements During Mechanical Ventilation

Application of hot wire anemometry or pneumotachography to time-cycled pressure- limited ventilation systems
allows measurement of inspiratory and expiratory tidal volumes, minute ventilation (VE), and air leak, the
difference between tidal volumes measured during inspiration and expiration at any combination of ventilator
settings. Tidal volume can be correlated with peak inspiratory pressure (PIP), inspiratory gas flow rate, TI and
PEEP. Knowledge of tidal volume from bedside measurements helps allow determination of the optimal PIP to
achieve optimal tidal volume. For many clinical situations, this value is between 3 and 5 ml/kg. This knowledge
allows minimizing PIP, which, if excessive, otherwise may induce or exacerbate the small airway injury of CLD.
However, knowledge of tidal volume and VE does not provide knowledge of distribution of inspired ventilation
and of [V with dot above]A/[Q with dot above] matching. Distribution of tidal volume may vary with the
associated PIP; low PIP can result in tidal volume distribution only to already overinflated lung regions, resulting
in worsened [V with dot above]A/[Q with dot above] matching, development or exacerbation of high [V with dot
above]A/[Q with dot above] areas, and worsening of CO2 retention, despite normal or elevated [V with dot
above]E.
Many ventilators now have pulmonary graphic systems to monitor breath-by-breath pressure-volume and flowvolume loops. These graphics can be helpful in evaluating the extent of airway instability during inspiration and
especially expiration. Pressure-volume curves may be helpful in detecting overdistention and/or air trapping or
small airway obstruction. One limitation in the use of pulmonary graphics is the considerable breath-to-breath
variability in the pattern exhibited, probably more than exists with assisted ventilation of adults. There is as yet
no strong evidence suggesting that a clinically important outcome can be modified in neonates with reliance on
interpretation of these patterns of airflow and pressure.
Automated bedside EEV measurements are now available. The devices utilize helium dilution methodology and
early trials demonstrated safe use and reproducible measurements. Their use may allow a more individualized
approach to application of PEEP in infants.
Investigation of New Forms of Assisted Ventilation

The gold standard approach to the introduction of new therapies in medicine is the performance of prospective,
definitive, randomized clinical trials (RCT) with predetermined clinically relevant primary outcomes. This
approach has been followed belatedly but now enthusiastically in Neonatology. Part of the acceptance of the
results of RCTs devolves from effective blinding of the treatment group assignment, and the widespread
agreement that the control group received the best available therapy known at the time. When these principles
are applied to trials testing new approaches in assisted ventilation, there are several obvious limitations.
It is difficult to conduct a blinded clinical study of two different types of assisted ventilation, especially if different
ventilators are used. It is difficult to agree on what constitutes optimal standard of care to be used for the
control group. This applies not only to the minutia of management of the ventilation device itself, but also to the
ancillary supportive care. Even if agreement about standards of care is achieved, it is difficult to sustain
adherence to these standards of care.
Contemporary ventilators designed for application to neonates are very complex devices. This complexity
mandates a balance in the performance of clinical trials between the rigid prescriptions of ventilator
management vs. the field level studies with greater tolerance for physician specific practices. The latter study
design may have perhaps more applicability to broad populations, but runs the risk that center-to-center
variance in approach and in outcome may be greater, potentially diluting the signal of interest. Although these
issues are inherent in all prospective definitive trials, they are particularly acute when applied to testing modes
of assisted ventilation in neonates.
Identifying a clinically relevant outcome can be difficult. Death is the obvious outcome of interest, but for most
populations of neonates treated with assisted ventilation, it may be an insensitive marker and not informative
for other important outcomes. Utilizing a surrogate marker for death (e.g., need for ECMO) has worked well for
clinical trials of near term infants, but will not work well for babies ineligible for ECMO, (those less than 34 to 35
weeks gestation). For preterm infants, evidence of pulmonary injury (i.e., CLD) in combination with death is
commonly used. However, even these outcomes pose problems because of the poorly understood pathogenesis
of CLD and the role of factors other than assisted ventilation that contributes to its development and severity.
Combined outcomes melded into one also have the problem that reduction in one part of the primary outcome
(death) may increase the incidence of the adverse-associated outcome. Some studies testing different forms of
assisted ventilation were plagued by the problem that to qualify for the study, infants first were treated with one
form of assisted ventilation for a variable period before the application of different patterns of ventilation could
be undertaken. More recent trials, especially those testing HFV (47,48,49), have largely avoided this problem.
Some studies, even those purporting to show a benefit from an experimental form of ventilation, may
demonstrate the benefit because the control patients demonstrated a
P.612
higher than expected rate of adverse outcomes. Successful recruitment of sufficiently large populations in
sufficiently short time and the ability to stratify them in a meaningful way can also prove to be difficult. This is
increasingly true, given the role of economics in the decision regarding whether eligible patients for definitive
clinical trials are moved from one center to another to participate in those trials (50).
Despite these limitations and others, advances have occurred in the application of assisted ventilation using
randomized clinical trials. RCTs will remain valuable to better understand the limitations and strengths of the
many different patterns of assisted ventilation.
Conventional Ventilation
Available Modes and Details of Use
Several manipulations of applied pressure allow for an increase in PaO2. The physician decides whether to
increase FiO2 or PAW[gas] by considering the prior settings and balancing the possible harmful effects of
increasing PAW[gas] against those of increasing FiO2, recognizing that threshold limits are arbitrary. If the FiO2
is approaching 1.0, and SpO2 or PaO2 are unacceptably low, then other options must be invoked. During timecycled, pressure-limited ventilation, if PEEP is already 5 to 6 cm H2O, the PIP or the TI is increased. The use of
PEEP helps to maintain patent small airways and prevent collapse to airlessness of those alveoli already open.
Inspiratory pressures of <15 cm H2O usually are required to open collapsed or fluid-filled acinar areas. A
combination of an increase in the TI with 6 cm H2O PEEP may be helpful during the initial phase of assisted
ventilation for RDS. With subsequent opening of air spaces, the optimal TI may need to be decreased. The use
of an end-inspiratory pause or plateau should improve the distribution of inspired gas if there are regional
differences in airway resistance. However, if the alveolar pressure exceeds capillary pressure, there will be
tamponade of the pulmonary circulation and development of high [V with dot above]A[Q with dot above]areas.
Hypoxemia may persist during all combinations of ventilator settings in some conditions, and other underlying
abnormalities should be suspected. The clinician should always consider the degree of air leak around the
endotracheal tube when adjusting pressure and flow rates. Echo-cardiographic evaluation of the infant for
coexisting pulmonary vascular hypertension or structural or functional heart disease is then indicated. Other
management considerations are listed in Table 31-4.
If there is airway obstruction, as may occur with CLD or meconium aspiration, optimal ventilator settings may
differ from those used for RDS. Because there is a relatively long time constant, the gas flow rate should not be
too rapid, and there should be adequate time for expiration. One of the multiple available newer modes of
assisted ventilation may then be useful.
TABLE 31-4 MANAGEMENT CONSIDERATIONS
Utilize alternative mode of ventilation (e.g., HFV, SIMV,
A/C, PRVC)
Raise the hematocrit to 45-50% with packed erythrocyte
transfusions
Reposition baby into prone position if supine or into left or
right lateral positions

Use of sedation with or without paralysis; if infant is paralyzed,
discontinue its use

Change to a larger size endotracheal tube to diminish air leak
Consider repeated doses of exogenous surfactant beyond
24 h of age
Administer diuretics to change pulmonary fluid concentration
Increase cardiac output and/or systemic blood pressure with
agents
Undertake trial of NO
Consider use of corticosteroids
A/C, assist/control; HFV, high-frequency ventilation; IMV, intermittent mandatory ventilation; SIMV,
synchronous intermittent mandatory ventilation; PRVC = pressure regulated volume control.
Patient-Triggered Ventilation
The goal of patient-triggered ventilation is to maximize the efficiency of spontaneous breathing efforts although
minimizing the risk of insufficient ventilation or trauma to airways (51). All patterns of patient-triggered
ventilation require a rapidly responding sensor and transducer that can detect the onset of spontaneous
inspiratory effort and provide the mechanical initiation of machine-assisted ventilation during the early phase of
the infant's inspiration. Currently used methods to signal the initiation of inspiratory effort are listed in Table 315. The current methodology allows transduction to be accomplished in as short a time as 30 to 50 ms,
approximately one-tenth the duration of the inspiratory phase of a spontaneous respiratory cycle (Fig. 31-8)
(52). The means by which this signal is provided and the addition of other subtle but potentially important
changes in the capabilities of particular ventilators differentiate one type of conventional neonatal ventilator
from another (53). A list of available modes and their theoretical advantages is found (Table 31-6).
Although it is not clear that modern means of patient-triggered ventilation have achieved their optimum, these
methods have already gained widespread acceptance for three reasons. These include the clinical impression
that infants are more comfortable and less distressed although being ventilated with patient-triggered
ventilation; there may be at least modest improvements in pulmonary gas exchange (53) during patienttriggered ventilation; and there appears to be decreased need for sedation and muscle relaxation. Even with
patient-triggered ventilation, it is important to recognize the pitfalls that may occur when the synchronized
intermittent mandatory ventilation (SIMV) rates are too high or when the patient is allowed to breathe in the
assist/control mode (Fig. 31-9) (53). With assist/ control, hyperventilation may occur, especially if the sensor
P.613
for initiation of respiration is inappropriately sensitive and triggers ventilator breaths that are not associated with
patient inspiratory effort. If the goal is to avoid breaths triggered late in inspiration or during expiration, then
flow-triggering systems are less prone to auto triggering and have a shorter and more consistent response time
than impedance-triggered systems (54). The Pressure-Regulated Volume Control mode available with the
Siemens Series 300 ventilator has proven useful in correcting gas exchange problems in larger preterm infants
or term infants. However, the set tidal volume is arbitrary because of loss of volume to the compliant breathing
circuit leading to inaccuracy of its measurement (55). If end tidal CO2 is measured with the ventilator, then a
second measure of tidal volume (VT) is made at the breathing tube, which may be more reflective of actual VT.
No definitive clinical trial of this variable flow mode of ventilation compared to conventional patterns of
ventilation in the treatment of low-birth-weight (LBW) infants has been reported.
ventilation in the treatment of low-birth-weight (LBW) infants has been reported.

TABLE 31-5 AVAILABLE MECHANISMS FOR DETECTING ONSET OF RESPIRATION
Advantage
Body movement measurement

Transthoracic impedence
Airway pressure sensors
Airflow sensors
Pneumotachygraphy
Hotwire anaemometer
Disadvantage
No deadspace
No deadspace
Little deadspace
Position-dependent
Position-dependent
Speed
Sensitive
Sensitive
Added deadspace
Too sensitive
A limitation of many devices is that the individual breaths generated by the ventilator are monomorphic.
Proportional assist ventilation may be one way to overcome the problem and provide improved individualization
of support. With proportional assist ventilation, the relationship between patient-induced inspiratory effort and
ventilator response is interactive (56). During proportional assist ventilation the ventilator amplifies the patient's
effort throughout the inspiratory phase of the cycle. With each spontaneously generated breath, the patient can
individualize the machine-initiated tidal volume and flow patterns. Sensors monitor instantaneous flow rate and
volume of gas from ventilator to patient; the applied pressure then changes according to the equation of motion.
This system may allow for both greater patient comfort and reduction of peak airway pressure required to
sustain ventilation, with less likelihood of overventilation compared to assist/control modes. (56). Another newer
patient-interactive system is available with the Drager 8000 neonatal ventilator. The modality is called pressuresupport ventilation with volume guarantee. The goal is to provide stable tidal volumes but with the capacity to
alter inspiratory time and peak inflating pressure.
Figure 31-8 System response time, also known as trigger delay. The flow change trigger is set at 1.0 LPM. It
took 25 ms from the time this threshold was reached (vertical line on the left) until there was a measurable rise
in airway pressure (vertical line on the right). From Donn SM, Sinha SK. Controversies in patient-triggered
ventilation. Clin Perinatol 1998;25:49, with permission.
TABLE 31-6 MODES OF ASSISTED VENTILATION VIA ENDOTRACHEAL TUBE
Modes
IMV
SIMV
A/C
PAV
Theoretical Advantages
Pressure and time limited breaths

Fresh gas flow allowing efforts at spontaneous breathing
Synchronized, controllable number of assisted breaths
Every spontaneous breath assisted with a monomorphic wave
form
Respiratory pressure is servocontrolled and is proportional to
endogenous volume and flow

PSV
Inspiratory time and flow are variable but PIP is still fixed
PSV + VG
Time variable inspiration
Constant flow
Automatic attempt to deliver same tidal volume
IMV = Intermittent mandatory ventilation, SIMV = Synchronized intermittent mandatory ventilation, A/
C = Assist/Control, PAV = Proportional assist ventilation, PSV = Pressure support ventilation, VG =
Volume guarantee
Figure 31-9 Tidal volume tracings (inspiration = upward) demonstrating three patterns of ventilator
interaction with spontaneous breathing. In this illustration, tidal volume of spontaneous breaths is less than
that of ventilator breaths. A: Asynchronous intermittent mandatory ventilation (IMV) with ventilator breaths
delivered during spontaneous expiration. During IMV, ventilator breaths occur at a constant rate, with random
timing with respect to spontaneous breaths. B: Synchronous intermittent mandatory ventilation (SIMV) with
ventilator breaths delivered early in selected spontaneous inspirations. During SIMV, ventilator breaths occur
more irregularly, but the ventilator delivers the set rate synchronously with spontaneous breaths. C: Assist/
control mode, with ventilatory breaths delivered early in all spontaneous inspirations. The assist/control mode
delivers ventilatory breaths synchronously with all spontaneous breaths and may lead to increased ventilation.
From: Cleary JP, Bernstein G, Mannino FL, Heldt GP. Improved oxygenation during synchronized intermittent
mandatory ventilation in neonates with respiratory distress syndrome: a randomized, crossover study. J Pediatr
P.614
No studies have been reported to date which measured a clinically important outcome after a significant duration
of exposure to a patient interactive device compared to another similar device. One study did examine in a
crossover design in infants with respiratory distress syndrome the effect of pressure-support ventilation plus
volume guarantee, a volume targeted form of ventilation. Results of measurements of end expiratory volume
and of minute ventilation are shown Figs. 31-10 and 31-11 (57). No obvious improvement could be detected in
VE, a/A, or EEV when PCO2 was controlled.
The ability to reduce assisted ventilatory support in a controlled manner with patient-interaction devices is
improved compared to that with ventilator models available previously. Decremental changes in back-up rate,
PIP and FiO2 can be accomplished smoothly and cumulatively reduce the reliance on the trial-and-error
approach to reduction in ventilatory support necessary with older generation devices. One established adjunctive
therapy for VLBW infants is the use of methylxanthine therapy with discontinuation of assisted ventilation.
Routine use of this medical therapy reduces the need for reintubation and reintroduction of assisted ventilation
(58).
Figure 31-10 The relationship between the VE and the mode of ventilation (SIMV or PSV + VG). The difference
between the two modes is significant (*p = 0.012). From Olsen SL, Thibeault DW, Truog WE. Crossover trial
comparing pressure support with synchronized intermittent mandatory ventilation. J Perinatol 2002;22: 461,
with permission.
Pressure-Regulation vs. Volume Guidance

It is helpful to think of current modes of conventional rate assisted ventilation as either pressure-controlled or
volume targeted (59). The reintroduction of volume-controlled ventilation after unsuccessful use of this modality
in the 1970s allows more choices after decades of pressure- and time-controlled neonatal ventilation. Some
adjunctive features of volume-control ventilation include variable and limited Ti, decelerating inspiratory flow
rates and patient-ventilator interaction. They can provide variable pressure support to maintain relatively
constant minute ventilation. During pressure-controlled assisted breathing, there can also be pressure-support
breathspressure once generated is constant during Ti but flow is decelerating. In summary, availability of
many choices within the realm of conventional rate assisted ventilation allows greater individualization of
assistance between patients and in the same patient during different stages of pulmonary insufficiency.
However, definitive evidence of improvement in mortality or of decrease in clinically important morbidity with
these new tools, compared to older modes of assisted ventilation, is lacking.
Figure 31-11 The relationship between the EEV and the model of ventilation (SIMV or PSV + VG). The
difference between the two modes is significant (*P = 0.011). From Olsen SL, Thibeault DW, Truog WE.
Crossover trial comparing pressure support with synchronized intermittent mandatory ventilation. J Perinatol
2002; 22:461, with permission.
P.615
High-Frequency Ventilation
Principles of Use
High-frequency ventilation (HFV) is defined as the use of small tidal volumes at supraphysiologic rates to provide
oxygenation and ventilation. It generally refers to the use of rates above 150 breaths per minute. In some
cases, VT is less than the dead space. Thus, the mechanisms of high-frequency ventilation differ from the
conventional combination of conduction and diffusion. This mode can still achieve adequate gas exchange. Many
theoretical and experimental studies have demonstrated that a number of convective and diffuse mechanisms
act in concert to affect gas transport during high-frequency ventilation (Fig. 31-12) (60). There is augmented
mixing of gas in the airways, through increased energy of the gas molecules at high ventilator frequencies and
high flows. The end result is that fresh gas reaches the alveoli. Various patterns of high-frequency ventilation
are illustrated in Fig. 31-13 (61).
The major advantage of high-frequency ventilation seems to be its potential to prevent some deleterious
consequences of mechanical ventilation. When used in animal models with a strategy of optimizing lung
inflation. High-frequency oscillatory ventilation (HFOV) improved gas exchange and lung mechanics, promoted
uniform inflation, reduced air leak, and decreased the concentration of inflammatory mediators in the lung, as
compared with conventional mechanical ventilation (62,63). Attempts have been made to differentiate highfrequency ventilators based on several ventilator-specific factors (Table 31-7) (61). The distinction among the
types of high-frequency ventilators may be relevant to the appropriate matching of any one type of highfrequency ventilator to a particular part of the natural history of a neonatal pulmonary disorder.
Figure 31-12 Gas-Transport Mechanisms During High-Frequency Ventilation. The major gas-transport
mechanisms that are operative under physiologic conditions in each region (convection, convection and
diffusion, and diffusion alone) are shown. There are seven potential mechanisms that can enhance gas
transport during high-frequency ventilation: turbulence in the large airways, causing enhanced mixing; direct
ventilation of close alveoli; turbulent flow with lateral convective mixing; pendelluft (asynchronous flow among
alveoli as a result of asymmetries in airflow impedance); gas mixing as a result of velocity profiles that are
axially asymmetric (leading to the streaming of fresh gas toward the alveoli along the inner wall of the airway
and the streaming of alveolar gas away from the alveoli along the outer wall); laminar flow with lateral
transport by diffusion (Taylor dispersion); and collateral ventilation through nonairway connections between
neighboring alveoli. From Slutsky AS, Drazen JM. Perspective. Ventilation with small tidal volumes. N Engl J
Med 2002;347 (9):630, with permission.
Figure 31-13 Respiratory rate vs. tidal volume. HFJV, high-frequency jet ventilation; HFO, high-frequency
oscillation; HFPPV, high-frequency positive-pressure ventilation. From Slutsky AS. Nonconventional methods of
ventilation. Am Rev Respir Dis 1988; 138:175, with permission.
P.616
Ventilation Devices and Their Indications

Three main types of high-frequency ventilators are used to treat neonates: high-frequency jet ventilators
(HFJV), high-frequency flow interrupters (HFFI), and HFOV. These devices differ both in the way they generate
HFV breaths and in the parameters that can be adjusted to control ventilation and oxygenation (Table 31-7).
High-Frequency Jet Ventilation

High-frequency jet ventilators are designed for the delivery of high-flow, short-duration pulses of pressurized
gas directly into the upper airway through a specifically designed endotracheal tube lumen. These systems
operate at rates of 150 to 600 breaths/min. Exhalation is passive. Effective tidal volumes are greater than the
anatomic dead space. The ventilator is servocontrolled and operates to maintain a constant pressure at the
endotracheal tube tip. The technique of high-frequency jet ventilation has been investigated to determine its
efficacy in the management of VLBW infants to reduce mortality or the development of chronic lung disease in
this population (64,65). Using chronic lung disease as a outcome variable is difficult in any study because of the
multiple confounding factors that may develop (66).
The current status of jet ventilation as a primary mode of support for low-birth-weight infants is unclear. Keszler
and associates (65) conducted a multicenter study in which 130 patients were enrolled. Half were treated with
HFJV and half were treated with a continuation of conventional mechanical ventilation (nonsynchronized
intermittent ventilation). Virtually all the infants had been treated with exogenous surfactant. Mortality and need
for ventilatory support at 28 days of life were the same in the two groups, but there was a reduction from 40%
to 20% for continuing respiratory support at 36 weeks postconceptional age among infants treated with highfrequency ventilation. This study also showed the HFJV was superior to tidal ventilation for preventing
pulmonary interstitial emphysema. It is interesting to note that the 28-day BPD rate and the survival rate were
comparable to that found 10 years earlier in a study comparing HFOV with conventional ventilation in infants
without prior delivery of surfactant (67). These encouraging results are contrasted with another prospective
randomized study in a comparable but smaller population of VLBW infants (68). In that study with HFJV there
was an increased risk of adverse outcomes, defined as large intraventricular hemorrhage, periventricular
leukomalacia, or
P.617
death. No difference in incidence of chronic lung disease was noted. The finding of leukomalacia had been
previously associated with hypocarbia produced by treatment with HFJV during the first 3 days of life (68),
although no cause-and-effect relationship could be established.

TABLE 31-7 HIGH-FREQUENCY VENTILATORS
Flow Interruption
Jet Ventilator
Oscillatory Ventilators
Commercially
available
device
Infant Star
(Infrasonics, San
Diego, CA)
Bunnell Life Pulse Sensormedics

(Bunnell, Salt
(Sensormedics,
Lake City, UT)
Anaheim, CA)
Indications
Failure of
PIE; intractable
conventional
ventilation in VLBW
infants; PIE
Rate of conventional
breaths; frequency
of high-frequency
respiratory failure;prevention of ECMO
Yes
air leak; failure of

conventional
ventilation
Variable
inspiratory ontime rate of 240600 beats/min
Yes
Passive(?)
Passive
Active
Sudden decrease
in PCO2;
Sudden decrease in PCO2; respiratory
Variables
Use with
conventional
ventilation
Expiratory phase
Special precautions >2 kg; no

publishedtrials
Draeger Babylog
8000 plus (Draeger
Medical Inc,
Telford, PA)
Primary or rescue treatment of
Rate; mean airway

pressure; IE ratio
No
Can be used as
continuous flow,
cycled ventilator or
as HFOV*, with
flow and volume
monitoring
Passive
(conventional
vent.)/active
(HFOV)
alkalosis
respiratory
alkalosis
* HFOV is not available in the USA.
Hybrid Patterns of Ventilation

A hybrid device combining elements of jet and flow interruption to generate high-frequency ventilation, coupled
with low-rate conventional ventilation, is available (Infrasonics, San Diego, CA). This device is operated at 10 to
15 Hz, with pauses in the high-frequency ventilator operation during the conventional breath inspiratory phase
of ventilation. When operated in its mixed conventional and high-frequency mode, this device generates highpressure, short, and nonvariable inspiratory flow with a variable shut-off phase, allowing passive lung deflation.
One attractive feature is the minimal patient disturbance involved in changing from high-frequency to
conventional forms of ventilation. No published study has demonstrated that this device offers any significant
advantage as primary or rescue therapy compared to other types of high-frequency ventilation.
High-Frequency Oscillatory Ventilation

HFOV differs from the other two types by delivering smaller tidal volumes at higher rates and has an active
expiratory phase. The Sensormedics 3100A high-frequency ventilator (Sensormedics, Anaheim, CA) has an
additional feature of operating with an independently adjustable TI. This device has been licensed for use for
primary initial treatment of assisted ventilation and for alternative therapy in intractable respiratory failure
unresponsive to conventional mechanical ventilation in neonates. The oscillating movement of the diaphragm
causes active inspiratory and expiratory phases that drive the mixing of gas (Fig. 31-14). The amplitude of the
pressure wave (measured as delta pressure) is proportional to the tidal volume of the breath being delivered
to the patient, and is adjusted to manipulate changes in ventilation. The mean airway pressure affects
oxygenation, by changing the rate of gas flow into the circuit and the resistance to the gas flow out the
expiratory side of the circuit (69).
Figure 31-14 HFOV Principle.
Several prospective, randomized, controlled studies have been conducted using different types of HFOV (4749,67,70-73) compared to conventional ventilation. The rationale for the postulated superiority of HFOV was
that diminished distending pressures, combined with more adequate recruitment of and maintenance of lung
volume, would be associated with fewer dysplastic cellular changes of the lung and less frequent or milder BPD.
Cumulatively, these studies provide evidence that HFOV can be used safely and effectively in a wide variety of
newborn infants, although its effectiveness as rescue therapy is still debated (74,75).
One question raised, but not answered, by clinical studies of HFOV is the potential benefit gained by its
application at birth. Two studies using premature primate models of RDS demonstrated improved gas exchange
if HFOV was used without any prior conventional mechanical ventilation (76,77). Jackson and associates (76)
found decreased proteinaceous alveolar edema and improved gas exchange after 6 hours of HFOV applied from
the first breath. However, all animals in both treatment groups had evidence of pulmonary cellular injury. The
finding of both short-term benefit and concomitant cellular damage implies that HFOV used from birth could still
be associated with significant lung injury. These findings are consistent with those of Solimano and associates
(78), who showed that fluid and protein leaks in preterm lambs still occurred, although HFOV was applied from
the first breath after delivery.
A possible second indication for high-frequency ventilation is to preclude the need for more invasive pulmonary
support with extracorporeal membrane oxygenation (ECMO). Studies enrolling patients eligible for ECMO
demonstrated that approximately one-half responded favorably to HFOV and did not need ECMO (79,80). Most
babies responding to high-frequency ventilation were larger infants suffering from severe RDS. Other acute
pulmonary disorders, such as meconium aspiration pneumonia with severe airway obstructive changes as shown
by chest radiography, may not respond as well to oscillatory ventilation. Kinsella and associates (81) showed
that the delivery of inhaled nitric oxide (INO), in association with HFOV, produced an improved outcome
compared to either conventional ventilation with nitric oxide (NO) or HFOV alone. There was a reduced need for
ECMO, especially in near-term infants with RDS or meconium aspiration syndrome (Fig. 31-15). The synergistic
effect of these two treatment modalities may help reduce dependence on invasive ECMO support in this group of
fragile term infants.
Two large contemporary trials comparing high-frequency oscillatory ventilation and conventional ventilation in
preterm infants at high risk were done in the United Kingdom (49) and in the United States (48). The results of
the study by Johnson and associates (49) (that enrolled infants with a gestational age of 23 to 28 weeks,
assigned either to HFOV (n = 400) or to conventional ventilation
P.618
(n = 397)) were similar to those of a previous recent trial that also showed no difference between the modes of
ventilation (47); the combined primary outcome, death or chronic lung disease (BPD), occurred in approximately
two thirds of the infants, and similar proportions of infants in the two treatment groups survived without BPD. In
contrast, Courtney and associates (48) found that in infants with birth weight of 601 to 1200 g, high-frequency
oscillatory ventilation conferred a small but significant benefit in survival without bronchopulmonary dysplasia
(56% vs. 47%; p = 0.046). Infants assigned to HFOV (n = 234) had a lower rate of pulmonary hemorrhage, a
slightly higher rate of pulmonary interstitial emphysema, and were successfully extubated one week earlier
compared to infants assigned to synchronized intermittent mandatory ventilation (n = 250). Neither trial showed
an in-creased incidence of intracranial abnormalities or risk for air leak, two previously described complications
(82,83).
Figure 31-15 Percentage of patients responding to high-frequency oscillatory ventilation (HFOV); inhaled nitric
oxide (INO); or combined HFOV plus INO by disease category. More patients with RDS or MAS responded to
combination therapy with HFOV plus INO than to either treatment alone. Response to INO during conventional
ventilation was more effective than response to HFOV in patients without significant lung disease (other
category). Asterisk indicates p < 0.05. From Kinsella JP, Truog WE, Walsh WF, et al. Randomized, multicenter
trial of INO and high-frequency oscillatory ventilation in severe, persistent pulmonary hypertension of the
newborn. J Pediatr 1997;130:55, with permission.
There are differences between the designs of both trials that could explain the different results. However, the
majority of available evidence suggests that in the usual clinical circumstances, the choice of the mode of
ventilation does not affect the pulmonary outcome, which may be influenced more by prenatal risk factors, initial
resuscitation, and other aspects of neonatal care (84).
Complications of Assisted Ventilation

Airway-Associated Problems
Acute and chronic complications of mechanical ventilation may develop in neonates. All patients with tracheal
intubation demonstrate some degree of mucosal injury, usually squamous metaplasia or mucosal necrosis. In
most infants, there seems to be spontaneous healing without significant sequelae. Hoarseness and stridor often
occur after prolonged intubation but usually resolve in a few days. Persistent lesions, including laryngomalacia
and subglottic stenosis, may develop in some patients occasionally requiring tracheostomy. Duration of
intubation, pressure from oversized tubes, inadequate humidification, and airway pressure produced by the
ventilator contribute to the development of laryngotracheal lesions.
The duration of intubation is dictated by the need for continued ventilatory support. In some infants requiring
lengthy assisted ventilation tracheostomy may facilitate development of normal feeding patterns and social
interaction and bypass already traumatized areas of the upper airway. However, infant tracheostomy itself is
associated with a significant morbidity and mortality (85,86). Complications can include fatal loss of airway after
accidental extrusion of the tube, paratracheal soft tissue infection, severe paratracheal air leak resulting in
inadequate generation of PIP, and tracheal cartilage softening, making decannulation difficult. A bedside
evaluation of the airway using flexible fiberoptic endoscopy can be helpful in assessing need for additional, more
invasive procedures (87).
Pulmonary Air Leak Complications

Pulmonary air leak may occur as a complication of any of the life-threatening disorders of the newborn or as a
result of their treatment. The air leak may take the form of pulmonary interstitial emphysema,
pneumomediastinum, pneumoperitoneum, pneumopericardium, or pneumothorax. Pneumopericardium and
pneumothorax occurring
P.619
during positive-pressure ventilation require immediate treatment by evacuation of the free air.
Pneumothorax must be considered if there is abrupt worsening of the respiratory or circulatory status of an
infant at risk. Unilateral hyperresonance, decreased breath sounds, a shift of the apical cardiac impulse, and skin
mottling are useful clinical clues. High-intensity illumination may demonstrate the presence of a pneumothorax if
the room can be adequately darkened. A definite diagnosis often can be made only by radiographic examination.
The volume of the extrapulmonary air collection is not always a valid indication of tension. Interstitial
emphysema, often a precursor of pneumothorax, causes the lung to remain partly expanded, even when
intrapleural pressure is high. Bilateral pneumothorax may lead rapidly to death and must always be considered
in cases of severe deterioration.
Pneumothorax in otherwise asymptomatic infants often resolves without therapy. However, marked mediastinal
shift, coexisting pulmonary disease, or use of mechanical ventilation indicates a need for evacuation of the air.
Aspiration with a syringe and needle may be done as an emergency procedure but is rarely adequate by itself
and should be followed with tube thoracostomy.
Thoracostomy tubes should be sterile and made of nonreactive rubber or plastic. The wall thickness should be
sufficient to prevent kinking, and the lumen should be large enough to prevent occlusion by exudate. The
presence of at least two holes in the tube reduces the likelihood of occlusion by tissue. Polyvinylchloride feeding
tubes or 8Fr (2.6 mm) to 10Fr (3.3 mm) catheters are suitable for thoracostomy use. The tube is inserted by
grasping the tip with a clamp and pressing through a previously made incision through the pleura. The catheter
is inserted after a skin incision has been made and can be Z-tracked over a rib for a better seal. Catheters with
a trocar within the lumen often require considerable force to insert, and puncture of the lung and liver has been
reported with their use.
The thoracostomy tube is connected to continuous suction at a negative pressure of 10 to 15 cm H2O with an
underwater seal. A chest radiograph should be obtained soon after thoracostomy. If the pneumothorax has not
been evacuated, the infant should be repositioned and the tube stripped or, if necessary, a second tube should
be inserted.
A thoracostomy tube is left in place until air ceases to bubble from the tube and until the risk of recurrent
pneumothorax is reduced (i.e., until respiratory distress has subsided or mechanical ventilation is no longer
required). The tube is then clamped. If there is neither clinical nor radiographic evidence of recurrent
pneumothorax, the thoracostomy tube can be removed.
Liquid Ventilation
One experimental approach to reducing the lung injury associated with gas ventilation is liquid ventilation. Lung
injury may be the result of abnormal inflation patterns produced because of elevated alveolar surface tension.
Liquid inflation of the lungs with saline eliminates the alveolar gas-lung liquid interface with its tendency to
induce collapse and perhaps injury. However, saline is too poor a carrier of oxygen to supply the body. Liquid
perfluorocarbon solutions are able to dissolve large volumes of oxygen and carbon dioxide at 1 atm.
Perfluorocarbon solutions have been used as carriers for O2 and CO2 in moribund human infants (88).
Two methods of liquid ventilation are being tested. The first is total liquid ventilation, which uses a completely
perfluorocarbon-filled ventilator circuit, and a membrane oxygenator to prime the inspired liquid flow. The
second is perfluorocarbon-assisted gas exchange, in which a portion of the lung volume, the EEV, is filled with
perfluorocarbon liquids and the lungs are ventilated with a conventional infant mechanical ventilator. Liquid
ventilation is an attractive form of pulmonary rescue therapy because the perfluorocarbons seem to be nontoxic
and the technique is less invasive than ECMO. Definitive trials in neonates are yet to be reported.
MANAGEMENT IN THE INTENSIVE CARE SETTING

Management of respiratory problems requires specialized personnel working effectively as a team and
specialized equipment. A physician skilled in neonatal intensive care techniques should be available within the
unit at all times, and other specialists should be immediately accessible for consultation.
Nurses must be carefully trained in the techniques for the intensive care of infants. Necessary skills include
application of ventilatory support equipment, recognition of equipment malfunction, airway management,
assessment of ventilation, and use of monitoring equipment. Respiratory therapists are critical to the effective
use of respiratory equipment. Maintenance and calibration of all oxygen administration and oxygen-measuring
devices require the presence of a respiratory therapist within the hospital at all times.
Equipment needs for neonatal intensive care include wall sources of compressed air and oxygen, oxygen
dilutors, heating and humidification devices, and oxygen-monitoring systems with alarms. Critically ill infants
need continuous monitoring of temperature, respiratory rate, and heart rate by electrical devices with alarm
systems. In the acute phase of illness, or if an umbilical or peripheral arterial catheter is in use, blood pressure
monitoring and electrocardiographic display must be available.
The parents of severely ill infants need understanding and support. They experience feelings of anxiety, fear,
guilt, and hostility. Most families are ill equipped for the unexpected emotional and financial burden imposed by
the child's hospitalization. A social worker should be available exclusively to the neonatal intensive care unit to
provide assistance to parents by delineating parental concerns and
P.620
helping coordinate communication with the medical and nursing staff and other hospital personnel.
The physical design of the intensive care unit must facilitate the management of acute respiratory problems.
Each patient area should be large enough to accommodate the necessary personnel and the enormous amount
of equipment, without generating intolerable crowding. A small number of patients in each area of the nursery
facilitates parental visits and alleviates the overall level of stress. The proliferation of monitors and alarms may
result in monitor fatigue, with the risk that monitors are ignored. Although neonatology is often thought of as an
acute care specialty, it more accurately is categorized as a chronic care specialty, because of the many days and
weeks of specialized care that small, sick infants require. The physical design of intensive care units is catching
up to that new reality.
Adequate attention paid to these ancillary features of intensive respiratory care of the newborn helps ensure the
optimal outcome expected for critically ill infants.
REFERENCES
1. Tin W. Oxygen therapy: 50 years of uncertainty. Pediatrics 2002;110(3):615.
2. The STOP-ROP Multicenter Study Group. Supplemental therapeutic oxygen for prethreshold retinopathy of
prematurity (STOP-ROP), a randomized, controlled trial. I. Primary outcomes. Pediatrics 2000;105:295.
3. Askie L, Henderson-Smart D, Irwig L, et al. The effect of differing oxygen saturation targeting ranges on
long term growth and development of extremely preterm, oxygen dependent infants: the BOOST trial. Pediatr
Res 2002;51:378(abst).
4. Chow LC, Wright KW, the CSMC Oxygen Administration Study Group, et al. Can changes in clinical practice
decrease the incidence of severe retinopathy of prematurity in very low birth weight infants? Pediatrics
2003;111(2):339.
5. Schulze A, Whyte RK, Way RC, et al. Effect of the arterial oxygenation level on cardiac output, oxygen
extraction, and oxygen consumption in low birth weight infants receiving mechanical ventilation. J Pediatr
1995;126(5):777.
6. Benaron DA, Benitz WE. Maximizing the stability of oxygen delivered via nasal cannula. Arch Pediatr
Adolesc Med 1994;148(3):294.
7. Dunn PM. Localization of the umbilical catheter by post-mortem measurement. Arch Dis Child 1966;41:69.
8. Jackson JK, Biondo DJ, Kilbride H, et al. Can an alternative umbilical arterial catheter solution and flush
regimen decrease iatrogenic hemolysis while enhancing nutrition? Ped Res 2002;51(4): 373A.
9. Cassady G. Transcutaneous monitoring in the newborn infant. J Pediatr 1983;103:837.
10. Peabody JL, Gregory GA, Willis MM. Transcutaneous oxygen tension in sick infants. Am Rev Respir Dis
1978;118:83-87.
11. Hay WW Jr, Brockway J, Eyzaquirre M. Neonatal pulse oximetry: accuracy and reliability. Pediatrics
1989;83:717.
12. Hay WW, Thilo E, Curlander JB. Pulse oximetry in neonatal medicine. Clin Perinatol 1991;18:441.
13. Epstein MF, Cohen AR, Feldman HA, et al. Estimation of PaCO2 by two noninvasive methods in the
critically ill newborn infants. J Pediatr 1985;106:282.
14. McEvedy BAB, McLeod ME, Kirpalani H, et al. End-tidal carbon dioxide measurement in critically ill
neonates: a comparison of sidestream and mainstream monitors. Can J Anaesth 1990;37:322.
15. Rozycki HJ, Sysyn GD, Marshall MK, et al. Mainstream end-tidal carbon dioxide monitoring in the neonatal
intensive care unit. Pediatrics 1998;101(4 Pt 1):648.
16. Laffey JG, Kavanagh BP. Hypocapnia. N Engl J Med 2002;347(1):43.
17. Garland JS, Buck RK, Allred EN, et al. Hypocarbia before surfactant therapy appears to increase
bronchopulmonary dysplasia risk in infants with respiratory distress syndrome. Arch Pediatr Adolesc Med
1995;149:617.
18. Wiswell TE, Graziani LJ, Kornhauser MS, et al. Effects of hypocarbia on the development of cystic
periventricular leukomalacia in premature infants treated with high-frequency jet ventilation. Pediatrics
1996;98:918.
19. Wyatt JS, Edwards AD, Cope M, et al. Response of cerebral blood volume to changes in arterial carbon
dioxide tension in preterm and term infants. Pediatr Res 1991;29:553.
20. Kraybill EN, Runyan DK, Bose CL, et al. Risk factors for chronic lung disease in infants with birth weights
of 751-1000 grams. J Pediatr 1989;115:115.
21. Poets CF, Sari B. Change in intubation rates and outcome of very low birthweight infants: a populationbased study. Pediatrics 1996;98:24.
22. Verder H, Robertson B, Greisen G, et al. Surfactant therapy and nasal continuous positive airway pressure
for newborns with respiratory distress syndrome. N Engl J Med 1994;331:1051.
23. Martin RJ, Nearman HS, Katona PG, et al. The effect of a low continuous positive pressure on the reflex
control of respiration in preterm infants. J Pediatr 1977;90:976.
24. Miller MJ, Waldeman AC, Martin RJ. Continuous positive airway pressure selectively reduces obstructive
apnea in preterm infants. J Pediatr 1985;106:91.
25. Miller MJ, DiFiore JM, Strohl KP, et al. Effects of nasal CPAP on supgraglottic and total pulmonary
resistance in preterm infants. J Appl Physiol 1990;56:141.
26. Narendran V, Donovan EF, Hoath SB, et al. Early bubble CPAP and outcomes in ELBW preterm infants. J
27. So BH, Tamura M, Kamoshita S. Nasal continuous positive airway pressure following surfactant
replacement for the treatment of neonatal respiratory distress syndrome. Acta Paediatr 1994;35:280.
28. So BH, Tamura M, Mishima J, et al. Application of nasal continuous positive airway pressure to early
extubation in very low birthweight infants. Arch Dis Child 1994;72:F191.
29. Annibale DJ, Halsey TC, Engstrom PC, et al. Randomized controlled trial of nasopharyngeal continuous
positive airway pressure in the extubation of very low birthweight infants. J Pediatr 1994;124:455.
30. Tapia JL, Bancalari A, Gonzalez A, et al. Does continuous positive airway pressure (CPAP) during weaning
from intermittent mandatory ventilation in very low birth weight infants have risks or benefits? A controlled
trial. Pediatr Pulmon 1995;19:269.
31. Khalaf MN, Brodsky N, Hurley J, et al. A prospective randomized, controlled trial comparing synchronized
nasal intermittent positive pressure ventilation versus nasal continuous postive airway pressure as modes of
extubation. Pediatrics 2001;108(1):13.
32. Angus DC, Linde-Zwirble WT, Clermont G, et al. Epidemiology of neonatal respiratory failure in the United
States. Projections from California and New York. Am J Respir Crit Care Med 2001;164:1154.
33. Dreyfuss D, Saumon G. Ventilator-induced lung injury. Am J Respir Crit Care Med 1998;157:294.
34. Thibeault DW, Lang MJ. Mechanisms and pathobiologic effects of barotrauma. In: Merritt TA, Northway
WH Jr, Boynton BR, eds. Bronchopulmonary dysplasia. Contemporary issues in fetal nenoatal medicine.
Boston: Blackwell Scientific, 1988;82.
35. Thibeault DW, Mabry SM, Ekekezie II, et al. Lung elastic tissue maturation and perturbations during the
evolution of chronic lung disease. Pediatrics 2000;106(6):1452.
36. Thibeault DW, Mabry SM, Ekekezie II, et al. Collagen scaffolding during development and its deformation
with chronic lung disease. Pediatrics 2003;11:766-776.
37. Thibeault DW, Truog WE, Ekekezie II. Acinar arterial changes with chronic lung disease of prematurity in
the surfactant era. Pediatr Pulm 2003;36:482-489.
38. Rodenstein DO, Perlmutter N, Stanescu DC. Infants are not obligatory nasal breathers. Am Rev Respir Dis
1985;131:343.
39. Polgar G, Kong GP. The nasal resistance of newborn infants. J Pediatr 1965;67:557.
40. Wung JT, Stark FI, Indyk L, et al. Oxygen supplementation during endotracheal intubation of the infant.
41. Fanconi S, Duc G. Intratracheal suctioning in sick preterm infants: prevention of intracranial hypertension
and cerebral hypoperfusion by muscle paralysis. Pediatrics 1987;79:538.
42. Kelly MA, Finer NN. Nasotracheal intubation in the neonate: physiologic responses and effects of atropine
and pancuronium. J Pediatr 1984;105:303.
P.621
43. Tochen ML. Orotracheal intubation in the newborn infant: a method for determining depth of tube
insertion. J Pediatr 1979;95:1050.
44. Kohelet D, Goldberg A, Goldberg M. Depth of endotracheal tube placement in neonates. J Pediatr
1982;101:157.
45. Tarnow-Mordi WO, Reid E, Griffiths P, et al. Low inspired gas temperature and respiratory complications in
very low birth weight infants. J Pediatr 1989;114:438.
46. Golombek S, Truog WE. Acute effects of exogenous surfactant treatment in near term infants with RDS. J
Invest Med 1995;43:463.
47. Thome U, Kossel H, Lipowsky G, et al. Randomized comparison of high-frequency ventilation with highrate intermittent positive pressure ventilation in preterm infants with respiratory failure. J Pediatr
1999;135:39.
48. Courtney SE, Durand DJ, the Neonatal Ventilation Study Group, et al. High-frequency oscillatory
ventilation versus conventional mechanical ventilation for very-low-birth-weight infants. N Engl J Med
2002;347(9):643.
49. Johnson AH, Peacock JL, the United Kingdom Oscillation Study Group, et al. High-frequency oscillatory
ventilation for the prevention of chronic lung disease of prematurity. N Engl J Med 2002;347(9):633.
50. Stark AR, Davidson D. Inhaled nitric oxide for persistent pulmonary hypertension of the newborn:
implication and strategy for future high-tech neonatal clinical trials. Pediatrics 1995;96(6): 1147.
51. Amitay M, Etches PC, Finer NN, et al. Synchronous mechanical ventilation of the neonate with respiratory
disease. Crit Care Med 1993;21:118.
52. Donn SM, Sinha SK. Controversies in patient-triggered ventilation. Clin Perinatol 1998;25(1):49.
53. Cleary JP, Bernstein G, Mannino FL, et al. Improved oxygenation during synchronized intermittent
mandatory ventilation in neonates with respiratory distress syndrome: a randomized, crossover study. J
Pediatr 1995;126(3):407.
54. Hummler HD, Gerhardt T, Gonzalez A, et al. Patient-triggered ventilation in neonates: comparison of a
flow-and an impedance-triggered system. Am J Respir Crit Care Med 1996;154:1049.
55. Castle RA, Dunne CJ, Mok Q, et al. Accuracy of displayed values of tidal volume in the pediatric intensive
care unit. Crit Care Med 2002;30(11):2566.
56. Younes M, Puddy A, Roberts D, et al. Proportional assist ventilation, a new approach to ventilatory
support. Am Rev Resp Dis 1992;145:114.
57. Olsen SL, Thibeault DW, Truog WE. Crossover trial comparing pressure support with synchronized
intermittent mandatory ventilation. J Perinatol 2002;22:461.
58. Henderson-Smart DJ, Davis PG. Prophylactic methylaxanthines for extubation in preterm infants.
Cochrane Database Syst Rev 2002.
59. Sinha SK, Donn SM. Volume-controlled ventilation. Variations on a theme. Clin Perinatol 2001;28(3):547.
60. Slutsky AS, Drazen JM. Perspective. Ventilation with small tidal volumes. N Engl J Med 2002;347(9):630.
61. Slutsky AS. Nonconventional methods of ventilation. Am Rev Respir Dis 1988;138:175-183.
62. Meredith KS, deLemos RA, Coalson JJ, et al. Role of lung injury in the pathogenesis of hyaline membrane
disease in premature baboons. J Appl Physiol 1989;66:2150.
63. Yoder BA, Siler-Khodr T, Winter VT, et al. High-frequency oscillatory ventilation: effects on lung function
mechanics, and airway cytokines in the immature baboon model for neonatal chronic lung diseases. Am J
Respir Crit Care Med 2000;162:1867.
64. Keszler M, Donn SM, Bucciarelli RL, et al. Multi-center controlled trial comparing high-frequency jet
ventilation and conventional mechanical ventilation in newborn infants with pulmonary interstitial emphysema.
J Pediatr 1991;119:85.
65. Keszler M, Madanlou HD, Brudno DS, et al. Multicenter controlled clinical trial of high-frequency jet
ventilation in preterm infants with uncomplicated respiratory distress syndrome. Pediatrics 1997;100:593.
66. Bhuta T, Henderson-Smart DJ. Elective high frequency jet ventilation versus conventional ventilation for
respiratory distress syndrome in preterm infants (Cochrane Review). In: The Cochrane Library, Issue 2, 2003.
Oxford: Update Software.
67. The HIFI Study Group. High-frequency oscillatory ventilation compared with conventional mechanical
ventilation in the treatment of respiratory failure in preterm infants. N Engl J Med 1989; 320:88.
68. Wiswell TE, Graziani LJ, Kornhauser MS, et al. High-frequency jet ventilation in the early management of
respiratory distress syndrome is associated with a greater risk for adverse outcomes. Pediatrics 1996;98:1035.
69. Durand DJ, Asselin JM. Physiology of high-frequency ventilation. In: Polin R, Fox WJ, eds. Fetal and
neonatal physiology, 2nd ed. Philadelphia: WB Saunders, 1998.
70. Clark RH, Gerstmann DR, Null DM Jr, et al. Prospective randomized comparison of high-frequency
oscillatory and conventional ventilation in respiratory distress syndrome. Pediatrics 1992;89:5.
71. Gerstmann DR, Minton SD, Stoddard RA, et al. Results of the Provo multicenter surfactant high frequency
oscillatory ventilation controlled trial. Pediatrics 1996;98:1044.
72. HIFO Study Group. Randomized study of high frequency oscillatory ventilation in infants with severe
respiratory distress syndrome. J Pediatr 1993;122:609.
73. Ogawa Y, Miyaska K, Kawano T, et al. A multicenter randomized trial of high frequency oscillatory
ventilation as compared with conventional mechanical ventilation in preterm infants with respiratory failure.
Early Hum Dev 1993;32:1.
74. Bhuta T, Clark RH, Henderson-Smart DJ. Rescue high frequency oscillatory ventilation vs. conventional
ventilation for infants with severe pulmonary dysfunction born at or near term (Cochrane Review). In: The
Cochrane Library, Issue 2, 2003. Oxford: Update Software.
75. Bhuta T, Henderson-Smart DJ. Rescue high frequency oscillatory ventilation versus conventional
ventilation for pulmonary dysfunction in preterm infants (Cochrane Review). In: The Cochrane Library, Issue
2, 2003. Oxford: Update Software.
76. Jackson JC, Truog WE, Standaert TA, et al. Effect of high-frequency ventilation on the development of
alveolar edema in premature monkeys at risk for hyaline membrane disease. Am Rev Respir Dis
1991;143:865-871.
77. Meredith KS, de Lemos RA, Coalson JJ, et al. Role of lung injury in the pathogenesis of hyaline membrane
disease in premature baboons. J Appl Physiol 1989;66:2150.
78. Solimano A, Bryan C, Jobe A, et al. Effects of high-frequency and conventional ventilation on the
premature lamb lung. J Appl Physiol 1985;59:1571.
79. Carter MJM, Gerstmann DR, Clark MRH, et al. High-frequency oscillatory ventilation and extracorporeal
membrane oxygenation for the treatment of acute neonatal respiratory failure. Pediatrics 1990;85:159.
80. Clark RH, Yoder BA, Sell MS. Prospective, randomized comparison of high-frequency oscillation and
conventional ventilation in candidates for extracorporeal membrane oxygenation. J Pediatr 1994;124:447.
81. Kinsella JP, Truog WE, Walsh WF, et al. Randomized, multicenter trial of inhaled nitric oxide and highfrequency oscillatory ventilation in severe, persistent pulmonary hypertension of the newborn. J Pediatr
1997;130:55.
82. Moriette G, Paris-Llado J, Walti H, et al. Prospective randomized multicenter comparison of high-frequency
ventilation and conventional ventilation in preterm infants of less than 30 weeks with respiratory distress
syndrome. Pediatrics 2001;107:363.
83. Henderson-Smart DJ, Bhuta T, Cools F, et al. Elective high-frequency oscillatory ventilation versus
conventional ventilation for acute pulmonary dysfunction in preterm infants (Cochrane Review). In: The
Cochrane Library, Issue 3. Oxford: Update Software, 2001.
84. Stark AR. High-frequency oscillatory ventilation to prevent bronchopulmonary dysplasia: are we there yet?
N Engl J Med 2002; 347(9):682-683.
85. Massie RJ, Robertson CF, Berkowitz RG. Long-term outcome of surgically treated acquired subglottic
stenosis in infancy. Pediatr Pulmonol 2000;30(2):125-130.
86. Kremer B, Botos-Kremer AI, Eckel HE, et al. Indications, complications, and surgical techniques for
pediatric tracheostomiesan update. J Pediatr Surg 2002;37(11):1556-1562.
87. Downing GJ, Kilbride HW. Evaluation of airway complications in high-risk preterm infants: application of
flexible fiberoptic airway endoscopy. Pediatrics 1995;95:567.
88. Leach CL, Greenspan JS, Rubenstein SD, et al. Partial liquid ventilation with perflubron in premature
infants with severe respiratory distress syndrome. N Engl J Med 1996;335:761.
Chapter 32
Extracorporeal Membrane Oxygenation
Billie Lou Short
In 1944 Kolff and Berk observed that blood became oxygenated as it passed through
cellophane chambers of their artificial kidney membrane (1). This historic observation led
to the recognition by those involved in the fast-developing field of cardiopulmonary bypass
that blood could be oxygenated through a semipermeable membrane lung. In the bubble
and disk oxygenators used during the early 1950s for open-heart surgery, oxygen and
blood were mixed directly. This mixing resulted in considerable damage to blood products
and the potential for producing lethal fibrin emboli, making these systems unsuitable for
prolonged clinical use (2). For this reason, and in light of Kolff and Berk's findings,
attention was directed to the development of semipermeable membrane oxygenators that
separated blood and oxygen, decreasing or eliminating the risks of the earlier oxygenators.
The first membrane lung, which used an ethylcellulose membrane, was described by
Clowes in 1956 and was used successfully in open-heart surgery 1 year later (3). With this
report began the study of prolonged cardiopulmonary bypass and potential application of
extracorporeal membrane oxygenation (ECMO) as an artificial lung.
The 1960s witnessed intensive research on materials and techniques (2). Silicone
polymers, available in thin sheets that enhanced gas transfer through membranes, began
to be characterized and developed. The development of the Kolobow silicone membrane
lung made the field of ECMO possible, and clinical trials, using prolonged bypass or ECMO
as an artificial lung, began in the late 1960s (2,4,5 and 6).
The concept of an artificial placenta, a device capable of continuing ex utero the gasexchange functions of the placenta, developed in parallel with that of an artificial lung. In
1961 Callaghan and colleagues began using animal models of respiratory distress
syndrome of the newborn to test the efficacy of an extracorporeal oxygenation circuit as
an artificial placenta (7). During the early 1960s investigators, including Rashkind, White,
Dorson, and Avery, used ECMO as an artificial placenta for premature infants (8,9 and
10). Although the infants died, this was an extremely important period for the
development and refinement of the mechanical and surgical techniques that laid the
foundation for the subsequent success of ECMO.
It was not until ECMO therapy was applied to the term infant through the pioneering work
of Bartlett and colleagues that its full potential as a powerful therapy for infants in severe
respiratory failure was recognized. In 1976 Bartlett and his associates reported the first
neonatal ECMO survivor, a term infant with severe meconium aspiration syndrome (MAS)
(11). During the next 10 years, neonatal ECMO was used to treat 99 term infants with
respiratory failure in three centers in the United States and produced an overall survival
rate of 65%. Since 1986, ECMO therapy has developed explosively, and more than 19,000
infants have been treated in more than 90 ECMO programs, with an overall survival rate
of 77% (12). ECMO is also being used to support the cardiac patient postoperatively and
the older child in respiratory failure. This chapter will only address the use of ECMO in the
neonatal patient in respiratory failure.

The most common use for ECMO is in the term or near-term infant with failure to
oxygenate as a result of MAS, idiopathic persistent pulmonary hypertension (PPHN),
congenital diaphragmatic hernia (CDH), sepsis and pneumonia, or hyaline membrane
disease. Although overall survival is 77% nationally, the best results are for the MAS
(94%) and PPHN (79%) groups (Table 32-1). In the last 3 years, the survival rate for the
MAS patient at Children's National Medical Center in Washington, DC has been 100%,
PPHN 94%, and sepsis/pneumonia 100%. So although new therapies such as inhaled nitric
oxide (iNO) and high frequency ventilation (HFV) may be delaying the initiation of ECMO,
it does not appear that they have affected survival (13). As with Gill and associates, the
time at Children's National Medical Center on ECMO has increased from 5 days to 8.9 days
with the initiation
P.623
of these additional therapies pre-ECMO. At Children's National Medical Center, a 34- to 36week gestation infant with respiratory failure has not been placed on ECMO in the last 3
years due to therapies such as surfactant therapy and HFV. The survival rate for the CDH
population requiring ECMO has not increased over time and remains between 50% to 60%
nationally, perhaps because of the heterogeneity of this group of infants, who have
various degrees of pulmonary hypoplasia and pulmonary hypertension (14,15). Only 30%
to 40% of the CDH population require ECMO therapy, and those patients not requiring
ECMO have a survival rate close to 100%. Therefore, with ECMO therapy, the overall
survival rate for the CDH population as a whole has increased to more than 70% and in
some centers has been reported to be as high as 80% (16,17). Most centers are now
repairing the diaphragmatic defect in the critical CDH either on ECMO or immediately postECMO (14).
TABLE 32-1 RESULTS OF EXTRACORPOREAL OXYGENATIONa
Diagnosis
Meconium aspiration syndrome

Persistent pulmonary hypertension
Sepsis and pneumonia
Congenital diaphragmatic hernia
Total
Survival After Therapy (%)
94
79
75
84
53
77
From the Extracorporeal Life Support Neonatal Registry, July 2003, 19,061
patients.
INDICATIONS
Among the most controversial aspects of ECMO therapy have been the clinical criteria
used to determine its use (18,19,20,21,22,23,24,25 and 26). Because of the invasive
nature of ECMO therapy and the potential risks associated with this therapy, the criteria
are designed to select a population of infants who have an 80% or greater mortality risk
with conventional therapy. Assumptions about the ability of ECMO to increase survival are
valid only if the criteria are specific for each high-risk population. The ultimate test for the
efficacy of ECMO, and the predictive value of ECMO criteria, is a prospective, randomized
trial. Although two randomized trials have been completed, most centers have used
historic controls to develop their criteria (18,19,24,26). In the prospective, randomized
trial reported by O'Rourke and colleagues, a crossover design was used, which may have
skewed the predictions of their criteria. The criteria used in O'Rourke's study, which were
thought to predict a mortality over 80% based on retrospective data, predicted a mortality
of only 40% when used prospectively (24). It is imperative that all centers continually
evaluate their criteria, especially as less invasive therapies become available. The United
Kingdom collaborative randomized trial did not involve a crossover design and therefore
represents a more ideal trial for ECMO therapy (27). Of the 185 patients enrolled in the
study, 93 were allocated to ECMO therapy and 92 to conventional therapy, which included
high-frequency ventilation and iNO therapy. Mortality was significantly different between
the ECMO and conventional groups, 32% versus 59% (P = 0.0005). This benefit of ECMO
was sustained when severe disability at 1 year of age was taken into account (P = 0.002).
The CDH population in this study had the highest mortality rate, 82% in the ECMO group
and 100% in the conventional group. Severity of illness was also a predictor, with
mortality higher in the patients with an oxygen index (OI) of 60 or greater at the time of
entry into the study, indicating that early transfer of these patients to an ECMO center is
essential. The authors also noted that mortality was lower if the referring hospital was a
teaching hospital.
The potential risks associated with ECMO therapy include those associated with ligation of
the carotid artery (in venoarterial ECMO) and jugular vein, prolonged exposure to
systemic heparinization, alterations in pulsatile blood flow patterns, exposure to potential
toxins such as phthalate esters (i.e., plasticizer) from the circuit, and others yet to be
determined (28,29). With its long-term outcome still unknown, use of ECMO should be
limited to the term or near-term infant who has a 20% or less chance of survival with
conventional therapy. Although criteria developed at other centers are available, they are
based on the clinical management and patient populations in those centers and may not
be valid when applied to patients in other institutions (19,20,24). What is considered
maximal conventional therapy (e.g., hyperventilation) in one institution may not be used
in others. Differences in patient populations, such as the percentage of patients who are
inborn versus outborn, may significantly alter applicability of criteria from one center to
another. All ECMO centers should attempt to develop criteria based on their own
management techniques and patient population.
Several important inclusion criteria for ECMO are based on known complications of the
procedure. These are listed in Table 32-2.
Age and Weight Limitations

The requirement for systemic heparinization of the ECMO patient places significant
limitations on the population
P.624
that can be treated. Use of ECMO in the late 1960s and early 1970s in premature infants
weighing less than 2,000 grams or younger than 34 weeks of gestation resulted in a
significant mortality rate from intracranial hemorrhage (ICH) (30,31 and 32). This
increased risk may be a result of the combination of systemic heparinization with a more
direct effect of ECMO on the brain (33,34). Data from a review of the premature infant
treated with ECMO by Hirschl and colleagues indicates that the ICH rate in the infant down
to 32 weeks of gestation is lower than previously noted (35). Although the ICH rate is
lower in infants at 32 to 33 weeks of gestation than in the earlier experience with this
group, the rate is still close to 50%; thus, if one is considering ECMO therapy for this
population, the information given to parents should appropriately address the high risk for
ICH. A better understanding of the pathophysiology of the intracranial bleeds seen in the
ECMO population may allow alteration of the risk for this complication and thus lower the
gestational age cutoff in the future (33,34).
TABLE 32-2 INCLUSION CRITERIA FOR EXTRACORPOREAL MEMBRANE
OXYGENATION
Gestational age 34 weeks or birth weight 2,000 g

No uncontrolled bleeding complications
No major intracranial hemorrhage, e.g., not greater than grade II
Mechanical ventilation provided for 10 to 14 days
Reversible lung disease
Hematologic Limitations
The requirement for systemic heparinization places the infant with a significant
coagulopathy or with bleeding complications at extreme risk. All attempts should be made
to correct any coagulopathy before instituting ECMO.
The septic infant is of particular concern because of the commonly associated
coagulopathy. Although these infants are at an increased risk for bleeding complications
on ECMO, correction of their coagulopathy and meticulous heparin management have
resulted in successful treatment (36).
The necessity for heparinization during ECMO precludes the treatment of any infant with a
major ICH. Infants with grade I to II intraventricular hemorrhages or small parenchymal
hemorrhages can be treated if heparin management is monitored closely and activated
clotting times (ACTs) are kept low (e.g., 160 to 180 seconds).
Prior Mechanical Ventilation

The limit of 10 to 14 days of assisted ventilation before ECMO therapy is imposed because
of the probable development of chronic lung disease after aggressive assisted ventilation
of this duration. ECMO is unable to reverse this disease process within a safe period. After
3 weeks of ECMO, the risks for complications related to the ECMO procedure itself, such as
clot formation, nosocomial infections (e.g., neck wound infections), and mechanical
failures (e.g., tubing ruptures), begin to increase. The maximal time that a neonatal
patient can be kept on the ECMO circuit is unknown, but in view of the increasing risk of
complications and usual lack of response beyond this time period, most centers limit time
on the circuit to around 3 weeks. Infants with diseases, such as chronic lung disease, that
do not improve in a short period should not be considered for ECMO unless there is a lifethreatening underlying disease state, such as acute pulmonary hypertension or sepsis/
pneumonia, that can be rapidly reversed by ECMO.
Cardiopulmonary Disease
Candidates for ECMO must have reversible lung disease. Because of the cardiopulmonary
support provided by this therapy, ECMO has allowed many infants thought to have
irreversible lung disease to live. The diagnosis of irreversible lung disease has become
progressively more difficult to make (14,37,38).
Significant cardiac disease must be ruled out before ECMO, but infants with severe
reversible lung disease superimposed on congenital heart disease may be candidates for
ECMO support before cardiac surgery.
Risk Assessment and Mortality Criteria

If the infant is failing maximal conventional therapy, ECMO should be considered. The next
task is to predict which infants have a high mortality risk without ECMO.
Commonly used criteria (Table 32-3) are the alveolar-arterial oxygen gradient (AaDO2),
OI, and arterial partial pressure of oxygen (PaO2) levels less than 50 mm Hg over a
specific time period (19,20,23,39). The AaDO2 can be calculated as follows:
when the fraction of inspired oxygen (FiO2) is 1.00, PB is the barometric pressure, 47 is
the water vapor pressure, and PaCO2 is the arterial carbon dioxide pressure. The OI can
be calculated with the following equation:
in which MAP is the mean airway pressure.

Deciding when to transfer an infant to an ECMO center is a difficult task. Most infants with
disorders treated by ECMO improve without ECMO. Before the infant becomes too
moribund for transport, the referring physician must attempt to determine which infants
are at high risk for failing maximal conventional therapy. This is an enormous
responsibility, which can be eased by early consultation
P.625
with ECMO center personnel. Earlier studies by Boedy and associates (40) found that 12%
of their ECMO referrals died before arrival of the transport team or during interhospital
transport. Of importance, 32% of the deaths occurred in infants with CDH, indicating that
early referral of these patients was warranted. This concern becomes even more
important with the addition of therapies such as iNO. Centers providing iNO without ECMO
capabilities need to have a close association with an ECMO center, so consultation can be
obtained to determine when infants should be transported to the ECMO center (41,42).
TABLE 32-3 NEONATAL EXTRACORPOREAL MEMBRANE OXYGENATION CRITERIAa
b
aaDO2 605-620 mm Hg for 4 to 12 hr

b
Oxygen index 35 to 60 for 0.5 to 6 hr

b
PaO2 35 to 50 mm Hg for 2 to 12 hr
pH <7.25 for 2 hr with hypotension

Acute deterioration PaO2 <30 to 40 mm Hg
a
Criteria used only after maximal therapy instituted; 50% of centers use more
than one criterion.

AaDO2, alveolar-arterial oxygen gradient; PaO2, arterial partial pressure of oxygen.
b
Represents the ranges used in various centers
The studies that are performed before transfer of a patient to an ECMO center include an
echocardiogram to rule out heart disease; cranial ultrasound scan to rule out significant
ICH; coagulation studies, including a partial thromboplastin time, prothrombin time,
fibrinogen level, fibrin degradation products, and platelet count; calcium and electrolyte
levels; leukocyte count with a differential analysis; and hemoglobin and hematocrit levels.
These studies help the team at the ECMO center determine whether the patient should be
considered for ECMO and if so, assist them in anticipating difficulties.
Once the patient is admitted to the ECMO center, it must be determined if the patient is
an appropriate ECMO candidate. The ultrasound examination of the central nervous
system (CNS) is repeated to ensure that an ICH did not occur during transport. The
cardiac evaluation is repeated if there is any residual question about the possibility of
cardiac disease. Doppler flow techniques are used to document the severity of pulmonary
hypertension. This information can be used later if the infant does not wean from ECMO
appropriately. Serum electrolyte and calcium levels; hemoglobin and hematocrit; clotting
studies including fibrinogen level, fibrin degradation products, partial thromboplastin and
prothrombin times; platelet count; and baseline ACT should be obtained on admission to
detect abnormalities that require correction before ECMO.
Many ECMO candidates receive muscle relaxants before admission, making the neurologic
status difficult to evaluate. It is imperative to obtain a complete perinatal history,
including Apgar scores, resuscitation and seizure activity, and description of the neurologic
status of the infant before paralysis. Infants who have sustained severe irreversible
neurologic damage should not be considered for ECMO.
PROCEDURE
Venoarterial Method
Venoarterial (VA) ECMO involves the use of two catheters: the venous outflow catheter in
the right internal jugular vein with the tip in the right atrium and the arterial return
catheter in the right carotid artery with the tip at the junction with the aortic arch. Blood is
removed through the jugular catheter by means of gravity drainage into a venous
reservoir (Fig. 32-1). Blood is pulled out of the reservoir by a roller occlusion pump and
pushed through the membrane lung, where gas exchange occurs. Gas transfers across the
silicone membrane lung into the blood because of pressure gradients, increasing the
oxygen level and removing carbon dioxide (Fig. 32-2). Blood then enters the heat
exchanger, where it is warmed to body temperature and returned to the infant through
the arterial catheter.
Figure 32-1 Components of the venoarterial extracorporeal membrane oxygenation

circuit. (From Short BL. Physiology of extracorporeal membrane oxygenation (ECMO). In
Polin RA, Fox WW, eds. Fetal and neonatal physiology. Philadelphia: WB Saunders,
1992:932, with permission.)
This form of bypass provides pulmonary and cardiac support. Although most infants
requiring ECMO have only a pulmonary disorder, some have cardiac dysfunction secondary
to severe hypoxia and require the cardiac support that VA ECMO provides. Oxygenation is
achieved by allowing the pump to support as much of the cardiac output as is needed to
oxygenate the infant, usually 120 to 150 mL/kg/minute in the first few days.
It is easy to support and oxygenate with VA ECMO, and it remains the gold standard for
ECMO therapy. However, ligation of the carotid artery, alteration of pulsatile arterial blood
flow patterns, and the possibility that particles or air in the circuit may enter the cerebral
or coronary circulation remain concerns.
Venovenous Method
Venovenous (VV) techniques for ECMO have been developed because of the concerns
about carotid ligation. VV ECMO is currently achieved using a single double-lumen catheter
placed through the internal jugular vein into the right atrium (Fig. 32-3) (43,44). This
catheter has inflow and outflow ports that attach into the circuit. Because blood return and
outflow happen in the right atrium, recirculation can occur, resulting in limited
oxygenation with this technique (Fig. 32-4). Because the heart is the pump for VV ECMO,
the use of this catheter depends on intact cardiac function (43,44). The advantages of this
technique are the lack of necessity for ligation of the carotid artery,
P.626
maintenance of normal pulsatile blood flow, and the theoretical advantage that particles
entering the circuit enter the lungs rather than the cerebral or coronary circulation.
Disadvantages are the lack of cardiac support and limited oxygenation.
Figure 32-2 The silicone membrane lung promotes gas transfer across a gradient for
oxygen and carbon dioxide. The pore size does not allow blood products to cross. PCO2,
carbon dioxide partial pressure; PO2, oxygen pressure; PvCO2, mixed venous carbon
dioxide partial pressure; PvO2, mixed venous oxygen partial pressure. (From Short BL.
Physiology of extracorporeal membrane oxygenation (ECMO). In Polin RA, Fox WW, eds.
Fetal and neonatal physiology. Philadelphia: WB Saunders, 1992:932, with permission.)
Figure 32-3 The inflow and outflow characteristics of the venovenous catheter in the
right atrium. (From Short BL, O'Brien A, Poindexter C, eds. CNMC ECMO training manual.
Washington, DC: Children's National Medical Center, 1993, with permission.)
EQUIPMENT AND SYSTEMS

Most equipment currently used for ECMO therapy is modified cardiopulmonary bypass
equipment designed for short-term use. To ensure safe and effective use of the ECMO
equipment, the limitations of each piece of equipment should be understood and
considered before it is used for long-term bypass.
Figure 32-4 Recirculation occurs with the use of the venovenous extracorporeal
membrane oxygenation catheter. Flows greater than 400 mL/min result in greater than
50% recirculation and decreased oxygenation. (From Anderson HL 3rd, Otsu T, Chapman
RA, et al. Venovenous extracorporeal life support in neonates using a double lumen
catheter. ASAIO Trans 1989;35: 650-653, with permission.)
TABLE 32-4 EQUIPMENT FOR NEONATAL EXTRACORPOREAL MEMBRANE

OXYGENATION
Roller occlusion pump

Pump base
Venous return monitor or pressure monitoring module
Heating unit
Coagulation timer
Membrane mounting board
Oxygen blender
Carbon dioxide or carbogen tank
O2 and CO2 flowmeters
Inline temperature probes
Inline oxygen saturation monitor
P.627
There is no single ECMO machine. Each ECMO center must design an ECMO system by
using equipment evaluated and designed to meet space and other specific requirements of
their center. Bioengineering experts and cardiopulmonary perfusionists should be
consulted in the design and evaluation of the ECMO system. The basic equipment needed
for a complete system is listed in Table 32-4.
Space requirements and the possible need to transport the patient on ECMO should be
taken into consideration in the design of an ECMO system (Figs. 32-1 and 32-5). The
system should be on a cart or other movable base. The nondisposable equipment required
includes a roller occlusion pump, water bath to maintain normothermic bypass
temperatures, venous return monitor (VRM) or pressure monitoring system to insure
adequate blood flow through the circuit, gas flow meters for CO2 and O2 delivery into the
membrane lung, an inline venous saturation monitor, and membrane pressure monitors.
Necessary disposable equipment includes the ECMO catheters (i.e., venous and arterial for
VA ECMO; double-lumen venous for VV ECMO), tubing packs designed for the system,
venous reservoir, heat exchanger, and membrane lung.
Figure 32-5 The extracorporeal membrane oxygenation system used at the Children's
National Medical Center has a modular design.
Those involved in providing ECMO therapy must know the potential complications related
to each piece of equipment, especially the thrombogenic characteristics and flow
dynamics. Several of these concepts are discussed in this chapter; however, additional
information can be found elsewhere (45).
In VA ECMO, the venous catheter is the oxygenation catheter because the rate of blood
flow through this catheter determines the percentage of the cardiac output that is
supported by the ECMO pump. Oxygenation is determined by the percentage of cardiac
output passing through the ECMO membrane (i.e., artificial lung) and bypassing the
patient's lungs. If a small-gauge venous catheter is placed, minimal flow through the
ECMO circuit occurs, and oxygenation may be compromised. Blood flow rates (Q) are
directly proportional to the fourth power of the radius (r) of the tubing and inversely
proportional to the length (L) of the tubing:
To maximize flow into the circuit, a relatively short venous catheter with as large a lumen
as possible is used.
The arterial catheter, which supplies return of flow from the circuit into the arch of the
aorta, is the smallest-diameter component in the circuit and acts as the major resistance
component in the circuit. A small arterial catheter may cause significant backpressure and
lead to restricted blood flow, hemolysis, or eventual rupture of the circuit. Resistance to
flow (R) depends on the dimensions and geometry of the tubing and the characteristics of
the fluid used. Poiseuille's Law defines this concept by the following equation:
in which 8/ is the constant of proportionality, L (cm) is the length of the tubing, is the
coefficient of viscosity (i.e., Poise = dyne X sec/cm2), and r (cm) is the radius of the
tubing. The longer the tubing and the smaller the radius, the greater is the resistance. An
8-F (2.6-mm) catheter has much greater resistance than a 10-F (3.3-mm) catheter. A
short, large-lumen catheter is ideal, but the size of catheter placed is limited by the
diameter of the patient's carotid artery (46).
The double-lumen catheter used in VV ECMO has a small lumen for blood return, which
produces a relatively
P.628
high backpressure in the circuit. Kinking of this catheter significantly increases circuit
pressures and must be diligently avoided.
The VRM is an electronic device that monitors the blood flow from the patient into the
ECMO circuit (Fig. 32-1). Some systems do this through pressure monitoring instead of
volume monitoring. The VRM functions as a servoregulator for the ECMO system and
sounds an alarm and stops the pump if venous flow from the patient slows, ensuring that
aortic blood input equals venous output. If a VRM system is not in place and venous return
decreases without servoregulation, the roller pump continues to pump and causes the
tubing to collapse, creating a negative pressure that pulls gas out of solution in the blood
and air into the circuit at the connection points. The most common causes of loss of
venous return are malplacement of the venous catheter (usually in the inferior vena
cava), pneumothorax, or pneumopericardium; unrecognized bleeding (e.g., ICH,
hemothorax); kinking of the venous catheter; or placing an anchoring suture too tightly
around the catheter during cannulation.
The only membrane lung approved for long-term ECMO use in the United States is the
silicone membrane lung made by Medtronics (Minneapolis, MN). The 0.8-m2 membrane is
most commonly used for neonates and can support oxygenation up to a blood flow rate of
1 L/min. CO2 transfer is so efficient with this membrane that CO2 must be added to the
gases flowing into the membrane.
PATIENT MANAGEMENT
A team approach to the management of the ECMO patient is critical. Duties of the bedside
nurse, respiratory therapist, and ECMO specialist should be clearly delineated to ensure
efficient and effective care (47).
Daily Medical Management

Most neonatal patients, with the exception of those with CDH, require ECMO support for 7
to 8 days. During this period, the patient who was in respiratory failure and dying before
ECMO shows evidence of reversal of disease, can be slowly weaned off ECMO to minimal
ventilator settings, and can usually be extubated within 24 to 48 hours after coming off
the ECMO circuit. The rapidity of recovery is remarkable, given the severity of the illness
suffered by these infants before ECMO. For this level of recovery to occur in such a short
time, many physiologic changes must take place rapidly, making daily care of the infant a
fine art. Routine care must incorporate the fact that these infants are systemically
heparinized, and tasks such as suctioning of the airway should be done with caution.
As the lungs improve, less blood flow is required to pass through the artificial lung, and
the ECMO blood flow can be reduced. In the first few days, a blood flow rate of 120 to 150
mL/kg/min is required to oxygenate the infant (47). With improvement of the infant's
lungs, arterial blood gases improve, and the ECMO blood flow can be decreased by 10 to
20 mL/min. The venous saturation of blood in the ECMO circuit can be monitored
continuously, providing a representation of a mixed venous saturation level. However, this
saturation is measured in blood from the right atrium, and right atrial blood saturation
does not represent true mixed venous saturation if there are intracardiac shunts. Because
most infants on ECMO develop left-to-right shunts, often occurring at the level of the
foramen ovale, venous saturations must be interpreted in terms of other clinical signs. The
following concepts must be understood:
in which CVO2 is venous oxygen content, CaO2 is arterial oxygen content, [V with dot
above]O2 is oxygen consumption, flow is cardiac output, Hb is hemoglobin, and PO2 is
oxygen pressure.
For venous oxygen saturation to represent a true indication of relative arterial oxygen
content, several assumptions must be made: the cardiac output remains stable,
hemoglobin concentrations remain stable, and metabolic rate of the patient does not
change. Any one of these factors can cause a change in the venous saturation. Therefore,
the patient should be carefully evaluated before this parameter is used alone to wean the
ECMO flows. Obviously this parameter cannot be used in VV ECMO, where the venous
saturation only represents recirculation. Patient arterial blood gases are needed to
determine the pH and PaCO2 status of the patient and membrane lung. If the arterial
PaCO2 of the membrane decreases below 35 mm Hg during VA ECMO, a decrease in
respiratory rate may result because the brain detects the blood gas levels in blood from
the membrane lung during VA ECMO. If the patient's respiratory rate falls, and he or she
is on low bypass, the result is deterioration in blood gas status. The problem is corrected
by increasing the CO2 coming from the membrane to stimulate the infant to breathe. A
high membrane PCO2 may indicate membrane failure and is an emergency, and a normal
membrane PCO2 with an abnormal patient PCO2 indicates a change in the patient's clinical
condition such as development of pneumothorax or secondary pneumonia.
After being stabilized on ECMO, the infant is placed on lung-rest settings on the ventilator
(e.g., FiO2 =0.21, peak inspiratory pressures [PIP] =15 to 18 cm H2O; peak endexpiratory pressures [PEEP] =5 to 6 cm H2O; rate =10 to 15 breaths/min). It is typical for
the lungs to appear opaque on chest radiographs during the first 1 to 3 days of ECMO
(48,49,50). This is probably caused by the acute decrease in ventilatory settings, capillary
leak, activation of complement as a result of interaction of blood products with the
artificial surfaces in the circuit, and surfactant deficiency secondary to lung injury (51,52).
Lotze and associates (53,54) showed that surfactant replacement therapy in infants on
ECMO can decrease time on ECMO for all infants except those with CDH. A study
conducted
P.629
using surfactant before ECMO revealed that a significant number of infants in the lowmortality group (OI 15 to 22) could be kept off ECMO with the use of surfactant (55).
Lung compliance studies can help in predicting successful decannulation, especially in the
infant who is borderline, and when a decision is being made about removing an infant
from ECMO because of complications (49). A typical lung compliance curve is shown in
Fig. 32-6. An old-fashioned but effective technique for assessing pulmonary improvement
is to hand-ventilate the infant daily. When the chest moves easily with a peak pressure of
20 cm H2O or less, the infant can successfully come off ECMO.
Figure 32-6 The typical lung compliance curve for an infant with meconium aspiration
syndrome on extracorporeal membrane oxygenation (ECMO). Infants in this study were
successfully taken off ECMO if a lung compliance (CL) of 0.8 mL/cm H2O/kg was attained.
PDA, patent ductur arteriosus. (From Lotze A, Short BL, Taylor GA. Lung compliance as a
measure of lung function in newborns with respiratory failure requiring extracorporeal
membrane oxygenation. Crit Care Med 1987;15:226-229, with permission.)
Heparin is administered continuously into the ECMO circuit to prevent clotting (47).
Heparin management varies, depending on events before and during ECMO. Optimal
heparin management can achieve the level of heparinization needed to decrease the risk
for fibrin and clot formation in the circuit while minimizing the risk for bleeding
complications in the patient. Because heparinization must be evaluated rapidly and at the
bedside, most centers use the ACT (56). The ACT is determined in a system that uses
activators, such as glass beads, to initiate the clotting cascade. The specimen is warmed
to accelerate the clotting process. This test gives values of 80 to 120 seconds in a
nonheparinized infant, compared with standard nonactivated bleeding time values longer
than 5 minutes (56,57).
The primary cause of death in the ECMO population is ICH (30,31,58). The risk factors
associated with the development of an ICH include significant hypoxic or ischemic cerebral
insult before ECMO, sepsis with coagulopathy, or gestational age less than 37 weeks.
Initial heparin management is based on pretreatment risk factors (45).
Fibrin formation is related to flow rate; if there are low blood flow rates in the circuit, the
heparin dose is increased to decrease the risk for clot formation. At the beginning of an
ECMO run, blood flows are high, and the ACTs can be maintained in a lower range. At the
end of a run, the ACTs are increased, especially when the idling phase (i.e., 60 to 80 mL/
min) is reached. When the blood flow rate in the circuit is below 150 mL/min, ACTs are
increased to 200 to 220 seconds. Clinical factors that affect the ACT values are renal
function (heparin excretion is directly proportional to urine output), transfusion of
nonheparinized blood products or platelets, and a significant patent ductus arteriosus with
a left-to-right shunt that may decrease renal blood flow.
Fluid requirements while on ECMO range from 80 to 120 mL/kg/day. Electrolyte
requirements are significantly different from those before ECMO is started. Most infants
require little sodium, usually 1 to 2 mEq/kg/day, and a large amount of potassium, usually
4 to 5 mEq/kg/day. The rationale for these requirements is unknown. Although renin
levels increase on bypass, aldosterone levels decrease, and atrial natriuretic peptide levels
do not change (59,60). Calcium requirements range from 40 to 50 mg/kg of elemental
calcium per day.
Systemic hypertension is a common medical complication of ECMO (59,60 and 61).
Hypertension (mean blood pressure greater than 65 mm Hg for more than 3 hours) can
affect as many as 70% of these patients. Hypertension usually develops shortly after
cannulation and is transient, but 1% to 5% of infants require long-term antihypertensive
therapy. The risk of ICH is increased by hypertension (60). Although a subject of
controversy, the hypertension may be related to an increase in serum renin levels
(59,60). Fluid restriction and diuretic therapy may decrease the risk for prolonged
hypertension.
As pulmonary vascular resistance decreases, it is common to develop a left-to-right shunt
across the patent ductus arteriosus, resulting in oxygenation difficulty (62). Most of these
shunts close with fluid restriction and diuretic therapy. Few patients require surgical
intervention. Indomethacin should not be used in this population because it decreases
platelet aggregation. After the shunt closes, an immediate increase in the patient's PaO2 is
observed.
Cardiac stun is an interesting complication of ECMO therapy that also occurs in patients on
cardiopulmonary bypass (63). Cardiac stun occurs in infants with severe hypoxia or in
infants in whom the tip of the arterial catheter is placed too close to the coronary arteries.
This syndrome is characterized by a pulse pressure of 10 mm Hg or less on ECMO, with
the patient's PaO2 equal to or within 50 to 100 mm Hg of the pump PO2 (e.g., a pump PO2
of 400 mm Hg while on an FiO2 of 1.00, with a patient PaO2 between 300 and 350 mm
Hg). The electrocardiographic pattern is normal, indicating that electrical conduction is

normal, but cardiac output is markedly reduced. The pathophysiology is not understood
(63,64). When data on infants with cardiac stun while on ECMO at the Children's National
Medical Center were evaluated, there was a significant difference in pre-ECMO blood gases
relative to infants who did not develop cardiac stun. Infants in stun were sicker before
ECMO and had a higher
P.630
incidence of death from ICH, indicating the marked hypoxic-ischemic insult suffered by
these patients before ECMO. Treatment of cardiac stun consists of maintaining ECMO
blood flow rates high enough to supply appropriate cardiac output. Afterload reduction
does not improve cardiac function in this population (65). As stun resolves, the pulse
pressure returns to normal, and the patient's PaO2 no longer mirrors the pump PO2.
Weaning from VA ECMO occurs slowly as the arterial blood gases and venous saturations
improve. Idling flows, defined as flows at 10% of the cardiac output (i.e., 60 to 80 mL/
minute) are continued for 6 to 8 hours, and if blood gases are normal during this period,
most infants can be successfully taken off ECMO. A lung compliance measurement of at
least 0.8 mL/cm H2O/kg indicates that the infant can successfully come off ECMO (Fig. 326) (50). For patients on VV ECMO, the flows can be kept high and the membrane lung can
be capped off, so no gas flow occurs. Monitoring blood gases for 2 to 4 hours will
determine if the patient is ready to come off ECMO. Typical ventilator settings for infants
after completing an ECMO course are FiO2 of 0.30 to 0.40; rate of 30 to 40 breaths/
minute; PIP of 18 to 20 cm H2O; and PEEP of 5 to 6 cm H2O.
Patient Care after Extracorporeal Membrane

Oxygenation
A neuromuscular blocking agent with an intermediate half-life is used at decannulation to
help the patient breathe spontaneously as soon as possible after coming off ECMO. By the
end of the ECMO run, it is common for the ECMO patient to require high PaCO2 levels (i.e.,
45 to 55 mm Hg) to stimulate respiratory drive. Many of these infants have been receiving
narcotics such as fentanyl and may need to be weaned slowly from this therapy to avoid
withdrawal symptoms. PaO2 levels of 60 to 70 mm Hg are accepted, and every effort is
made to wean the patient from the ventilator. At Children's National Medical Center, the
average time to extubation post ECMO is 23 hours. The CDH patient is an exception; these
patients require slower weaning to ensure that pulmonary hypertension does not recur.
Hemoglobin, hematocrit, calcium, and electrolyte measurements should be obtained 6 to 8
hours after ECMO. These can then be monitored every 24 hours or as clinically indicated.
The platelet count should be followed closely (i.e., values every 12 hours for 24 hours)
because rebound thrombocytopenia may occur. Intravenous sodium is increased to 2 to 3
mEq/kg/day, and potassium supplementation is decreased to 1 to 2 mEq/kg/day.
After extubation, the infant usually requires oxygen therapy for another 5 to 7 days. Most
ECMO infants feed poorly and may require feeding by tube for a few days. The cause of
this problem is uncertain, but it is usually transient and does not indicate long-term
developmental problems.
Because not all intracranial abnormalities are detected by ultrasound, a computed
tomographic or magnetic resonance scan is recommended before discharge (66). A
baseline hearing screen and neurologic assessment are also recommended before
discharge. All infants should be followed in a neonatal high-risk follow-up program.
Outcome Data
Developmental outcome is encouraging, with most centers reporting that 60% to 70% of
ECMO survivors are normal at 1 to 2 years of age (67,68 and 69). Risk factors associated
with poor outcome include finding a severe abnormality on neuroimaging, chronic lung
disease, prematurity, and group B streptococcal sepsis (67,70-72). When ECMO survivors
were evaluated at 5 years of age though, Glass and associates (73) found that 37% were
at risk for school failure; however, Rais-Bahrami and colleagues (74) found a similar risk
in the population of near-miss ECMO patients, defined as those referred for ECMO but
who improved without ECMO. Wagner and associates (75) followed the ECMO population
studied in the Glass study into school and found that indeed, a high percentage had
academic problems (37%). Therefore, although a great majority of post-ECMO patients
are doing well, a thorough neuropsychologic evaluation prior to starting school should be
considered to identify those children who may benefit from special education programs.
The need for carotid artery ligation for VA ECMO has caused concern that right-sided CNS
lesions may result, but most studies have not shown this to be true (58,66,70,72).
Analysis of the first 360 patients treated at Children's National Medical Center did not
reveal lateralizing hemorrhagic or nonhemorrhagic abnormalities, but there was a high
incidence of posterior fossa hemorrhage, raising the concern that jugular venous ligation
might increase venous backpressure and the risk of hemorrhage (76). Data published by
Taylor and Walker (77) showed that decreased sagittal sinus blood flow velocity is
associated with ICH (70%) in the ECMO population. Whether this is cause or effect has yet
to be determined. Also noted in this study was a marked decrease in sagittal sinus blood
flow when the infant's head was turned to the left, obviously obstructing the left internal
jugular vein when the right was ligated, resulting in obstruction of venous flow. The
association of this with cerebral hemorrhage could not be determined because of the small
number of patients, but it does address the need for keeping the infants head midline
during the ECMO run. Because of this concern, many ECMO teams are now placing jugular
bulb catheters in the right internal jugular vein, advanced up to the jugular bulb area, to
drain the venous outflow from the brain into the venous side of the circuit, and thus
reduce the potential obstruction caused by the venous catheter and ligation of the jugular
vein.
The infant with CDH may have unique long-term problems, including significant
gastroesophageal reflux and chronic lung disease (14,38). These infants require close
follow-up in a multidisciplinary clinic to prevent problems such as failure to thrive and
respiratory compromise.
P.631
SUMMARY
Care of the ECMO patient requires highly trained nurses, respiratory therapists,
perfusionists, and physicians. The team must continually evaluate the treatment
modalities and use the information to improve techniques and define the indications for
ECMO therapy. Prior to the use of iNO, it was estimated that only 1,000 to 2,500 term
infants required ECMO each year in the United States (78). That number is obviously
going to be smaller since iNO has been shown to reduce the need for ECMO, making the
need for tailoring the development and expansion of ECMO centers based on regional need
even more important. Regionalization of ECMO programs can help to maintain cost control
and optimize quality of care.
REFERENCES
1. Kolff WJ, Berk HT. Artificial kidney: a dialyzer with a great area. Acta Med Scand
1944;17:121-134.
2. Kenedi RM, Courtney JM, Gaylor JDS, et al, eds Artificial organs. Baltimore:
University Park Press, 1976:87-88.
3. Clowes GHA Jr, Hopkins AL, Neville WE. An artificial lung dependent upon diffusion of
oxygen and carbon dioxide through plastic membranes. J Thorac Surg 1956;32:630-637.
4. Kolobow T, Stool EW, Sacko KL, et al. Acute respiratory failure. Survival following ten
days' support with a membrane lung. J Thorac Cardiovasc Surg 1975;69:947-953.
5. Zapol WM, Snider MT, Hil JD, et al. Extracorporeal membrane oxygenation in severe
acute respiratory failure. A randomized prospective study. JAMA 1979;242:2193-2196.
6. Gille JP, Bagniewski AM. Ten years of use of extracorporeal membrane oxygenation
(ECMO) in the treatment of acute respiratory insufficiency (ARI). Trans Am Soc Artif
Intern Organs 1976;22:102- 109.
7. Callaghan JC, de los Angeles J. Long-term extracorporeal circulation in the

development of an artificial placenta for respiratory distress syndrome of the newborn.
Surg Forum 1961;12:215-217.
8. Rashkind WJ, Freeman A, Klein D, et al. Evolution of a disposable plastic, low volume,
pumpless oxygenator as a lung substitute. J Pediatr 1965;66:94-102.
9. White JJ, Andrews HG, Risemberg H, et al. Prolonged respiratory support in newborn
infants with a membrane oxygenator. Surgery 1971;70:288-296.
10. Dorson W Jr, Baker E, Cohen ML, et al. A perfusion system for infants. Trans Am
Soc Artif Intern Organs 1969;15:155-160.
11. Bartlett RH, Gazzaniga AB, Jefferies MR, et al. Extracorporeal membrane
oxygenation (ECMO) cardiopulmonary support in infancy. Trans Am Soc Artif Intern
Organs 1976;22:80-93.
12. Extracorporeal Life Support Organization quarterly report, July 2003.
13. Gill BS, Neville HL, Khan AM, et al. Delayed institution of extracorporeal membrane
oxygenation is associated with increased mortality rate and prolonged hospital stay. J
Pediatr Surg 2002;37:7-20.
14. Van Meurs K, Short BL. State-of-the-art management of congenital diaphragmatic

hernia: a neonatologist's perspective. Pediatr Rev 1999;20:e79-e87.
15. Sanchez LS, O'Brien A, Anderson KD, et al. Best postductal PO2 and PCO2 do not
predict outcome of CDH infants in extremis, stabilized with ECMO prior to surgical
repair. Pediatr Res 1992;31: 221A.
16. Boloker J, Bateman DA, Wung JT, et al. Congenital diaphragmatic hernia in 120
infants treated consecutively with permissive hypercapnea/spontaneous respiration/
elective repair. J Pediatr Surg 2002; 37:357-366.
17. Kays DW, Langham MR Jr, Ledbetter DJ, et al. Detrimental effects of standard
medical therapy in congenital diaphragmatic hernia. Ann Surg 1999;230:340-348.
18. Bartlett RH, Roloff DW, Cornell RG, et al. Extracorporeal circulation in neonatal
respiratory failure: a prospective randomized trial. Pediatrics 1985;76:479-487.
19. Beck R, Anderson KD, Pearson GD, et al. Criteria for extracorporeal membrane
oxygenation in a population of infants with persistent pulmonary hypertension of the
newborn. J Pediatr Surg 1986;21:297-302.
20. Cole CH, Jillson E, Kessler D. ECMO: regional evaluation of need and applicability of
selection criteria. Am J Dis Child 1988;142: 1320-1324.
21. Dworetz AR, Moya FR, Sabo B, et al. Survival of infants with persistent pulmonary
hypertension without extracorporeal membrane oxygenation. Pediatrics 1989;84:1-6.
22. Krummel TM, Greenfield LJ, Kirkpatrick BV, et al. Alveolar-arterial oxygen gradients
versus the neonatal pulmonary insufficiency index for prediction of mortality in ECMO
candidates. J Pediatr Surg 1984;19:380-384.
23. Marsh TD, Wilkerson SA, Cook LN. Extracorporeal membrane oxygenation selection
criteria: partial pressure of arterial oxygen versus alveolar-arterial oxygen gradient.
Pediatrics 1988;82:162- 166.
24. O'Rourke PP, Crone RK, Vacanti JP, et al. Extracorporeal membrane oxygenation
and conventional medical therapy in neonates with persistent pulmonary hypertension
of the newborn: a prospective randomized study. Pediatrics 1989;84:957-963.
25. Wung JT, James LS, Kilchevsky E, et al. Management of infants with severe
respiratory failure and persistence of the fetal circulation, without hyperventilation.
Pediatrics 1985;76:488-494.
26. Sacks H, Kupter S, Chalmero TC. Are uncontrolled clinical studies ever justified? N
Engl J Med 1980;303:1059.
27. UK Collaborative ECMO Trial Group. UK collaborative randomised trial of neonatal

extracorporeal membrane oxygenation. Lancet 1996;348:75-82.
28. Karle VA, Short BL, Martin GR, et al. Extracorporeal membrane oxygenation exposes
infants to the plasticizer, di(2-ethylhexyl)- phthalate. Crit Care Med 1997;25:696-703.
29. Schneider B, Schena J, Truog R, et al. Exposure to di(2-ethylhexyl)phthalate in

infants receiving extracorporeal membrane oxygenation. N Engl J Med 1989;320:1563.
30. Cilley RE, Zwischenberger JB, Andrews AF, et al. Intracranial hemorrhage during
extracorporeal membrane oxygenation in neonates.Pediatrics 1986;78:699-704.
31. Revenis ME, Glass P, Short BL. Mortality and morbidity rates among lower birth
weight infants (2000 to 2500 grams) treated with extracorporeal membrane
oxygenation. J Pediatr 1992;121: 452-458.
32. Toomasian JM, Snedecor SM, Cornell RG, et al. National experience with
extracorporeal membrane oxygenation for newborn respiratory failure. Data from 715
cases. ASAIO Trans 1988;34: 140-147.
33. Short BL, Walker LK, Bender KS, et al. Impairment of cerebral autoregulation during
extracorporeal membrane oxygenation in newborn lambs. Pediatr Res 1993;33:289-904.
34. Short BL, Walker LK, Gleason CA, et al. Effects of extracorporeal membrane
oxygenation on cerebral blood flow and cerebral oxygen metabolism in newborn sheep.
Pediatr Res 1990;28:50-53.
35. Hirschl RB, Schumacher RE, Snedecor SN, et al. The efficacy of extracorporeal life
support in premature and low birth weight newborns. J Pediatr Surg 1993;28:13361340.
36. McCune S, Short BL, Miller MK, et al. Extracorporeal membrane oxygenation
therapy in neonates with septic shock. J Pediatr Surg 1990;25:479-482.
37. Newman KD, Anderson KD, Van Meurs K, et al. Extracorporeal membrane
oxygenation and congenital diaphragmatic hernia: should any infant be excluded? J
Pediatr Surg 1990;25:1048-1052.
38. Van Meurs KP, Newman KD, Anderson KD, et al. Effect of extracorporeal membrane
oxygenation on survival of infants with congenital diaphragmatic hernia. J Pediatr
1990;117:954-960.
39. Ortiz RM, Cilley RE, Bartlett RH. Extracorporeal membrane oxygenation in pediatric
respiratory failure. Pediatr Clin North Am 1987;34:39-46.
40. Boedy RF, Howell CG, Kanto WP Jr. Hidden mortality rate associated with
extracorporeal membrane oxygenation. J Pediatr 1990; 117:462-464.
41. American Academy of Pediatrics. Committee on Fetus and Newborn. Use of inhaled
nitric oxide. Pediatrics 2000;106:344- 345.
P.632
42. Clark RH. How do we safely use inhaled nitric oxide? Pediatrics 1999;104:296-297.
43. Anderson HL 3rd, Otsu T, Chapman RA, et al. Venovenous extracorporeal life
support in neonates using a double lumen catheter. ASAIO Trans 1989;35:650-653.
44. Rais-Bahrami K, Walton DM, Sell JE, et al. Improved oxygenation with reduced
recirculation during venovenous ECMO: comparison of two catheters. Perfusion
2002;17:415-419.
45. Short BL. Pre-ECMO considerations for neonatal patients. In: Arensman RM, Cornish
D, eds. Extracorporeal life support. Cambridge, MA: Blackwell Scientific, 1993:156-174.
46. Van Meurs KP, Mikesell GT, Seale WR, et al. Maximum blood flow rates for arterial
cannulae used in neonatal ECMO. ASAIO Trans 1990;36:M679-M681.
47. Short BL. Clinical management of the neonatal ECMO patient. In: Arensman RM,
Cornish D, eds. Extracorporeal life support. Cambridge, MA: Blackwell Scientific,
1993:195-206.
48. Taylor GA, Short BL, Kriesmer P. Extracorporeal membrane oxygenation:

radiographic appearance of the neonatal chest. AJR Am J Roentgenol 1986;146:12571260.
49. Lotze A, Short BL, Taylor GA. Lung compliance as a measure of lung function in
newborns with respiratory failure requiring extracorporeal membrane oxygenation. Crit
Care Med 1987;15: 226-229.
50. Keszler M, Subramanian KN, Smith YA, et al. Pulmonary management during
extracorporeal membrane oxygenation. Crit Care Med 1989;17:495-500.
51. Lotze A, Whitsett JA, Kammerman LA, et al. Surfactant protein A concentrations in
tracheal aspirate fluid from infants requiring extracorporeal membrane oxygenation. J
Pediatr 1990;116:435-440.
52. Anderson JM, Kottke-Marchant K. Platelet interactions with biomaterials and artificial
devices. In: Williams DF, ed. Blood compatibility, vol I. Boca Raton, FL: CRC Press,
1987:127-.
53. Lotze A, Knight GR, Martin GR, et al. Improved pulmonary outcome after exogenous
surfactant therapy for respiratory failure in term infants requiring extracorporeal
membrane oxygenation. J Pediatr 1993;122:261-268.
54. Lotze A, Knight GR, Anderson KD, et al. Surfactant (beractant) therapy for infants
with congenital diaphragmatic hernia on ECMO: evidence of persistent surfactant
deficiency. J Pediatr Surg 1994;29:407-412.
55. Lotze A, Mitchell BR, Bulas DI, et al. Multicenter study of surfactant (beractant) use
in the treatment of term infants with severe respiratory failure. Survanta in Term
Infants Study Group. J Pediatr 1998;132:40-47.
56. Hattersley PG. Activated coagulation time of whole blood. JAMA 1966;196:436-440.
57. Kay LA, ed. Essentials of haemostasis and thrombosis, 2nd ed. New York: ChurchillLivingstone, 1988.
58. Taylor GA, Short BL, Fitz CR. Imaging of cerebrovascular injury in infants treated
with extracorporeal membrane oxygenation. J Pediatr 1989;114:635-639.
59. Marinelli KA, Short BL, Martin GR, et al. Extracorporeal membrane oxygenation: its
effect on renin, aldosterone and natriuretic peptide. Pediatr Res 1989;25:241A.
60. Sell LL, Cullen ML, Lerner GR, et al. Hypertension during extracorporeal membrane
oxygenation: cause, effect, and management. Surgery 1987;102:724-730.
61. Boedy RF, Goldberg AK, Howell CG Jr, et al. Incidence of hypertension in infants on
extracorporeal membrane oxygenation. J Pediatr Surg 1990;25:258-261.
62. Martin GR, Short BL. Doppler echocardiographic evaluation of cardiac performance
in infants on prolonged extracorporeal membrane oxygenation. Am J Cardiol
1988;62:929-934.
63. Martin GR, Short BL, Abbott C, et al. Cardiac stun in infants undergoing
extracorporeal membrane oxygenation. J Thorac Cardiovasc Surg 1991;101:607-611.
64. Marban E. Myocardial stunning and hibernation. The physiology behind the
colloquialisms. Circulation 1991;83:681-688.
65. Martin GR, Chauvin L, Short BL. Effects of hydralazine on cardiac performance in
infants receiving extracorporeal membrane oxygenation. J Pediatr 1991;118:944-948.
66. Bulas DI, Taylor GA, O'Donnell RM, et al. Intracranial abnormalities in infants
treated with extracorporeal membrane oxygenation: update on sonographic and CT
findings. Am J Neuroradiol 1996; 17:287-294.
67. Glass P, Miller M, Short B. Morbidity for survivors of extracorporeal membrane

oxygenation: neurodevelopmental outcome at 1 year of age. Pediatrics 1989;83:72-78.
68. Schumacher RE, Palmer TW, Roloff DW, et al. Follow-up of infants treated with
extracorporeal membrane oxygenation for newborn respiratory failure. Pediatrics
1991;87:451-457.
69. Towne BH, Lott IT, Hicks DA, et al. Long-term follow-up of infants and children
treated with extracorporeal membrane oxygenation (ECMO): a preliminary report. J
Pediatr Surg 1985;20: 410-414.
70. Glass P, Bulas DI, Wagner AE, et al. Severity of brain injury following neonatal
extracorporeal membrane oxygenation and outcome at age 5 years. Dev Med Child
Neurol 1997;39:441-448.
71. Schumacher RE, Barks JDE, Johnston MV, et al. Right-sided brain lesions in infants
following extracorporeal membrane oxygenation. Pediatrics 1988;82:155-161.
72. Bulas DI, Glass P, O'Donnell RM, et al. Neonates treated with ECMO: predictive
value of early CT and US neuroimaging findings on short-term neurodevelopmental
outcome. Radiology 1995; 195:407-412.
73. Glass P, Wagner AE, Papero PH, et al. Neurodevelopmental status at age five years
of neonates treated with extracorporeal membrane oxygenation. J Pediatr
1995;127:447-457.
74. Rais-Bahrami K, Wagner AE, Coffman C, et al. Neurodevelop-mental outcome in

ECMO vs near-miss ECMO patients at 5 years of age. Clin Pediatr (Phila) 2000;39:145152.
75. Wagner AE, Coffman CE, Short BL, et al. Neuropsychological outcome and
educational adjustment to first grade of ECMO-treated neonates. Pediatr Res 1996;39
(Suppl 2):283. (abstr.)
76. Bulas DI, Taylor GA, Fitz CR, et al. Posterior fossa intracranial hemorrhage in infants
treated with extracorporeal membrane oxygenation: sonographic findings. AJR Am J
Roentgenol 1991;156: 571-575.
77. Taylor GA, Walker LK. Intracranial venous system in newborns treated with
extracorporeal membrane oxygenation: Doppler US evaluation after ligation of the right
jugular vein. Radiology 1992; 183:453-456.
78. Rais-Bahrami K, Short BL. The current status of neonatal extracorporeal membrane
oxygenation. Semin Perinatol 2000;24:406-417.
79. Short BL. Physiology of extracorporeal membrane oxygenation (ECMO). In Polin RA,
Fox WW, eds. Fetal and neonatal physiology. Philadelphia: WB Saunders, 1992:932.
Chapter 33
CARDIAC DISEASE
Michael F.Flanagan M.D Scott B. Yeager M.D.
Incidence
The incidence of congenital heart disease detectable by routine clinical
examination has been estimated to be 7.5 per 1,000 live births (1). The
incidence of congenital heart anomalies in neonates seen with detailed
echocardiographic examination is four- to 10-fold higher, with most of the
difference being clinically insignificant ventricular septal defects (20-50
per 1,000) and nonstenotic bicommissural aortic valves (2). Severe forms
of cardiac anomalies requiring cardiac catheterization or surgery, or
resulting in death, occur in 2.5 to 3 infants per 1,000 births (2,3). Almost
one-half of these are diagnosed during the first week of life. Additionally,
moderately severe forms of cardiac anomalies occur in another 3 per
1,000 live births, and another 13 of 1,000 live births have a
bicommissural aortic valve that may require care eventually (2). The
distributions of congenital heart anomalies in newborns seen at a primary
and a tertiary pediatric cardiac center are shown in Table 33-1.
Infant Mortality
Before aggressive intervention, Mitchell found that 2.3 of 1,000 live births
died with cardiac problems in infancy (1). Infant cardiac mortality in
developed countries has progressively declined over the last several
decades with better prenatal and postnatal recognition, and with the
development and refinement of definitive interventions and periprocedural management. The infant cardiac fatality rate in the United
States was 0.15 per 1,000 births in 2000, ranking tenth among leading
causes of infant death (4). At high-volume surgical centers more
commonly occurring cyanotic cardiac anomalies such as transposition of
the great arteries and tetralogy of Fallot have surgical mortality rates of
1% to 5% or less. Complex anomalies with the highest risk have also had
significant improvements in neonatal survival in developed countries. For
example, surgical survival of neonates with hypoplastic left heart
syndrome increased from 40% to 60% to 75% to 93% at specialized
surgical centers (5), with overall mortality in infancy approximately twice
this. Although surgical surgery outcomes are much improved, including in
babies with prematurity and multiple anomalies, prematurity and
associated noncardiac anomalies still strongly influence the potential for
salvaging infants with cardiac disease (Table 33-2) (4). In some
situations, the mortality attributable to these problems is considerable.
Long-Term Survival
This chapter focuses on infancy; however, discussions with parents concerning
their newborn with a cardiac anomaly often quickly, and appropriately, move to
the length and quality of life anticipated in later childhood and adulthood. It is
important that parents are accurately advised of likely and potential long-term
outcomes with current therapies by cardiologists or other practitioners up-to-date
with recent advances and outcome findings. In general, those with common
acyanotic anomalies such as uncomplicated septal defects or pulmonary valve
stenosis, and most of those with cyanotic anomalies such as simple transposition
of the great arteries now can anticipate an essentially near-normal life after
appropriate intervention. Even with many complicated anomalies, most can expect
to survive to at least mid-adulthood, although long-term survival is highly
dependent on the specific diagnosis, with the highest mortality now generally
occurring in infancy (see Table 33-2). With few exceptions, a cardiac operation or
catheter intervention can lengthen and improve the quality of life of a child with
heart disease. Even when the nature of the long-term management is unclear,
care has proceeded with the conviction that childhood survival often allows yet to
be planned later interventions, resulting from future progress in the field, to
provide even longer and better survival. The palliative shunt operations of 20 to
40 years ago unexpectedly produced candidates for later Fontan procedures. The
central principle continues to be where there is life, there is hope.
With reductions in mortality, and new generations of adults and older children
with repaired and palliated cardiac anomalies, it has become evident that cardiac
anomalies, and procedures utilized to treat them, sometimes have late residua
and sequlae, including neurologic and cognitive morbidity not evident until
preschool or school
P.634
age (6,7 and 8), and arrhythmias and ventricular dysfunction developing in
adolescence and adulthood. Most difficult, and not yet fully delineated have been
the neurologic and cognitive morbidity. These abnormalities appear rarely with
common septal and valvar malformations, and tend to be more frequent and
severe with more complicated anomalies and repairs. Cardiac, neurologic and
cognitive outcomes after repair of septal defects have appeared generally normal
(8). Although most children with complex anomalies and procedures, such as
hypoplastic left heart syndrome and operative circulatory arrest, have neurocognitive outcome within the normal range, on average, they have slower
development, lower IQ scores, and higher rates of learning disabilities and special
needs than normal. Moreover, a significant number have major impairments (8).
The etiologies are complex, and include possible genetic issues, coexistent brain
anomaly, potentially diminished cerebral blood flow or oxygenation in utero with

some anomalies (e.g., hypoplastic left heart syndrome), pre- and/or perioperative low cardiac output, cyanosis (See color plate), thrombo-embolism, or
intracerebral hemorrhage, and operative use of hypothermic circulatory arrest
(6,7 and 8). The possibility that brain injury or other injury may be acquired
prenatally, pretreatment or in the process of treatment should be understood.
TABLE 33-1 NEONATAL (FIRST MONTH) CARDIAC DIAGNOSTIC
DISTRIBUTIONa
Percent
Ventricular septal defect
41
Atrial septal defect secundum

Valvar pulmonary stenosis
12
11
Coarctation of the aorta

Tetralogy of Fallot
6
5
Cardiomyopathy
Transposition of the great arteries

Endocardial cushion defects
Hypoplastic left heart syndrome
3
3
2
Aortic stenosis
Tricuspid atresia
2
1
Malposition
Total anomalous pulmonary venous connection
Truncus arteriosis
1
1
1
Aortic-Pulmonary window
Hemitruncus
Interupted Aortic Arch
L-Transposition of the great arteries
Tricuspid valve diseases
Pulmonary atresia and intact interventricular septum
Single ventricle
1
<1
<1
<1
<1
<1
<1
Based on 361 patients diagnosed by echocardiography by age 1 month at

Childrens Hospital at Dartmouth. Neonates with patent ductus arteriosus,
persistent fetal circulation, arrhythmia and normal heart exams are not
included.
The goal is to provide a satisfying life through childhood and adulthood. The longrange future of patients undergoing intracardiac repair, arterial switch operations,
staged multiple complex palliative operations, Fontan operations, or cardiac
transplantation requires detailed discussion. The ramifications of treating a child
who has major anomalies must be discussed with the parents in understandable
language. The expected physical capabilities of the patient after treatment should
be delineated. After the physician is confident that the parents thoroughly

understand the known facts, he or she is free to express an opinion about what
may be best for the child.
Etiology
Heart formation is a fantastic metamorphosis regulated by many sequences of
genes. Given the remarkably complex orchestration of molecular and morphologic
processes in formation of the heart, even small genetic and/or environmental
changes in the control of these processes can have major, and variable,
consequences. Truly, it is wonderful that development occurs and it does, as often
as it does. Nevertheless, understandably, parents ask why their baby was born
with a cardiac abnormality and whether it is likely to recur with a subsequent
pregnancy.
Occasionally, children with isolated cardiac anomaly have a parent who has
survived with cardiac anomaly, or another family member with a cardiac anomaly,
but more often, there is no family history. However, an inheritable defect in a
single gene (e.g., Marfan syndrome) or a chromosomal anomaly (e.g., trisomy) is
identified in a significant minority of patients with cardiac malformations. More
commonly there may be susceptibility from inherited or acquired mutations in two
or more genes, perhaps with additional alterations in gene transcription or
posttranscriptional processes from maternal-fetal folate metabolism, or fetal
exposure to specific pharmacologic, biochemical, infectious and environmental
factors that cumulatively surpass a threshold of liability (9). These result in
pathogenetic changes in embryonic development, including defects in
mesenchymal tissue migration (tetralogy of Fallot, truncus arteriosus, interrupted
aortic arch, malalignment conal-septal ventricular septal defects, transposition of
great arteries), extracellular matrix defects (endocardial cushion defects),
abnormal cell death (muscular ventricular septal defect, Ebstein anomaly),
targeted growth (anomalous pulmonary vein connection, single atrium), defective
situs and cardiac looping (heterotaxy syndromes, L-transposition), and secondary
effects from alterations in blood flow in the right heart (secundum atrial septal
defect, pulmonary valve stenosis and atresia) or left heart (hypoplastic left heart
syndrome, coarctation of aorta, aortic valve stenosis, patent ductus arteriosus)
(9,10,11 and 12).
Approximately 13% of children with cardiac anomalies have chromosomal
syndromes associated with cardiovascular malformation. Another approximately
8% to 13% of children have inheritable syndromes with associated cardiovascular
abnormalities (11,12 and 13). The genes affected in many of these syndromes
have been identified (9,14,15 and 16) (Table 33-3). The most common human
mutations are in a critical 30 gene region of chromosome 22q11 involved in neural
crest and cardiac development that cause DiGeorge
P.635
and velocardiofacial syndromes and associated conotruncal and aortic arch

malformations (15). Other mutations in the genes encoding the extracellular
matrix proteins fibrillin-1 and elastin are responsible, respectively, for Marfan and
Williams syndrome. Genetic syndromes associated with abnormal cardiac
development are generally associated with specific cardiac malformations, for
instance Down syndrome and endocardial cushion defects, Williams syndrome and
supravalvar aortic stenosis, DiGeorge syndrome and tetralogy of Fallot, truncus
arteriosus, and interrupted aortic arch (Table 33-3). Recognition that a child has a
syndrome associated with a cardiac anomaly, or vice versa, should prompt an
investigation for possible associated anomalies (16,17,18,19 and 20).
TABLE 33-2 INFANT (FIRST-YEAR) CARDIAC DIAGNOSTIC
DISTRIBUTION AND MORTALITY
Diagnosis
Frequency (%)
Mortality (%)
Ventricular Septal Defect
31
0.7
Pulmonary Valve Stenosis

Atrial Septal defect
19
12
0.5
0.2
Aortic Valve Stenosis

Atrio-ventricular Canal
Septal Defects
8
5
4
1.8
Coarctation of Aorta
D-Transposition of Great
Arteries
4
4
3.5
1.0
Tetralogy of Fallot with

pulmonary outflow stenosis
2.2
Malposition
Hypoplastic Left Heart
Syndrome
Pulmonary Valve Atresia
with Intact Ventricular
Septum
Pulmonary Valve Atresia
with Tetralogy of Fallot
Double Outlet Right Ventricle
Total anomalous pulmonary
venous connection
Truncus arteriosis
Single Ventricle
Tricuspid Atresia
Interupted Aortic Arch
Aortic-Pulmonary window
Hemitruncus
3
2
9
22
11
0.9
0.8
6
7
0.7
0.7
0.5
0.3
0.2
0.1
5
<0.1
<0.1
<0.1
<0.1
<0.1
L-Transposition of Great
0.1
Arteries
Total
2.2
<0.1
Data is based on 5,182 infants diagnosed at Childrens Hospital, Boston

between January 1,1998 and January 1,2002, and their overall mortality
by age 1 year. Babies with patent ductus arteriosus and primary
cardiomyopathy were not included. Mortality by age 1 year was
significantly influenced by coexistent prematurity and noncardiac disease.
Most children with cardiac malformations, even those with tetralogy of Fallot, have
isolated cardiac anomalies, without generalized syndrome or other apparent
abnormality. Many specific cardiac anomalies are rarely associated with a
noncardiac syndrome, for example, transposition of the great arteries, and
pulmonary atresia with intact ventricular septum (Table 33-4). Although the
molecular biology of cardiovascular development is being unraveled, the specific
genetic and/or environmental causes of isolated cardiac anomalies remain
unknown in most individual cases (14,15,21). Mutations in 22q11 cause 20% to
30% of isolated conotruncal and aortic arch malformations. Other mutations are
involved with at least some of the remainder (15). For example, alterations in
cardiac specific transcription factor control genes, or their function (e.g., NKX2.5,
TBX5), sometimes in combination with genes that they control, are being
identified in children with isolated cardiac anomalies (15).
Although the occurrence of cardiac malformation has been mostly constant year
by year, and by location, there are exceptions. Biochemical and chance events
during fetal development may play roles in causation of new genetic mutations
and in transcriptional and posttranscriptional processes (9,14,15,22). For
instance, the risk for development of conotruncal anomalies is significantly
influenced by maternal-fetal intake and metabolism of folic acid and homocysteine
(24,25,26,27 and 28). Although fetal exposure to specific pharmacologic,
biochemical, infectious and environmental factors may increase the risk for
developing a cardiac abnormality (Table 33-5) (9,29,30 and 31), these factors
alone do not appear to explain most cases. In individual cases, it is usually
difficult or impossible to identify specific extrinsic factors that may have modified
the baby's genotype or genotypic expression.
A singular abnormality results in characteristic complex cardiac malformation by
altering migration or function of embryonic primordial cells, such as in the neural
crest or endocardial cushion, before formation of cardiac structures in the
conotruncus or atrioventricular valves, respectively. Animal studies have
demonstrated that embryonic
P.636
P.637
P.638
cervical neural crest cells migrate into the thorax and contribute to formation of
the aortic arch and conotruncal outflow region of the heart. Blockage of the
normal function of these embryonic neural crest cells results in aortic arch
anomalies including aortic interruption; conotruncal abnormalities including
tetralogy of Fallot, truncus arteriosus, and transposition; and ventricular inlet
anomalies including tricuspid atresia and double-inlet single left ventricle (11,14).
Cells in the embryonic endocardial tissue undergo a different developmental
sequential process controlled by a large number of factors. Perturbation of specific
steps in these embryonic cell process changes developmental sequences in
characteristic ways and alters blood flow patterns affecting vascular growth
downstream in characteristic ways (9,10 and 11). Because growth of specific
cardiovascular structures is flow dependent, limitation of flow can cause additional
hypoplasia of downstream structures (10,11). For example, a mildly stenotic
bicommissural aortic valve may decrease blood flow through the aortic isthmus
and result in coarctation.
TABLE 33-3 CONGENITAL DISORDERS ASSOCIATED WITH CARDIAC DISEASE

Selected Disorders
Identified Gene
(s)
Chromosome
Location
% Heart
Disease
Cardiovascular
Anomalies
Autosomal Dominant
Alagille Arteriohepatic
dysplasia
Jagged 1
20p12
100
multiple PA stenosis
and hypoplasia,
PDA, ASD, VSD
Apert acrocephalosyndactyly FGFR-2

Beckwith-Wiedemann
BWS
syndrome
de Lange syndrome
CATCH 22/ DiGeorge
syndrome
10q26
11pter-p15.4
CDL
3q26.3
DiGeorge
22q11
15 ?
30
>50
chromosome
region
VSD, hypoplastic
PDA
HCM, ASD, VSD, TF,
PDA
VSD, ASD, PDA, AS,
EFE
TF, interrupted Ao
arch, truncus
arteriosus, right Ao
arch
Goldenhar hemifacial
microsomaia
Hereditary hemorrhagic
telagiectasia Osler-WeberRendu
Holt-Oram Heart-hand
HFM
endoglin
14q32
9q34.1
15
100
TBX5
12q24.1
100
syndrome
Klippel-Feil syndrome
Marfan syndrome
Neurofibromatosis type I
Noonan syndrome
Rubinstein-Taybi
Broad thumb-hallux
syndrome
TF, VSD, PDA,

COARC
pulmonary and
systemic AVM,
anuerysms,
angiectasia
ASD-2, VSD or PDA
in 2/3, conduc.
block, HLHS, TAPVC,
truncus art.
May-70
VSD, dextrocardia
up to 100 % Ao anuerysm; AR,
KFSL
fibrillin-1
8q22.2
15q21.1
NFI
NS 1
CREB binding
protein
17q11.2
12q22-qter
16p13.3
rare
7p21,10q26
22q11
?
80
5q32-q33.1
9q34
7q11.2
10
30
5080
Saethre-Chotzen syndrome TWIST, FGFRShprintzen velocardiofacial 3&2

syndrome
DiGeorge
chromosome
region
Treacher Collins syndrome TCOF1
Tuberous sclerosis
TSC1
Williams-Beuren syndrome elastin
35
MR, TR & prolapse

PS, COARC, renal
artery stenosis
PS/dysplasia, PDA,
HCM, COARC
VSD, PDA, ASD,
COARC, PS
various, subvalvar
AS
VSD, TF, right Ao
arch, aberrant right
subclavian artery
ASD, VSD, PDA
rhabdomyomas,
rarely Ao aneuerysm
supravalvar AS,
small aorta,
stenoses LCA,
multiple PAs,
cerebral & renal
arteries
Autosomal Recessive
Carpenter
?
acrocephalopolsndactyly type EVC
7q32-q34
4p16
iduronidase
4p16.3
iduronate 2-
Xq28
sulfatase
GNS
12q14
5q11-q13
Coffin-Siris fifth digit

syndrome
Ellis van Creveld
Mucopolysaccharidosis type1 arylsulfatase B

type 2
?
?
7q32.1
type 3D
SLOS
unknown
type 6
unknown
7q11.23
Pierre Robin syndrome

Smith-Lemli-Opitz
multiple-peroxin5,2,6,12
33
?
5060
>50
2q32.3q33.2
20/100
33
PDA, PS, VSD, TF,

TGA, ASD
ASD, VSD, TF, PDA
single atrium,
primum ASD,
COARC, HLHS
all types have
valvular disease,
coronary disease
(type 2)
TF, COARC,
pulmonary
hypertension
syndrome 1&2
Trombocytopenia absent
VSD, PDA, ASD, TF,

AV canal, COARC
radius
ASD, TF
Zellweger cerebro-hepato-
VSD, ASD, PDA
renal syndrome
Selected Chromosomal
Disorders
Trisomy 13 Patau syndrome
Trisomy 18 Edwards
syndrome
Trisomy 21 Down syndrome
+ 8 Mosaicism
+ 9 Mosaicism
XO Turners syndrome
4p- Wolfe syndrome
5p- Cri-du-Chat syndrome
7q13q18qring 18
10p trisomy
10q24 trisomy
22 + Cat eye syndrome
Fragile X
Syndromes with unknown

etiology
Asymmetric crying facies
CHARGE Association
VACTERL Association
13
80
PDA, VSD, ASD,
18
90100
COARC, AS, PS
21
8
9
4050
25
70
VSD, polyvalvular,
ASD, PDA
AV canal, VSD,
X
4p
>50
33
ASD1&2, PDA, TF
VSD, PDA, CoAo, PS
5p
7q
13q-
20
20
common
VSD, PDA, LSVC

bicusp AV, COARC,
Ao anuerysm, AS,
18q
18
25
20
VSD
VSD, ASD, COARC
10p
10q24
22
30
50
40
50
44
6575
10
VSD, PDA, ASD, PS

VSD, ASD, PDA, PS
COARC, PA
hypoplasia, HLHS,
LSVS
AV canal, VSD, ToF
TAPVC, TF
Ao root
enlargement, MR
VSD
TF, DORV, ASD,
VSD, PDA, COARC,
AV canal
VSD, ASD, TF
Nonrandom Associations
cleft lip and palate
diaphramatic hernia
lung agenesis
omphalocele
intestinal atresia
renal agenesis unilateral/
25
VSD, PDA, TGA, TF,
25
SV
20
20
TF
PDA, VSD, TF, TAPVC
10
17/75
TF, ASD
VSD
VSD
bilateral
Abbreviations: Ao, aortic; AR, aortic regurgitation; AS, aortic stenosis; ASD, atrial septal defect;
ASD-1, primum atrial septal defect; ASD-2, secundum atrial septal defect; AV, aortic valve; AV
canal, atrioventricular canal defect; AVM, arteriovenous malformation; COARC, coarctation of the
aorta; DORV, double-outlet right ventricle; EFE, endocardial fibroelastosis; HLHS, hypoplastic left
heart syndrome; LCA, left coronary artery; LSVC, left superior vena cava; MR, mitral regurgitation;
PAs, pulmonary arteries; PDA, patent ductus arteriosus; PS, pulmonary valve stenosis; TAPVC,
totally anomalous pulmonary venous connection; TF, tetralogy of Fallot; TGA, transposition of great
arteries; TR, tricuspid regurgitation; truncus art., truncus arteriosus; VSD, ventricular septal defect.
References 9,12,16,17,18,19,20.
TABLE 33-4 INCIDENCE OF SEVERE ASSOCIATED NONCARDIAC ANOMALIES AMONG 2,220
INFANTS WITH HEART DISEASE
Diagnosis
Incidence (%)
Endocardial cushion defect

43
31

Malpositions
Tetralogy of Fallot
24
13
10
Coarctation of aorta
Pulmonary atresia with intact septum
D-Transposition of the great arteries
9
1
1
TABLE 33-5 POSSIBLE TERATOGENS FOR CONGENITAL HEART DISEASE
Vitamin deficiency
a
Folate deficiency
Environmental agents
a
High altitude, trichloroethylene, irradiation

Drugs
a
Ethanol, folic acid antagonists (trimethoprim, sulfasalazine, triamterene, timethadione,

a
phenytoin, primidone, phemobarbitol, carbamazepine, methotrexate) , valproic acid, lithium,

a
thalidomide, retinoic acid, antineoplastic agents (?), coumadin, amphetamine, cocaine

Metabolic factors
a
Maternal diabetes, maternal phenylketonuria, homocysteine

Immune factors
Maternal autoimmune disease with anti-Ra anti-LA antibodies
Infectious agents
a
Rubella, mumps (?), cytomegalovirus (?)

a
It is generally accepted that these prenatal factors increase the risk for congenital heart disease.
From refs. 9,12,13,18,23,24,25,26,27,28,29,30,31,114.
Certain cardiac anomalies are associated with prematurity or low birth

weight. Because closure of the ventricular septum may be delayed until
the first months of life, it is not surprising that there is a somewhat
greater incidence of ventricular septal defect among premature infants.
The increased incidence of patent ductus arteriosus in prematurely born
infants can be viewed as the result of birth long before the programmed
time for closure of the ductus. Hypoxemia of pulmonary origin also
promotes ductal patency.
As with gross anatomic cardiac anomalies, specific causes of cardiac
muscle diseases unknown until recently are being identified. Many of the
hypertrophic and dilated cardiomyopathies previously known as idiopathic
are now known to be caused by specific gene mutations and pathogenetic
mechanisms (9,16,32,33 and 34). Most patients with isolated
hypertrophic cardiomyopathy have newly acquired or autosomally
dominant inherited mutations in the genes encoding sarcomeric
contractile proteins, most commonly cardiac B-myosin heavy chain,
cardiac myosin-binding protein C or troponin T2 (32,33). Isolated dilated
cardiomyopathy has been associated to date with 19 genetic loci,
involving contractile, cytoskeletal and other proteins, and identification of
more is likely (9,16,33). Dilated and hypertrophic cardiomyopathies also
occur in association with a large number of more generalized
neuromuscular and metabolic disorders occurring from specific nuclear
and mitochondrial genetic mutations (see Cardiomyopathy and Tables 3316 and 33-17) (9,16,34).
Some specific arrhythmia syndromes are also caused by specific genetic
mutations. These include patients with ventricular tachycardia associated
with prolonged QT syndrome and arrhythmogenic right ventricular
dysplasia and rare forms of Wolf-Parkinson-White syndrome with
supraventricular tachycardia. Long QT syndrome results from mutations
in genes encoding specific cardiac potassium and sodium ion channels
that regulate repolarization (35); the ensuing prolongation of
repolarization results in ventricular tachycardia.
Fetal Cardiology
Extensive information about the circulatory physiology of the fetus and newborn
has accumulated. The works of Barcroft (37), Dawes (38), Lind and associates
(39), and Rudolph (40) should be consulted for details, but the central features
are discussed here. The circulation before birth consists of parallel circuits (Fig. 331). Blood in the aorta may follow several routes to a capillary bed in the fetus or
the placenta, back to the heart, passing through either ventricle, and out again to
the aorta. The stream of newly oxygenated blood from the placenta passes
through the umbilical vein, the ductus venosus, the inferior vena cava,
P.639
and the right atrium. Unlike the circulation after birth, the streams of oxygenated
and unoxygenated blood are not completely separated, although the more
oxygenated blood from the inferior vena cava is mostly diverted through the
foramen ovale into the left atrium. Consequently, blood from the left ventricle
entering the ascending aorta and coronary and carotid circulations is somewhat
higher in oxygen than that entering the descending aorta from the right ventricle
by way of the ductus arteriosus.
The volume pumped by the right ventricle is normally about 55% of the combined
output of both ventricles. Because both ventricles pump against the systemic
resistance, the level of pressure in the two ventricles is comparable. The
resistance to blood flow through the lungs is relatively great; only minimal flow
through the lungs occurs in utero, and almost all of the right ventricular output
into the pulmonary artery passes through the ductus arteriosus to the descending
aorta. The parallel arrangement of the ventricles allows fetal survival despite a
wide variety of cardiac lesions. With total obstruction of either ventricle, the other
ventricle assumes the entire cardiac output. Reversal of the pulmonary arterial
and aortic streams of blood, as occurs in transposition of the great arteries,
produces no deleterious effect on the fetus. Additionally, ductal or ascending
Figure 33-1 Fetal circulation is in parallel, and the amount of blood handled by
the left and right ventricles is 125 and 90 mL, respectively. Only 40 mL passes
through the aortic arch to the descending aorta, and only a small fraction passes
through the lungs. The numbers inside the diagram represent relative blood flow
(mL); the numbers in italics are pressure measurements. Modified from Rudolph
AM. Congenital diseases of the heart. Chicago: Year Book, 1974, with permission.
aortic flow may reverse in the presence of severe semilunar valvar stenosis or
atresia. Despite this remarkable ability to adapt, grow, and survive, the fetus is
affected by limitations in myocardial contractility. Prolonged, severe pressure or
volume loading of the heart or primary myocardial disease may result in
congestive heart failure, manifested by hydrops fetalis. The interplay between the
metabolic effects of congestion in the fetus and the possible compensatory role of
the placenta is not understood. Because lesions that may be expected to cause
gross intrauterine difficulty are tolerated surprisingly well, the postulate that the
placenta helps compensate for the metabolic abnormalities resulting from
congestive heart failure is tenable.
Circulatory Adjustments at Birth

Changes in the Source of Oxygenated Blood and
in the Ductus Venosusand Ductus Arteriosus
With the first breath, the resistance to pulmonary blood flow drops sharply. The
oxygen content of the left heart and systemic circulation rapidly reaches levels
well above that of the fetal circulation. Oxygen saturation in the ascending aorta
rises from about 65% in the fetus, to about 93% immediately after birth. The
ductus venosus functionally closes, establishing the portal circulation as an
independent loop between two capillary beds. With removal of the low-resistance
placenta, systemic resistance increases. The relative fall in the pulmonary
resistance and rise in the systemic resistance result in a transitory left-to-right
shunt through the ductus arteriosus. In half of term babies the ductus is
completely constricted by the end of the first day of life, and is normally becoming
anatomically obliterated at about 10 days of age. Even among cyanotic newborns
who are duct dependent, the ductus may inexorably close, often severing the
infant's only source of pulmonary or systemic blood flow. The mechanisms causing
closure of the ductus arteriosus are not completely understood but involve
decreased prostaglandins and increased blood oxygen. Prostaglandin levels in the
blood decrease at birth as a result of removal of their placental source of
production from the circulation and increase in perfusion of the lungs, in which
prostaglandins are metabolized.
Foramen Ovale
Functional closure of the foramen ovale occurs soon after birth, largely as a result
of increased left atrial volume and pressure secondary to the increased pulmonary
venous return, the ductal left-to-right shunt, and the developing differences in
diastolic pressure of the two ventricles. Anatomic closure normally is delayed for
months or years.
P.640
Among infants with cardiac defects, lesions with increased right atrial pressure
favor indefinite patency of the foramen ovale (e.g., pulmonary stenosis), but
abnormally increased left atrial pressure promotes early anatomic closure (e.g.,
ventricular septal defect). Before birth, the pulmonary arterioles are relatively
muscular and constricted.
Pulmonary Vasculature
With the first breath, total pulmonary resistance falls rapidly because of the
unkinking of the vessels with expansion of the lungs and because of the
vasodilatory effect of inspired oxygen. The muscular constriction relaxes, and
gradually, during the subsequent days and weeks, the muscular wall of the
pulmonary arterioles thins. During the first weeks of life, the muscular arterioles
retain a significant capacity for constriction. Pulmonary alveolar hypoxia normally
produces an increase in pulmonary artery pressure at all ages, but in the young
infant, the response is more profound and occurs more rapidly. Therefore,
pulmonary hypertension equal to or greater than systemic pressure occurs
commonly in neonates with severe respiratory disease.
Ventricular Work
Before birth, the two ventricles share in supplying systemic blood flow and
placental flow, and after birth, the two ventricles sequentially and independently
handle the entire cardiac output. At birth, the volume of blood to be pumped by
the right ventricle decreases to the level of the systemic blood flow; right
ventricular pressure falls as a result of the decrease in pulmonary resistance and
closure of the ductus arteriosus. Although right ventricular work decreases, left
ventricular work increases (Fig. 33-2). At birth, the left ventricle abruptly becomes
the sole supplier of systemic blood flow, and the volume that it pumps is
fractionally increased. The left-to-right shunt through the ductus arteriosus adds
further volume work, and the elevated systemic resistance must be overcome.
Although this is a stressful time for the left ventricle, the magnitude of these
suddenly acquired burdens is not so great that detectable left ventricular
difficulties are seen normally, but any impairment of myocardial function may be
magnified as a consequence. Myocardial disease as a cause of symptoms is more
common in the first days of life than at any other time during infancy; 25% of
infants with myocardial disease presented in the first week of life (4).
Myocardial Function
Important changes occur in the fetus and neonate in many aspects of myocardial
biochemistry and structure. These include myocyte size and number,
microvascular structure, myocyte utilization of lactate and fatty acids, and
antioxidant systems. Many structures and proteins involved in calcium handling

within the myocyte, such as t-tubules, sarcoplasmic reticulum, Na+-Ca2+
exchange, Ca2+-ATPase, and phospholamban, have important developmental
changes. These changes influence the effects on ventricular rhythm and function
of normal development, prematurity, ischemia, cardioplegia, and various inborn
errors of metabolism.
Figure 33-2 Mature circulation is in series, and the amount of blood carried by
the two ventricles is approximately the same as before birth. The lungs carry an
amount equivalent to the cardiac output, as does the ascending aorta. The
numbers inside the diagram represent relative blood flow (mL); the numbers in
italics are pressure measurements.
Fetal Echocardiography
High-resolution two-dimensional ultrasound evaluation of the fetal heart is a
useful and accurate technique in the diagnosis and management of the fetus at
risk for structural or functional cardiac abnormalities. Indications for prenatal
echocardiography may include maternal, fetal, and genetic considerations (Table
33-6). The optimal time for performing fetal echocardiography is 18 to 24 weeks
of gestation. At this age, the fetal heart is usually large enough for detailed
anatomic evaluation, and the images are unimpaired by dense rib or spine
calcification. There is also a relatively large volume of amniotic fluid that facilitates
imaging from a variety of angles. For accurate diagnoses the examiner must be
experienced in the technical aspects of fetal ultrasonography and knowledgeable

in the anatomic patterns and physiologic consequences of congenital heart defects
(46). Trans-vaginal ultrasound can provide diagnostic information as early as 10
to 12 weeks gestation, and may be useful when there is a high suspicion of major
cardiac anomalies. Three dimensional echocardiography and magnetic resonance
imaging has been applied to fetal cardiac evaluation and may improve diagnostic
accuracy, although difficulty gating to the fetal heart rate has limited their utility
to date.
TABLE 33-6 INDICATIONS FOR FETAL ECHOCARDIOGRAPHY
Suspected cardiac malformation on general ultrasound

Other malformations noted on general ultrasound
Oligo- or polyhydramnios
Fetal dysrhythmia
Suspected or known chromosomal abnormality
Family history of congenital heart disease
Family history of chromosomal abnormality
Maternal diabetes
Maternal collagen vascular disease
Rubella exposure
Evidence of hydrops fetalis
Intrauterine growth retardation
Maternal drug exposure, including:
Lithium
Hormones
Anticonvulsants
Chemotherapy
Alcohol
P.641
Cardiac Anatomy
Virtually all major cardiac malformations can be detected prenatally using highresolution two-dimensional, real-time sector scanning by an experienced
examiner. The details of systemic and pulmonary venous connections, arterial
alignment, chamber size and orientation, and valve position and function can be
determined (Fig. 33-3) and abnormal structures demonstrated (Figs. 33-4 and 335).
Cardiac Physiology
Color Doppler provides a quick and sensitive means of evaluating the function of
atrioventricular and semilunar valves, the direction of flow in fetal vessels, and the
presence of normal and abnormal connections (Fig. 33-5). If abnormal flow is
detected, it can be evaluated further using the quantitative capabilities of pulsed
or continuous-wave Doppler.
Figure 33-3 Echocardiogram of the normal fetal heart in a four-chamber view

demonstrating the position of the heart and the cardiac chambers in a crosssection of the chest. A, anterior; L, left; LA, left atrium; LV, left ventricle; P,
posterior; R, right; RA, right atrium; RV, right ventricle; arrow denotes the spine.
Figure 33-4 Echocardiographic cross-sectional views of the fetal chest in an

infant with Ebstein anomaly of the tricuspid valve. The right atrium (RA) is
markedly dilated and fills much of the thorax. The left atrium (LA) and left
ventricle (LV) are of normal size but are dwarfed by the right-sided structures.
The severely regurgitant tricuspid valve has apical displacement of the septal
leaflet into the right ventricle (RV).
Cardiac Function
A qualitative assessment of cardiac function is obtained by visual inspection of
ventricular motion during real-time
P.642
sector scanning. When more quantitative information is desired, M-mode
recording can provide precise dimensions and an accurate measure of ventricular
shortening (Fig. 33-6). Doppler derived indices of cardiac function may also
provide insight into fetal cardiac performance and three-dimensional imaging and
fetal MRI may ultimately prove useful as well. Severe ventricular dysfunction may
manifest as generalized hydrops fetalis, which is readily recognized by ultrasound
as pleural and peritoneal fluid accumulation and cutaneous edema.
Figure 33-5 Echocardiographic cross-sectional view of a fetus with multiple

intramyocardial rhabdomyomas (arrows). The infant was subsequently diagnosed
with tuberous sclerosis. A, anterior; L, left; LA, left atrium; LV, left ventricle; P,
posterior; R, right; RA, right atrium; RV, right ventricle.
Figure 33-6 Echocardiogram of a fetus with a premature atrial beat. The upper
left image is a two-dimensional view with a cursor (line) through the fetal right
atrium (RA) and aortic valve (AOV), demonstrating the axis of the simultaneous
M-mode echocardiogram seen in the lower panel. The M-mode tracing depicts the
motion of the fetal right atrial wall and aortic valve in the cursor line over a time
frame of 3 seconds. A series of normal atrial wall contractions are interrupted by
a premature contraction (large arrow) followed by opening of the aortic valve
(small arrow), demonstrating a premature atrial beat conducted to the ventricle.
LV, left ventricle; RV, right ventricle.
Arrhythmias
Tachyarrhythmias, bradyarrhythmias, and irregular cardiac rhythms are common
reasons for referral for evaluation. Structural and functional abnormalities should
be excluded, as described above. The mechanism of the rhythm disturbance can
usually be elucidated by determining the timing of atrial and ventricular
contraction using an M-mode (Fig. 33-6), two-dimensional and Doppler
echocardiography, which simultaneously display atrial and ventricular wall and
valve motions and flows. By this means, the timing and sequence of atrial and
ventricular activation can be deduced. The most common rhythm disturbance is
isolated premature atrial contractions in a structurally normal heart or in
association with an atrial septal aneurysm. Sustained tachyarrhythmia usually
represents a reentrant or ectopic atrial tachycardia. These infants must be
monitored closely for the development of congestive heart failure and hydrops
fetalis, which would be an indication for induced delivery of the mature fetus or
maternal antidysrhythmic therapy in the immature fetus. Sustained
bradyarrhythmias may be secondary to heart block, nonconducted premature
atrial contractions, or noncardiac sources of fetal distress. The mechanism can be
inferred as described above and appropriate therapy initiated if indicated (see
Arrhythmias).
Fetal Cardiac Intervention

Improvements in prenatal diagnostic techniques have led to renewed
interest in fetal cardiac intervention. The restoration of more normal flow
and pressure relationships in the developing heart may encourage normal
or near normal growth and function, improving the postnatal outcomes
and surgical and medical options. Exteriorization of the fetus and surgical
intervention has been performed, but the inevitable development of
premature labor remains a major obstacle to the development of fetal
cardiac surgical techniques. Improvements in fetal imaging techniques
are compatible with interventional catheterization methodology. The
technical feasibility of aortic valve dilation in the fetus using new, lowprofile coronary balloons and wires to dilate aortic valves in fetuses likely
to acquire hypoplastic left heart syndrome without intervention has
recently been demonstrated. Additionally, there has been some success in
creating atrial communications in fetuses with premature closure of the
foramen ovale in association with left sided obstruction. Attempts have
also been made to re-establish communication between the right ventricle
and pulmonary arteries in fetuses with valvar pulmonary atresia. Issues
of patient selection, timing and outcome are only beginning to be
addressed, and the clinical utility of prenatal interventions is yet to be
determined.
Prematurity
The circulatory adjustments and myocardial biochemical changes at birth and in
the neonatal period are modified in direct relation to the degree of prematurity.
The muscular coat of the pulmonary arterioles develops late in gestation; the
more premature the infant, the less muscular are the pulmonary arterioles at
birth. The most notable consequence of this is that the difference between
systemic and pulmonary resistance after birth is greater among premature than
among normal infants. Shunting through a ductus arteriosus is often audible.
Developmental biological factors in the ductus arteriosus and hypoxia, so common
among premature infants, may be factors that contribute to the delay in closure of
the ductus in premature infants. The
P.643
propensity of the ductus to close at around 41 weeks after conception is clinically
recognized. Developmental changes in myocardial structure and biochemistry may
influence the function of the left ventricle in response to stress such as volume
overload associated with the left-to-right shunt through a patent ductus arteriosus.
Recognition of Clinical Features

Only a few infants are born in hospitals equipped for all eventualities. Infants with serious heart anomalies
require transportation to a specialized pediatric cardiac center for detailed diagnostic assessment with
echocardiography, and treatment that may include cardiac catheterization and/or surgery. Timely clinical
recognition of the likely presence of a specific cardiac anomaly, that without intervention will result in serious
deterioration of the baby's condition (e.g., critical coarctation of the aorta, pulmonary valve atresia), is
necessary for initiation of medical therapy to prevent and/or reverse clinical deterioration (e.g., administration of
prostaglandin, inotropic agents, oxygen, and ventilation) and thereby provide the time and conditions necessary
to transfer, evaluate, and treat the baby.
Initial evaluation includes assessment for cyanosis (See Color Plate) and of the infant's well-being, perfusion,
pulses and blood pressure in the extremities, respiratory work and rate, precordial activity, second heart sound
splitting, and murmur intensity, quality, pitch, and timing. Chest radiograph and electrocardiogram (ECG)
remain cost- and time-efficient tests that aid in the initial evaluation of suspected congenital heart disease. Any
Figure 33-7 The differential diagnosis of cardiac exam findings in acyanotic neonates. Anomalies in larger print
are more common. +/-, sometimes; up arrow, increased; down arrow, decreased; AP window,
aorticopulmonary window; AS, aortic stenosis; ASD, atrial septal defect; CAVC, complete atrioventricular canal
defect; CoAo, coarctation of the aorta; DCM, dilated cardiomyopathy; HCM, hypertrophic cardiomopathy;
Interrup. Ao, interrupted aortic arch; MS, mitral stenosis; PDA, patent ductus arteriosus; PS, pulmonary
stenosis; VSD, ventricular septal defect.
one of these alone is rarely diagnostic. A number of lesions result in cyanosis; quite a number of lesions are also
associated with loud murmurs; others are associated with little or no murmur; some cause shock (see Fig. 33-7
and 33-8). Others have chest
P.644
radiographs with increased pulmonary arterial or venous markings, others have diminished pulmonary vascular
markings. Most have an undistinguished electrocardiogram at birth, although some have left axis deviation on
electrocardiogram (see Fig. 33-9 and 33-10). Most cardiac anomalies vary in their characteristics at
presentation. Furthermore, often it is not possible to determine with certainty if the second heart sound is split
or not, or if the pulmonary vascular markings on chest radiograph are normal vs. increased or normal versus
decreased. Clinical analysis requires weighting of the categories of evidence regarding its certainty and other
possibilities. A classical diagnostic approach based on sequential analysis of data categories is limited by these
types of weakness in the clinical information and is no stronger than the weakest link in the chain of information.
However, interweaving of the findings provides a matrix of diagnostic information that remains intact even when
one category of findings is weak. Overlapping possible anomalies suggested from history, physical examination,
chest radiograph and electrocardiogram, as if with a series of Venn diagrams (see Figs. 33-7, 33-8, 33-9 and 3310), provides information that allows a careful observer to quickly determine which anomaly, or which two or
three possible anomalies, is likely present. Comparing the anomalies consistent with the clinical presentation
with the anomalies consistent with the murmur findings, other physical exam findings, chest radiograph findings,
and electrocardiographic findings, usually focuses the list of possible anomalies on one or two primary choices
(see Fig. 33-11). This may provide an important advantage in the timely and efficient management of
potentially life threatening anomalies. For example, the combination of cyanosis, soft or no murmur, single S2,
chest radiograph with decreased pulmonary vascular markings and normal heart size, and electrocardiogram R
axis of 50 suggests pulmonary atresia, which is an anomaly in which life depends on maintaining ductal
patency (see Fig. 33-8 and 33-10). Two-dimensional echocardiogram should be obtained promptly if significant
cardiac disease is suspected. This technique when done by personnel trained for evaluation
P.645
of congenital cardiac anomalies in neonates accurately demonstrates the anatomy, occasionally uncovering a
potentially lethal lesion before symptoms. Appropriate initial management (e.g., infusion of PGE1 in a cyanotic
infant suspected to have pulmonary atresia) need not await availability of echocardiography (see Fig. 33-12 and
Management Procedures for Severe Cardiac Disease).
Age of Presentation
Most children with critical congenital cardiac anomaly develop symptoms within the first weeks of life (8). The
age when infants develop cardiac symptoms is diagnostically useful. For instance, although ventricular septal
defect is far more common (Table 33-2), transposition of the great arteries, coarctation of the aorta, and the
hypoplastic left heart syndrome are the most common life-threatening anomalies presenting in the first days of
life (see Table 33-7). Isolated ventricular septal defect is not associated with cyanosis, the associated murmur
Figure 33-8 The differential diagnosis of cardiac exam findings in cyanotic neonates. +/-, sometimes; up
arrow, increased; down arrow, decreased; AS, aortic stenosis; CoAo, coarctation of aorta; collat., systemic to
pulmonary artery collateral vessels; Ebstein's, Ebstein anomaly of the tricuspid valve; HLHS, hypoplastic left
heart syndrome; Int Ao, interrupted aorta; PA, pulmonary atresia; PA-IVS, pulmonary atresia with intact
interventricular septum; PPH, persistent pulmonary hypertension syndrome; PS, pulmonary stenosis; SV, single
ventricle; TAPVC, totally anomalous pulmonary venous connection; TGA, transposition of great arteries; ToF,
tetralogy of Fallot; tricuspid atr., tricuspid atresia.
Figure 33-9 The differential diagnosis of chest radiographic and electrocardiographic findings in acyanotic
neonates. Abbreviations, see Fig. 33-7; CXR, chest radiograph; LAD; left axis deviation; LV, left ventricle;
Pulm. vasc. markings, pulmonary vascular markings; RV, right ventricle.
generally develops after several days or more, and respiratory symptoms usually do not develop until after the
first week of life. Among those whose problem is cyanosis, transposition of the great arteries is the leading
cause through the third week of life; after that time, tetralogy of Fallot becomes the dominant cause of cyanosis.
Among neonatal cardiac patients admitted because of respiratory symptoms, the hypoplastic left heart
syndrome is the leading cause in the first week, complex coarctation leads in the second week, and thereafter,
ventricular septal defect becomes the main cause (see Table 33-6).
Respiratory Symptoms
Persistent tachypnea may be the first clue to heart disease or lung disease. Cardiac abnormalities with excessive
pulmonary arterial flow or pulmonary venous hypertension cause pulmonary vascular engorgement, pulmonary
edema, and decreased lung compliance that often result in
P.646
increased respiratory effort and rate. Cardiac anomalies with decreased pulmonary blood flow often have intense
cyanosis that elicits a reflex peaceful tachypnea without respiratory distress. Experienced parents often
observe that the affected baby had always breathed too fast. Persistent respiratory rates of 60 per minute or
greater, often with minimally labored but persistently increased depth of respiration, commonly precede other
findings and may presage clinical deterioration. A chest radiograph may differentiate cardiac from pulmonary
disease.
Figure 33-10 The differential diagnosis of chest radiographic and electrocardiographic findings in cyanotic
neonates. Abbreviations, see Fig. 33-8 and 33-9.
Systemic Perfusion and Pressure

Decreased systemic cardiac output is an ominous sign that requires rapid assessment and rapid appropriate
management for the infant to survive. There are many potential noncardiac causes, the most common being
sepsis, and important cardiac causes (see Fig. 33-7 and 33-8). Signs of diminished systemic perfusion include
poorly perfused, cool, and/or mottled skin, listlessness, diminished peripheral pulse intensity, diminished systolic
and pulse pressure, decreased urine output, and metabolic acidosis. Blood pressure should be measured in all
four extremities in an infant who appears severely ill with these signs, particularly with coexistence of a murmur
or cyanosis. Oscillometric devices can quickly and noninvasively measure blood pressure, although their
correlation with centrally measured blood pressure declines in the presence of very low perfusion and pressure.
It is very important to establish if the pulse amplitude and systolic blood pressure are diminished only in the
postductal arterial distribution or throughout the body, that is, if the right arm systolic blood pressure is similar
to or higher than that in the other extremities. Significant differential in the systolic blood pressure between the
arms (usually right) and the legs (and umbilical artery) is diagnostic of an aortic obstruction. The additional
presence of a murmur, gallop, hepatomegaly, or cyanosis further strongly suggests that a cardiac anomaly is
causative of poor perfusion.
Figure 33-11 A process for diagnosing cardiac anomalies from findings on cardiac exam, chest radiograph, and
electrocardiogram. ASAP, as soon as possible; c/w, consistent with; FiO2, fractional percentage of inspired
oxygen; PDA, patent ductus arteriosus.
P.647
Murmur
Hearing a murmur is the most common means of recognizing the presence of heart disease in an infant.
Determining the diagnosis requires ascertaining the characteristics of the murmur. These include the history of
the baby's age when the murmur was first audible and examination findings of murmur timing in systole versus
diastole, loudness, pitch and the association of a thrill. The murmurs of valvar regurgitation and stenosis are
audible immediately after birth, and the murmurs of septal defects are usually delayed days to weeks, or as long
as several months in the case of atrial septal defects. Diastolic murmurs are rare but indicative of cardiac
pathology. A prominent continuous murmur in a cyanotic neonate, particularly in the back or axilla, is rare but is
very characteristic of tetralogy of Fallot with
P.648
pulmonary atresia and systemic-to-pulmonary-artery collateral vessels (the latter being the cause of the
murmur). The loudness of a murmur, in combination with other findings, may suggest the likelihood of various
anomalies but is often not proportional to the severity of the lesion. The absence of a murmur does not preclude
serious heart disease. To the contrary, many life-threatening cardiac anomalies may be associated with little or
no murmur. In a neonate with cyanosis and/or shock and suspected cardiac anomaly, the presence of little or no
murmur provides a diagnostic clue (see Fig. 33-7 and 33-8). The pitch of a murmur is associated with the
pressure gradient across the abnormality causing the murmur. Tiny ventricular septal defects develop a
characteristic fairly high-pitched murmur when the right ventricular pressure decreases to much less than the
left ventricular pressure. Severe pulmonary or aortic stenosis can sometimes be distinguished from mild stenosis
by a high-pitched harsh loud murmur and an associated thrill.
Heart Sounds
Auscultation of the heart sound splitting is the most difficult part of the cardiac examination in neonates because
of the relatively rapid heart and respiratory rates in neonates. However, when abnormalities of the first and
second heart sounds are strongly suspected or excluded, it provides important information. Detection of splitting
of the heart
P.649
sounds requires practice and a minute or so of focused attention on just that sound, using a quality stethoscope,
in a quieted baby. The absence of splitting may result from a heart rate too fast to discern splitting or a truly
single-component sound. A split first heart sound in a neonate suggests a click. The second heart sound
emanates from closure of the aortic and pulmonary valves. Determining that the second heart sound is split (i.
e., has two components) suggests that both aortic and pulmonary valves are not severely abnormal; that is, it is
against the presence of aortic or pulmonary valve atresia or severe stenosis. However, other serious anomalies
with two semilunar valves may still be present, for example, simple transposition of the great arteries. A second
heart sound that always appears single, particularly at heart rates not greater than 120 per minute, may be
caused by relatively early pulmonary valve closure associated with elevation of pulmonary artery pressures
comparable to aortic pressures, but it suggests that the pulmonary or aortic valve may be abnormal (as in
pulmonary atresia or critical stenosis, hypoplastic left heart syndrome, truncus arteriosus). Although difficult to
detect, constant splitting of the second heart sound, as opposed to the usual intermittent splitting, suggests a
atrial septal defect.
Figure 33-12 An approach to the diagnosis and management of cyanotic infants. #, cyanosis; see text for
additional details of assessment of cyanosis; +/-, possibly; *, see text concerning management of specific
anomalies; ASAP, as soom as possible; c/w, consistent with; FiO2, fraction inspired oxygen; Rx, treatment.
Cyanosis
Much more threatening than a murmur is the presence of cyanosis. Cyanosis without pulmonary disease is
almost invariably the result of a serious cardiac abnormality. Cyanosis may result from poor mixing of separate
parallel circulations (e.g., transposition of the great arteries, other anomalies with transposition physiology such
as Taussig- Bing-type double-outlet right ventricle); restricted pulmonary blood flow and right-to-left shunting of
un-oxygenated systemic venous blood to the systemic arterial circulation (e.g., tetralogy of Fallot, critical
pulmonary stenosis, tricuspid atresia); or right-to-left shunting from intracardiac mixing with normal or
increased pulmonary blood flow (e.g., total anomalous pulmonary venous connection without obstruction,
truncus arteriosus, single ventricle without pulmonary stenosis, hypoplastic left heart syndrome). Especially in
the first week of life, cyanosis may be the sole evidence of an important cardiac lesion. One-third of infants with
potentially lethal congenital heart disease have cyanosis as their major symptom; another one-third have
cyanosis associated with respiratory symptoms. Prompt cardiac evaluation of all cyanotic babies is mandatory
because prompt infusion of prostaglandin endothelin (E1) to open the ductus arteriosus or catheter intervention
to create an atrial septal defect may be necessary for survival, and most of the responsible lesions are amenable
to surgery.
The clinical recognition of cyanosis is dependent on the amount of oxygen desaturation of arterial hemoglobin
and therefore is influenced by the total blood hemoglobin concentration. An anemic infant may have severe
arterial oxygen unsaturation without obvious cyanosis, and infants with polycythemia may appear cyanotic with
near normal arterial oxygen levels. Hypothermic infants may seem blue; babies viewed in fluorescent lighting or
in blue surroundings may make the estimation of cyanosis more difficult. Persistent cyanosis secondary to
hypoglycemia or methemoglobinemia is rare. Cyanosis is particularly evident in the lips. Peri-oral or nail-bed
cyanosis without lip cyanosis is usually not caused by cyanotic heart disease. When cyanosis is suspected,
indirect assessment of arterial oxygen saturation by the transcutaneous method can provide a rapid noninvasive
check.
Acute Lung Disease and Cardiac Disease

Rapid determination of the diagnosis and initiation of appropriate management is most pressing when the infant
is dyspneic and cyanotic. A chest radiograph may suggest lung disease, particularly if the findings are
asymmetric. In the presence of diffuse symmetric changes possibly compatible with pulmonary edema or
increased vascular markings, caution is necessary, particularly in the full-term neonate. The differential
diagnosis between primary lung disease and heart disease causing pulmonary edema (e.g., total anomalous
pulmonary venous connection with obstruction) can be difficult. Persistent pulmonary arterial hypertension with
right-to-left shunting may coexist with lung disease and cause severe cyanosis. Although carbon dioxide
retention is usually prominent among babies with primary lung disease, some severely cyanotic infants with
cardiac anomalies can have marked hypercarbia. It can also
P.650
be difficult to differentiate cardiac anomalies with diminished pulmonary blood flow and little murmur (e.g.,
pulmonary valve atresia) from persistent pulmonary arterial hypertension without other, radiographically
apparent, lung parenchymal disease. Although the absence of hypercarbia suggests cardiac disease, some
severely cyanotic infants with pulmonary vascular disease have normal partial pressure of CO2 (PCO2) values.
The response of the arterial partial pressure of oxygen (PO2) to administration of 100% oxygen, with the
exceptions noted below, can differentiate cyanotic heart disease from lung disease. The infant who responds to
breathing pure oxygen with a rise in arterial PO2 to 220 mm Hg or more does not have cyanotic heart disease,
and the infant who does not raise his preductal arterial PO2 above 100 mm Hg is likely to have heart disease.
Transcutaneous estimation of arterial oxygen saturation is not an accurate alternative because any arterial PO2
greater than 70 mm Hg will result in arterial oxygen saturation greater than 95%. Because the hyperoxia test is
inconclusive when the arterial PO2 is between 100 and 220 mm Hg, these babies should be approached as
possibly having cyanotic heart disease (see Fig. 33-12). The arterial PO2 while the baby is breathing 100%
oxygen may initially be most rapidly measured from an umbilical artery catheter positioned in the descending
aorta. However, a low arterial PO2 measured in the descending aorta may be from right-to-left shunting through
a ductus arteriosus from coexisting persistent pulmonary hypertension. Comparison of the PO2 measured in
blood from the right radial artery with the PO2 measured in blood from the umbilical arterial catheter may help
to differentiate persistent pulmonary hypertension from cyanotic cardiac anomaly. The former may have a high
PO2 in the right radial artery. Simultaneous mechanical hyperventilation and administration of oxygen may
decrease pulmonary resistance and increase pulmonary flow, increasing the PO2 to greater than 220 mm Hg in
the descending aortic and/or right radial arterial blood, allowing differentiation of lung disease or persistent
pulmonary hypertension from cyanotic cardiac anomaly. Difficulties arise when pulmonary and cardiac pathology
coexist. For example, in the baby with both lung and heart disease or those with coexistence of persistent
pulmonary hypertension and predominant right-to-left shunting through the foramen ovale (see Chapters 29
and 32) or the baby with heart disease causing pulmonary venous hypertension and pulmonary edema, results
may be confusing. Arterial PO2 may not significantly increase in response to 100% oxygen with severe
persistent pulmonary hypertension and predominant right-to-left shunting through the foramen ovale. If doubt
persists, the physician can make the diagnosis with two-dimensional echocardiography (see Fig. 33-12).
Noncardiac Anomalies
It is useful to know the relative frequency of the cardiac diagnostic possibilities when there are associated
noncardiac anomalies (see Tables 33-3 and 33-4) or prematurity. Among premature infants, patent ductus
arteriosus, coarctation of the aorta, and ventricular septal defect occur more often. Chromosomal abnormalities
and congenital syndromes are also associated with lower birth weight cardiac malformations (e.g., Down
syndrome).
TABLE 33-7 TOP FIVE DIAGNOSES PRESENTING AT DIFFERENT AGES

Diagnosis
Age on admission: 0-6 days (n = 537)

D-Transposition of great arteries
Hypoplastic left ventricle
Tetralogy of Fallot
Others
Percentage of Patients
19
14
8
7
3
49

Hypoplastic left ventricle
Tetralogy of Fallot
Others
16
14
8
7
7
48

Tetralogy of Fallot
16
12
7
7
5
Others
53
Diagnostic Tools
Chest Radiography
Chest radiography is rarely diagnostic of specific cardiac lesions, but it is a relatively quick and relatively
inexpensive method to identify lung disease and screen for suspected cardiac anomaly in symptomatic infants.
Chest radiographs often appear normal or near normal in the first day or days of life with many cardiac
anomalies, and may not be cost effective in asymptomatic infants with an isolated murmur (i.e., no cyanosis
(See Color Plate) or congestive signs or symptoms). However, in cyanotic or symptomatic infants, the presence
or absence of cardiomegaly, increased pulmonary arterial or venous markings, diminished pulmonary arterial
markings, or right aortic arch provides important information regarding the presence of a cardiac anomaly (see
Fig. 33-9 and 33-10). In combination with physical exam and electrocardiographic findings, chest radiographic
findings may provide important information concerning the possible presence of specific cardiac anomalies that
may aid in early management before an echocardiogram can be obtained. The heart size should be differentiated
from the thymic shadow. Cardiomegaly is indicated by a cardiothoracic ratio greater then 0.6 in an anteriorposterior projection in the presence of an adequate inspiration. The aortic arch position can be assessed, even in
the presence of a large overlying thymus, by deviation of the trachea to the opposite side. Associated noncardiac
anomalies that provide clues to the cardiac diagnosis may be discovered by radiographic findings, for example,
heterotaxy (asplenia syndrome, malrotation), absence of the thymus gland (DiGeorge syndrome), vertebral
anomalies (VACTERL association), and abnormal sternal ossification (Down syndrome).
Electrocardiography
Electrocardiography can be useful in evaluation for cardiac anomaly and arrhythmogenic disorders and in
particular for diagnosis and management of arrhythmia (see Arrhythmias). Electrocardiographic interpretation in
neonates has several caveats. However, when placed within the context of other physical exam and radiographic
findings, electrocardiographic findings can provide a timely advantage in diagnosis and management of
suspected cardiac anomaly (see Fig. 33-9 and 33-10). Electrocardiography is also useful for timely recognition of
life-threatening arrhythmogenic disorders, particularly within the context of other findings; for example family
history and borderline prolonged QT interval (QTc).
TABLE 33-8 NORMAL MATURATIONAL ECG CHANGES
Age
0-1 day
1-3 days
3-7 days
7-30 days
Heart Rate*
(beats per
min)
R Axis*
(+ degrees)
93-154 (123) 59-192 (135)

91-159 (123) 64-197 (134)
90-166 (129) 77-187 (132)

1-7-182
65-160 (110)
(149)
1-3 months 121-179
31-114 (75)
(150)
* 2-98th percentile (mean)
from ref 43,45
R Amplitude V1*
(mm)
R Amplitude V6*
(mm)
T Amplitude V1*
(mm)
5-26
5-27
0-11
0-12
-30-+40
-41-+41
3-24
3-21.5
0.5-12
2.5-16
-45-+25
-1052
3-18.5
5-21
-1262
P.651
Interpretation of neonatal electrocardiograms requires knowledge of maturational changes in heart rate, and
electrocardiographic intervals, axes, voltages and repolarization that occur normally during the first days and
weeks of life (see Table 33-8 and Fig. 33-13) (43,44 and 45). Within the first days of life, there are significant
changes in repolarization, including T axis and rate-corrected QTc, and changes in R and S wave amplitudes,
that influence interpretation. Compared with a 1 year old, 1 day old babies normally have relatively fast heart
rates (93-154, average 123/minute), relatively rightward R axis (60-195, average 123 degrees) and relative
right ventricular hypertrophy (R in V1 5-16 mm) (44). During the first 4 days of life there is variability of the
QTc, an evolution of changes in T wave polarity and
P.652
voltage (e.g., in V1 from upright to inverted), and nonspecific ST segment changes are common. QTc is 440
milliseconds or less in 97.5% of newborns, but can be much longer with electrolyte abnormality (e.g.,
hypocalcemia, hypokalemia), drug effect, brain injury and genetic prolonged QT syndromes. Premature neonates
tend to have higher heart rates (141-160/min), slightly more leftward R axis 107-135 degrees), slightly less R
amplitude in V3R (median 3 mm), slightly greater R amplitude in V6 (median 6-7 mm) and slightly longer QTc
interval (411-412 ms) (46).
Figure 33-13 Electrocardiograms from healthy day old (top) and 5 day old infants (lower) demonstrate
normal neonatal repolarization changes. V1 T wave morphology chanbges from upright to inverted.
Ventricular hypertrophy in the electrocardiogram at birth is a consequence of the hemodynamic abnormalities

imposed on the fetal circulation. The hemodynamic changes and the ventricular hypertrophy associated with
various cardiac anomalies are often much different in the fetal circulation than postnatal. For example, neonates
with coarctation, and other obstructive anomalies with less than normal blood flow volume through the fetal left
ventricle and more than normal flow through the fetal right ventricle, often have right ventricular hypertrophy
from increased fetal right ventricular workload. Furthermore, anomalies associated with systemic-level right
ventricular hypertension later in infancy often have neonatal electrocardiographic right ventricular forces that
are difficult to differentiate unambiguously from normal in the neonate. Therefore, the differential diagnosis of
ventricular hypertrophy in the neonate is different from later in infancy (see Fig. 33-9 and 33-10). In term
neonates electrocardiographic findings of right ventricular hypertrophy include R amplitude in V1 above the 98th
percentile (44) (>26 mm in 1st week, >21 mm in 2nd to 4th week), presence of Q wave in V1, or persistence of
upright T waves in V1 beyond age 1 week. Findings of left ventricular hypertrophy include elevated R amplitude
in V6 (44) (>12 mm in 1st week, >16 mm in 2nd to 4th week) and deep Q wave in V6 (>4 mm), often
accompanied by T wave flattening or inversion in V6.
Although most anomalies do not have R axis deviation at birth, when axis deviation is present the
electrocardiogram can be very helpful in diagnosis (see Fig. 33-9 and 33-10). Left axis deviation in acyanotic
newborns is most often associated with endocardial cushion anomaly (e.g., primum atrial septal defect,
complete atrioventricular canal defect), although in cyanotic newborns left axis deviation occurs with tricuspid
atresia and other cyanotic anomalies in combination with an endocardial cushion anomaly (e.g., tetralogy or
double outlet right ventricle with complete atrioventricular canal defect).
Echocardiography
Examination of the heart by two-dimensional echocardiography with color Doppler ultrasound allows excellent
analysis of the intracardiac anatomy in small infants. The equipment is portable and can be readily brought to
the bedside of critically ill neonates thus avoiding transport elsewhere, interruption of ongoing care, or
anesthesia. The examination in neonates requires no or little sedation. Neonates are particularly good
candidates for echocardiographic imaging because they are less active and have excellent echocardiographic
imaging windows. Detailed segmental examination from sub-xiphoid, parasternal, apical, supra-sternal notch,
and additional modified views as necessary delineates almost all relevant cardiac anatomy and anomalies in
most neonates. The situs, ventricular relationship, great artery relationships, systemic and pulmonary venous
cardiac connections, atrial and ventricular septum, valve structure, great artery anatomy and coronary origins
can be accurately determined.
Doppler echocardiography demonstrates the direction and velocity of blood flow within the heart and vessels.
Color Doppler visualizes valve regurgitation and blood flow in valvar and sub-valvar stenoses, patent ductus
arteriosus, septal defects, abnormal coronary arteries, systemicvenous anomalies, and arterio-venous
malformation. Pulsed and continuous-wave Doppler techniques enable estimation of physiologic measurements
such as the pressure gradient across stenotic valves, septal defects, and patent ductus arteriosus (see Fig. 3314). When physiologic or pathologic tricuspid regurgitation is present, right ventricular peak systolic pressure
may be estimated by Doppler measurement of the magnitude of the pressure gradient between the right
ventricle and right atrium and the addition of the right atrial V-wave pressure, whether assumed or directly
measured through an umbilical vein catheter (usually 3 to 10 mm Hg) (see Fig. 33-15). Right ventricular systolic
pressure relative to left ventricular pressure can also be qualitatively assessed by the curvature of the
interventricular septum. Contrast echocardiography with injection of agitated saline or albumin into intravenous
or umbilical artery catheters can sometimes serve as a useful adjunct to color Doppler in detection of shunts.
The ventricular systolic performance, size, and wall thickness can be assessed. The shortening fraction of the left
ventricular internal short-axis dimension is the most commonly used measurement to assess left ventricular
systolic function in children. In the sick neonate right ventricular systolic pressure often is close to left
ventricular systolic pressure, resulting in flattening of the ventricular septal curvature; such that shortening
fraction may not be indicative of global systolic performance. When regional wall motion abnormalities are
present, ventricular systolic performance is assessed by estimation or measurement of the relative change in
ventricular volume with contraction, the ejection fraction. The shortening and ejection fractions measure left
ventricular performance, which is a function of contractility, afterload, preload, and heart rate. Contractility can
be independently assessed by measuring the relationship of end-systolic wall stress velocity to fiber shortening
using directed M-mode echocardiography, indirect central pulse tracing, and phonocardiography. This technique
is also impaired when right ventricular hypertension results in ventricular septal systole flattening.
Echocardiography has limitations. Because complete examination of cardiac anatomy in neonates is labor
intensive and requires expensive additional technology, the cost is often nearly that of a computerized
tomography
P.653
or magnetic resonance scans. The evaluation has often been unsatisfactory when performed in which the use of
echocardiography to recognize heart disease in neonates is infrequent and echocardiographic transducers with
frequencies appropriate for infants are not available. Training and performance standards for echocardiographic
examination of congenital heart disease in fetuses and children have been disseminated (46). Cases of delayed
transfer of babies because of erroneous diagnoses of inoperable congenital heart disease and erroneous
impression that there is no significant lesion have been encountered. With limited experience in the diagnosis of
congenital heart disease in neonates, it may be best to transport the infant to the nearest center for
echocardiographic examination. If personnel adequately trained in performing a complete study for congenital
heart disease are available, it may be possible to send or transmit a tape of the examination for a second
opinion.
Figure 33-14 Echocardiographic parasagittal parasternal view of a patent ductus arteriosus. Doppler analysis
demonstrates flow away from the transducer within the pulmonary artery and aortic isthmus and, in white, a
jet of flow toward the transducer through the patent ductus arteriosus into the pulmonary artery. B:
Quantification of the velocity of the flow jet through the ductus arteriosus with a continuous-wave Doppler
technique and application of the Bernoulli principle allows the aortic-to-pulmonary-artery systolic pressure
gradient to be measured. The pressure gradient by this technique is 4 (maximum instantaneous velocity)2.
The pulmonary artery peak systolic pressure can be estimated by the difference in the arterial systolic pressure
and the pressure gradient across the ductus arteriosus. AO, aorta; DAO, descending aorta; LA, left atrium;
MPA, main pulmonary artery; PDA, patent ductus arteriosus.
Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) can provide accurate assessment of intracardiac anatomy and function, and
perhaps more importantly, can provide detailed images of intrathoracic structures, such as peripheral pulmonary
arteries, systemic-to-pulmonary collateral vessels, and right ventricular function, that often are not adequately
provided by echocardiography (see Coarctation, Fig. 33-35). Use of echocardiography and magnetic resonance
imaging, in conjunction with the history and physical examination, enables precise diagnosis without resorting to
diagnostic cardiac catheterization in most neonates.
Diagnostic Cardiac Catheterization and

Angiography
Anatomy
Catheterization is rarely used to learn the basic anatomy of the heart. The diagnostic information
necessary for most cardiac surgical procedures in neonates is now obtained noninvasively by
echocardiography. Diagnostic cardiac catheterization is used to provide specific data unavailable
through echocardiography that are useful in planning management (47). What is the anatomy of
the pulmonary arteries and systemic-to-pulmonary collaterals in the patient with tetralogy of Fallot
and pulmonary atresia? (see Tetralogy, Fig. 33-22) Is catheter closure of the collaterals possible?
What is the anatomy of the coronary arteries in the patient with pulmonary atresia and intact
ventricular septum? (see Pulmonary Atresia, Fig. 33-26) What is the anatomy of the coronary
arteries with tetralogy of Fallot or transposition in which abnormality is suspected and/or
echocardiographic imaging is nondiagnostic? In selected patients with cardiomyopathy, light and
electron microscopic analysis of ultrastructural anatomy and biochemical analysis of myocardium
obtained by biopsy may provide a diagnosis. In many situations diagnostic catheterization may be
safer or more useful after initial palliative surgery, such as in hypoplastic left heart syndrome or
following shunt procedures in those with complex intracardiac anomalies and pulmonary atresia.
The decision to perform a cardiac catheterization should be guided by a careful assessment of the
long-term benefits in management versus the risk of the procedure. Before catheterization the
medical condition is optimized for the anomalies present and the rapidity with which
catheterization may provide critical information to further stabilize the situation. Infants with a
duct-dependent anomaly are best managed with an infusion of prostaglandin E1, begun before and
continued throughout the
P.654
catheterization. The potential for procedural morbidity and mortality is greater in the sick newborn
(47). Potential morbidity includes blood loss, hypothermia, metabolic and respiratory acidosis,
arrhythmia, electrolyte imbalance, hypoglycemia, thrombosis of femoral arteries, and angiographic
myocardial stains.
Figure 33-15 A: Echocardiographic apical four-chamber view in systole. Color Doppler analysis
of the right heart demonstrates a jet of tricuspid regurgitation depicted by the blue flow jet (white
arrow). B: Quantification of the velocity of the regurgitant jet by application of continuous-wave
Doppler technique along the axis of the dotted line in the upper panel. Application of the
measured triscuspid regurgitant velocity within the Bernoulli equation allows the right-ventricle-toright-atrial peak pressure gradient to be measured and estimation of the right ventricular
pressure (LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle; TR tricuspid
regurgitation). (See color plate)
Hemodynamic Measurements
Hemodynamic data obtained by catheterization can now largely be deduced from noninvasive
measurements of blood pressure, oxygen saturation, and echocardiographic Doppler
measurements of pressure gradients. Direct measurement does not help preoperative neonatal
surgical management of most anomalies. When catheterization is done primarily for delineation of
anatomy, hemodynamic measurements can be readily obtained and can facilitate delineating the
clinical status and management. Hemodynamic measurements are often used to guide
interventional catheterization such as valvuloplasty. Sometimes, particularly postoperatively,
information from implanted catheters is useful for management of sick babies in the intensive care
unit. Catheters in the right atrium placed through an umbilical vein, systemic vein, or transthoracically in the operating room may be used to obtain central venous pressure and blood
oxygen saturation. These data can be used to infer preload and adequacy of cardiac output and, in
combination with blood pressure measurements, to infer relative afterload. Catheters in the
pulmonary artery, placed trans-thoracically at surgery or trans-venously, can be used to measure
left-to-right shunts and to measure pulmonary pressure to titrate pulmonary vasodilators.
The hemodynamic principals for these calculations are based on Ohm's law and the Fick principle
(see Table 33-9) (47). The former, when applied to hemodynamics, is the pressure drop across a
vascular bed equals the product of the flow and resistance across it. Therefore, the resistance
equals the difference of the arterial and venous pressure divided by the flow. The flow can be
calculated from the Fick principal, which is based on the premise that oxygen delivery to the body
equals oxygen consumption by the body. Oxygen consumption is routinely measured in the
catheterization laboratory and, in the intensive care unit, can be assumed to be 200 to 240 mL/
min/m2 in neonates (48). Oxygen delivery is the product of flow and the arterial-venous oxygen
content difference. Arterial and venous oxygen contents are calculated from the products of the
measured blood oxygen saturations and blood hemoglobin concentration and the hemoglobin
oxygen-carrying capacity (1.36 mL O2/g hemoglobin). Systemic and pulmonary flow can be
calculated as in Table 33-8. In those without a hemodynamic shunt, cardiac output can be
measured by the thermodilution method.
Management Procedures for Severe

Cardiac Disease
Preprocedure
An infant in difficulty in the first days of life because of heart disease has the
potential for rapid deterioration. Too often, the baby looks as though he will
survive but is near death hours later. The earlier symptoms appear, the faster
deterioration can take place. By the time the infant has
P.655
reached 1 or 2 months of age, concern about sudden shifts in status is less
warranted.
TABLE 33-9 HEMODYNAMIC FORMULATIONS
O2 consumption = O2 delivery
= Q (arterial O2 contentvenous O2 content)
Blood O2 content (mL/L) = Hgb (g/dl) 10 (dL/L) 1.36 (ml O2/g Hgb)
Hgb O2 Sat.
Average neonate O2 consumption = 200-220 mL/min/m2
Qs (L/min/m2) = O2 consumption/Hgb 13.6 (arterial O2 Sat. venous
O2 Sat.)
Qp (L/min/m2) = O2 consumption/Hgb 13.6 (pulm. venous O2 Sat.
pulm. arterial O2 Sat.)
Qp/Qs = (arterial O2 Sat. venous O2 Sat.)
P (mmHg) = Q R (Woods units)
Rs = Qs/(arterial mean pressure - RA mean pressure)
Rp = Qp/(pulm. arterial mean pressure - LA mean pressure)
Q, cardiac output or blood flow; Hgb, blood hemoglobin concentration;
Sat., saturation; Qs, systemic flow or cardiac output; Qp, pulmonary flow;
P, arterial mean pressure minus arterial atrial mean pressure; pulm.,
pulmonary; R, vascular resistance; Rs, systemic vascular resistance; Rp,
pulmonary vascular resistance.
From ref. 47.
Infants who present with severe cyanosis (See Color Plate) in the first days to
weeks of life may do so because right ventricular outflow is critically obstructed
and adequate pulmonary blood flow is dependent on a closing ductus arteriosus,
or because the great arteries are transposed and adequate mixing of the
pulmonary and systemic circulations is decreasing as the ductus arteriosus
constricts. Babies with congestive heart failure in the first week of life often have
obstructed left ventricular or aortic outflow, with descending aortic flow supplied
by a closing ductus arteriosus. In these babies, survival may depend on persistent
patency of the ductus arteriosus; dependency should be suspected, and
prostaglandin E1 therapy considered. If possible, echocardiography should be
used to confirm a specific anatomic diagnosis, but this may not be available in
many primary care facilities, and the infant's condition may not provide the time
before starting treatments to transport to a facility in which echocardiography is
available. If a duct-dependent anomaly is suspected from physical examination,
ECG, and chest radiograph (e.g., pulmonary atresia, hypoplastic left heart
syndrome), or if the condition of a baby with undiagnosed cardiac anomaly is
significantly worsening so that arterial oxygen saturation is less than 70% (e.g.,
as in d-transposition of the great arteries or critical pulmonary stenosis) or there
is severe congestive heart failure because of ductal closure (e.g., as in critical
aortic stenosis or coarctation), prostaglandin E1 therapy should be initiated even if
echocardiography is not available (see Fig. 33-12). The usual starting dose of 0.1
mg/kg/min can frequently be reduced to 0.05 to 0.02 mg/kg/min after
stabilization. The occurrence of relatively common side effects, particularly central
apnea, vasodilation with hypotension, and fever should be anticipated.
Endotracheal intubation should be performed prior to transport in infants receiving
prostaglandins, to reduce the risk should later-onset apnea occur.
Despite prostaglandin therapy, these critically ill infants may have low cardiac
output that may respond to the correction of common metabolic perturbations
including hypothermia, intravascular hypovolemia, hypocalcemia, and
hypoglycemia, but frequently, inotropic support is needed (Tables 33-10 and 3311). Measurement of pressure in a central venous catheter may guide fluid
therapy and permit administration of concentrated infusions of dextrose, calcium,
and vasoactive amines. After appropriate steps to correct contributing metabolic
abnormalities, fluid can be given in 5- to 10-mL/kg doses until adequate response
is achieved or circulatory congestion occurs. Infusion of dopamine or dobutamine
(5 to 20 mg/kg/min) or amrinone, should be added to support pump function as
needed. Higher doses or continuous infusion of epinephrine can be considered to
support refractory neonates until surgical palliation can be achieved. Digitalis
preparations are much less desirable for acute inotropic support of critically ill
infants who have variable renal and hepatic functions and electrolyte status.
Hyperventilation should be avoided in babies with certain lesions in which the
pulmonary and systemic circulations are in parallel, such as hypoplastic left heart
syndrome. Hyperventilation and oxygen administration in these babies can drop
pulmonary vascular resistance to low levels, resulting in runoff into the pulmonary
vasculature, systemic hypotension, and very low systemic blood flow.
The acyanotic cardiac infant who develops symptoms of increased respiratory
work and poor feeding after 2 to 4 weeks of life often has congestive heart failure
from decreasing pulmonary vascular resistance and increasing left-to-right shunt.
Treatment with digoxin, diuretics, and, in refractory cases, systemic vasodilators
is often indicated (see Table 33-10). Rarely, these infants have left-sided
obstructive lesions or myocardial disease (e.g., anomalous left coronary artery)
that requires different treatment (see Therapeutic Catheterization).
Therapeutic Catheterization
Interventions now performed on neonates in the catheterization laboratory include
balloon atrial septostomy by
P.656
P.657
P.658
P.659
Rashkind technique for transposition of the great arteries, creation of an atrial
septal defect in mitral atresia or hypoplastic left heart syndrome with restrictive
atrial communication using Brockenbrough atrial puncture and balloon dilation,
pulmonary and aortic valvuloplasty, pulmonary artery angioplasty, angioplasty of
discrete aortic coarctation with otherwise normal caliber aortic arch, and closure of
systemic-to-pulmonary arterial collateral vessels (47,49,50,51 and 52). To
perform a therapeutic procedure and to extract vital diagnostic information with
the least danger to the patient requires vigilance against a multitude of
treacherous pitfalls and acute sense of the clinical cost and benefit of each
maneuver contemplated. The neonate undergoing study is ill, often critically ill,
and may have a widely fluctuating physiologic state. Before catheterization the
baby is medically stabilized as best possible as dictated by the baby's anomalies,
condition, and the rapidity with which catheterization may be required to further
stabilize the situation. Duct-dependent infants are managed with an infusion of
prostaglandin E1 (53,54). Careful and constant attention to maintenance of proper
thermal environment, minimization of blood loss, vascular access and hemostasis,
anticoagulation, metabolic status, respiratory status, and catheter manipulation
optimizes the outcome.
TABLE 33-10 COMMON ORAL DRUGS FOR THE TREATMENT OF CONGESTIVE HEART FAILURE
Genetic Drug
Digoxin
Proprietary Name
Lanoxin
Form
Dose
Elixir: 50 g = 0.05
mg/mL
Action
Digitalizing dose:
Premature, 20 g/kg
Toxicity
NaK ATPase inhibitor,

increases contractility
Atrioventricular block (monitor ECG during

loading), tachydysrhythmias, vomiting; use
Term, 30 g/kg Initial

dose, In 6 h, In
12h,
Maintenance dose:
Premature, 34 g/
kg/12 hr
Term, 45 g/kg/12 hr
Suspension: 10 mg/mL1 mg/kg/12 hr PRN to Loop of Henle Cl-pump
2.0 mg/kg/8 hr
inhibition, diuretic
Furosemide
Lasix
Chlorothiazide
Diuril
Spironolactone
Aldactone
Captopril
Capoten
with caution in renal failure and myocarditis;

decrease doseby one-half if used with
quinidine
Hyponatremia, hypokalemia, hypochloremic

alkalosis nephrocalcinosis
Suspension: 10 mg/mL1015 mg/kg/12 hr
Blocks distal tubular Na Hyponatremia, hypokalemiahypochloremic

reabsorption, diuretic
alkalosis, hyperbilirubinemia, hyperuricemia,
hyperglycemia
Suspension: 5 mg/mL 13 mg/kg/d, tab Blocks tubular
Hyperkalemia
Tablet: 25 mg
tab, crushed qodqd aldosterone receptor
diuretic
Tablet: 12.5 mg, 25
Term, 0.11.0 mg/
Angiotensin-converting Hypotension, azotemia, proteinuria, may
mg
kg/8 hr
enzyme inhibition
Preterm, 0.050.2 mg/ decreases afterload
kg/812 hr
cause hyperkalemia when given with,

spironolactone or potassium; use with
caution in low dose in premature neonates
TABLE 33-11 INTRAVENOUS VASOACTIVE DRUGS
Drug
Dopamine
Dose
(g/kg/
min)
25
520
Action
Systemic
Resistance
Preload
Pulmonary
Resistance
D, 1
+/-
+/-
D, ,
Dobutamine
220
Epinephrine
0.051.0 , 1, 2
1mild 2,
Heart
Rate
Contractility
Use
+/-
CO,
BP
Toxicity
Tachycardia,
dysrhythmias, necrosis
with extravasation,
+/-
CO
CO,
BP,
HR
renal blood flow at higher

doses
Tachycardia,
dysrhythmias, necrosis
with extravasation
Tachycardia,
dysrhythmias necrosis
with extravasation renal
blood flow
Isoproteronol
0.052.0 1, 2
CO, HR Marked tachycardia,

dysrhythmias,
hypotension when
Amrinone
510
Phosphodiesterase
inhibition
CO
Thrombocytopenia,
dysrhythmias
EDRF-like action
CO BP
Hypotension, V/Q
mismatch, thiocyanate
volume depleted
(load:
1.0 mg/
kg)
Nitroprusside
0.55
Nitroglycerin
15
EDRF-like action
CO,
preload
Phenylephrine
0.54.0
+/-
+/-
TF,
cyanotic
toxicity
Hypotension, V/Q
mismatch,
methemoglobinemia,
Cardiac output, renal
blood flow
spells
+/-, may or may not; , decrease; , increase; , -adrenergic; , -adrenergic; BP, blood pressure; CO, cardiac output; D, dopaminergic; EDRF,
endothelial-derived relaxing factor; HR, heart rate; TF, tetralogy of Fallot; V/Q, pulmonary ventilationperfusion ratio.
TABLE 33-12 DIFFERENTIAL DIAGNOSIS OF CYANOTIC HEART DISEASE
Diagnosis
Single S2, respiratory work,

pulse amplitude, perfusion,
+/- SRM
Transposition of great arteries

(IVS, VSD)
Truncus arteriosus
Total anomalous pulmonary

venous connection
Tricuspid atresia
Without PS
With PS
Radiographic Findings
Pulmonary arterial markings,

cardiomegaly
Electrocardiographic Findings
LV force usually, develops RAE, RAD, RVH
Split S2, +/- murmur, +/-
Pulmonary arterial markings, +/- Develops RAE, RAD, RVH

respiratory work (i.e., peaceful cardiomegaly with narrow
mediastinum (i.e., egg on a
cyanosis)
string)
Develops RAE, RAD, BVH
Split S2, multiple clicks, soft to
loud SEM, +/- DRM,
respiratory work
cardiomegaly
Narrow S2 split, +/- murmur,
Pulmonary venous markings,

diffuse interstitial markings
Develops RAE, RAD, RVH
Left axis deviation
respiratory work
Split S2, heave, SRM
Single S2, SEM
cardiomegaly Pulmonary arterial

markings +/- cardiomegaly
Tetralogy of Fallot
With PS
With PA
Single S2, SEM
+/- Pulmonary arterial
Single S2 continuous murmur,
markings, +/- boot-shaped heart
LSB back, axillae
Develops RAE, RAD RVH/
, Pulmonary arterial markings

and heart size
Pulmonary stenosis (IVS or SV)
Single S2, click, SEM
Pulmonary arterial markings
In IVS QRS axis 0- 100, develops RAE RAD, RVH
Pulmonary atresia (IVS or SV)
Single S2, soft SRM
Pulmonary arterial markings
In IVS QRs axis 0-80, RV forces, +/- develops Q

waves
Persistent pulmonary hypertension Narrow split or single S2, S2

loudness, +/- SRMs
Pulmonary arterial markings +/- Develops RAE, RAD, RVH

parenchymal infiltrates +/cardiomegaly
Single ventricle is usually associated with transposition of the great arteries, and in the absence of PS or PA, it presents similar to transposition with
ventricular septal defects.
+/-, may or may not be present;
septum; LSB, left sternal border;
enlargement; RV, right ventricle;
ventricle; VSD, ventricular septal
, decreased; , increased; BVH, biventricular hypertrophy; DRM, diastolic regurgitant murmur; IVS, intact ventricular
LV, left ventricle; PA, pulmonary atresia; PS, pulmonary stenosis; RAD, right axis deviation; RAE, right atrial
RVH, right ventricular hypertrophy; SEM, systolic ejection murmur; SRM, systolic regurgitant murmur; SV, single
defect.
Surgery
Life threatening heart disease in neonates often requires surgery. Early
recognition, safe transport to a cardiac center, accurate diagnosis, and an
experienced surgical team are needed for success. Anesthesiologists
familiar with the problems of neonatal cardiac patients and a wellequipped intensive care unit with trained personnel contribute to
successful management of these babies. The postoperative care requires
fine adjustment of blood volume, body temperature, fluid and electrolyte
balance, oxygenation, ventilation, and hemodynamic measurements.
Close cooperation between the cardiologists, intensivists, and surgeons
responsible for the care of these infants is mandatory.
The timing of surgical intervention depends on the anatomic diagnosis
and the relative outcome of surgery sooner or later. Single-stage repair
should be used if possible when outcomes are at least equivalent to those
for staged procedures to avoid the added jeopardy of performing and
additional procedure before a definitive procedure is done. There is also
increasing evidence that early repair results in improved cardiac status
and neurologic function (12,55). For complex anomalies, particularly
those with one functional ventricle, a staged approach is often required,
with initial life-saving, palliative operation done in infancy, followed
months or years later by additional reparative surgery.
Cyanotic Lesions
The differential diagnosis of cyanotic heart disease includes many disorders (Table 33-12). Lesions usually
associated with decreased pulmonary flow include the tetralogy of Fallot, pulmonary stenosis, tricuspid atresia,
pulmonary atresia with intact ventricular septum, and Ebstein disease. Cyanotic lesions usually associated with
increased pulmonary vascular markings include d-transposition of the great arteries, hypoplastic left heart
syndrome, total anomalous pulmonary veins, truncus arteriosus, and single ventricle.
Anomalies with Cyanosis Caused by Separate Transposed Systemic

and Pulmonary Circulations
d-Transposition of the Great Arteries
With transposition of the great arteries, the aorta arises from the right ventricle and the pulmonary artery from
the left ventricle. In the most common form, d-transposition, the aorta is anterior to and to the right of the
pulmonary artery rather than in its normal rightward and posterior position.
Transposition of the great arteries is one of the most common congenital heart lesions presenting in the
newborn period (see Tables 33-1 and 33-2), and is a frequent cause of death among un-operated neonates with
congenital heart disease. The male-female ratio is 1.8:1, and the average birth weight is greater than that for
other patients with congenital heart disease, although not for the general population. Transposition is associated
with other cardiac abnormalities, including ventricular septal defect, patent ductus arteriosus, pulmonary valve
stenosis, hypoplastic right ventricle, and coarctation.
Pathophysiology
The systemic and pulmonary circulations are normally in series with each other, but in complete transposition,
the circulations are in parallel. Deoxygenated systemic venous blood returns to the right atrium, enters the right
ventricle, and exits through the aorta. Maximally oxygenated pulmonary venous blood enters the left atrium and
the left ventricle, and then returns to the pulmonary arteries and the lungs. Without some communication
between the pulmonary and systemic circulations, survival is impossible; oxygenated blood cannot be delivered
to the systemic circulation, nor can systemic venous blood be directed to the lung to become oxygenated. An
atrial communication, ventricular defect, or patent ductus arteriosus, singly or in combination, may provide for
mixing between the circulations (Fig. 33-16 and 33-17). The foramen ovale and ductus arteriosus, both normally
patent in the fetus, usually close soon after birth. Infants with transposition and intact ventricular septum
become extremely cyanotic within the first few hours or days after delivery, as closure of the foramen ovale and
ductus arteriosus occurs and mixing between the circulations diminishes. The severe hypoxemia may lead to
metabolic acidosis. Survival depends on prompt supportive medical care and reestablishment of patency of the
ductus arteriosus and interatrial communication to improve mixing and oxygenation.
Figure 33-17 Diagram of the anatomy and physiology of transposition with intact ventricular septum in a 1day-old girl who was cyanotic at birth. At catheterization, she had less than systemic pressure in the left (i.e.,
pulmonary) ventricle. The patent ductus arteriosus shunts blood into the pulmonary circuit, and the foramen
ovale shunts an equal amount out of the pulmonary circuit. If this were not the case, relative blood volume
would shift to one side of the circulation in a matter of minutes. If the ductus spontaneously closed, the infant's
condition would become precarious. If the ductus were dilated with prostaglandins, the infant would become
pinker but might experience respiratory difficulty because of excess pulmonary flow. The numbers below the
chamber name are pressure measurements (mm Hg) determined at cardiac catheterization; the percentages
indicate oxygen saturation. LA, left atrium; LV, left ventricle; PA, pulmonary artery; PV, pulmonary vein; RA,
right atrium; RV, right ventricle; SVC, superior vena cava. Adapted from Mullins CE, Mayer DC. Congenital
heart disease: a diagramatic atlas. New York: Alan R Liss, 1988, with permission.
P.660
Infants born with transposition and a large ventricular septal defect are less cyanotic because the ventricular
defect allows mixing. These babies may not be recognized in the newborn period but appear in subsequent
weeks with congestive failure. The combination of a large pulmonary flow, pulmonary hypertension, and
elevation of left atrial pressure leads to the development of congestive heart failure (see Fig. 33-16) and later
pulmonary vascular obstructive disease. Anatomic changes during the first few months of life may result in
important hemodynamic changes. A large ventricular septal defect may spontaneously diminish in size or close,
reducing mixing and increasing hypoxemia. Increasing pulmonary stenosis may decrease pulmonary flow and
thereby increase cyanosis (See Color Plate) but improve congestive heart failure. Atrial septal defects created by
balloon septostomy and those made by surgical septectomy may spontaneously diminish in size or close.
Clinical Findings
Infants with transposition of the great arteries and an intact ventricular septum develop marked cyanosis
accompanied by mild tachypnea soon after birth. Often the infants, although tachypneic, do not seem distressed
(i.e., peaceful cyanosis). The cardiac examination, chest radiograph, and ECG may otherwise be normal. The
heart sounds are normal (i.e., the second heart sound splits), and there may be no significant murmur. Because
the usual clinical findings, besides cyanosis, can be unremarkable, one of the most important diagnostic tests is
the hyperoxia test. Failure of the arterial arterial oxygen pressure (PaO2) (often <30 mm Hg in room air) to rise
significantly after the inhalation of 100% oxygen for a 10-minute period is strong presumptive evidence for
cyanotic heart disease, most commonly complete transposition.
Figure 33-16 Diagram of the anatomy and physiology of transposition of the great arteries with a single
membranous ventricular septal defect in a 1-month-old baby who had mild cyanosis and controlled congestive
heart failure. There is equilibration of pressure between the ventricles and elevation of left-sided diastolic
pressures. After balloon septostomy, the arterial saturation rose to 75%. The numbers below the chamber
name are pressure measurements (mm Hg) determined at cardiac catheterization; the percentages indicate
oxygen saturation. LA, left atrium; LV, left ventricle; PA, pulmonary artery; PV, pulmonary vein; RA, right
atrium; RV, right ventricle; SVC, superior vena cava. Adapted from Mullins CE, Mayer DC. Congenital heart
disease: a diagramatic atlas. New York: Alan R Liss, 1988, with permission.
The ECG may show some excessive right ventricular forces. On chest radiograph the heart and pulmonary
vascularity may initially appear normal, although cardiac enlargement, a narrow mediastinum, and pulmonary
plethora are frequently present or develop.
Figure 33-18 Echocardiographic subxiphoid view of a neonate with transposition of the great arteries. (AO
aorta; ARCH, aortic arch; LA, left atrium; LV, left ventricle; PA, pulmonary artery; RV, right ventricle; TV,
tricuspid valve.)
P.661
Subcostal echocardiography reveals the diagnosis. The great artery arising from the left ventricle has an
abnormal course and then bifurcates into the right and left pulmonary artery. The right ventricle gives rise to a
great artery that passes relatively straight superiorly to the posterior arching aorta (Fig. 33-18).
Echocardiographic examination can also determine the patency of the foramen ovale and ductus arteriosus, the
nature of associated anomalies, and the coronary anatomy relevant to the surgical arterial switch procedure.
Infants with transposition and a large ventricular septal defect usually present with congestive failure and mild
cyanosis, between zero and 6 weeks of age. Poor weight gain, tachypnea, and excessive diaphoresis are
common, and wheezing occurs in older infants. A loud systolic murmur is present maximally at the lower left
sternal border, often associated with a mid-diastolic flow rumble. An S3 may produce a gallop rhythm. Rales
may be audible in the lungs.
ECG reveals right axis deviation and right atrial and right ventricular hypertrophy. Infrequently, if the right
ventricle is hypoplastic, right ventricular forces may be absent or reduced, and left ventricular hypertrophy is
present. Chest radiograph characteristically shows considerable cardiomegaly and pulmonary plethora.
Echocardiographic examination should identify the location of the ventricular septal defect, its relation to the
great arteries and the atrioventricular valves, and complex associated problems including straddling or abnormal
tricuspid valve, hypoplastic right ventricle, valvar or sub-valvar pulmonary stenosis, coarctation of the aorta,
juxtaposition of the atrial appendages, and anomalous systemic or pulmonary venous drainage.
Differential Diagnosis
Most infants with transposition and intact ventricular septum are readily recognized as cyanotic infants with little
respiratory distress and without significant murmur. Often a split second heart sound can be distinguished on
exam, and chest radiograph demonstrates cardiomegaly and increased pulmonary flow, helping to distinguish
transposition from other lesions with cyanosis and little murmur, pulmonary valve atresia with intact ventricular
septum, and total anomalous pulmonary venous connection. Diagnosis of transposition of the great arteries can
be complicated if other abnormalities, such as straddling tricuspid valve, hypoplastic right ventricle, coarctation
of the aorta, or pulmonary stenosis, exist (see Table 33-12). Depending on the type and severity of the
associated cardiac malformations, the clinical symptoms and findings in infants with complicated transposition of
the great arteries may closely resemble those of almost any other cyanotic heart lesion.
The clinical picture in infants with transposition of the great arteries, ventricular septal defect, and pulmonary
stenosis or atresia is virtually indistinguishable from that of tetralogy of Fallot or pulmonary atresia with a
ventricular septal defect. If d-transposition of the great arteries is associated with a large ventricular septal
defect and limited cyanosis, it is sometimes mistaken for other lesions with a large left-to-right shunt, such as a
ventricular septal defect with normal aortic root or total anomalous pulmonary venous return without
obstruction. The absence of cyanosis identifies the former, and echocardiography can differentiate all of these
anomalies.
Treatment
In those with established or suspected transposition with intact ventricular septum, prostaglandin E1 is infused
to open and maintain patency of the ductus arteriosus, to improve mixing and systemic oxygenation. Because
prostaglandin E1 may cause apnea and vasodilation, support with mechanical ventilation, volume infusion, and
sometimes inotropic agents may be required. Some babies also require a widely open atrial defect for adequate
oxygenation, and most do better with one. Balloon atrial septostomy usually results in considerable clinical
improvement in those sick from severe cyanosis, and some require it promptly for survival (see Fig. 33-19). The
anatomy of the coronary arteries and associated lesions may also be established by catheterization. The type of
surgical procedure employed depends on the associated cardiac defects. Anatomic repair with an arterial
switch operation has been demonstrated to be the procedure of choice in neonates with uncomplicated
transposition (56,57). The aorta and pulmonary arteries are transected, and the distal vessels are rejoined to
provide normal physiologic connections. A button of proximal aortic tissue surrounding each coronary artery
origin is cut, and both coronaries with the surrounding aortic button are moved from the native transposed aorta
and attached to the neo-aorta. The ductus arteriosus is ligated, and the atrial septostomy is closed. Infants with
transposition and isolated ventricular septal defect undergo arterial switch procedure and closure of the
ventricular septal defect. Long-term outcome of arterial
P.662
switch procedure is not yet known but appears promising (58). Although a small number have difficulty with the
coronary artery kinking, pulmonary artery compression, or anastomotic narrowing, the large majority do very
well and lead essentially normal lives in childhood.
Figure 33-19 Echocardiographic subxiphoid view demonstrating a balloon septostomy catheter with the
inflated balloon (arrow) being pulled from the left atrium (LA) to the right atrium (RA) through the foramen
ovale. This septostomy technique is used in neonates with transposition of the great arteries to create an atrial
septal defect to increase intracardiac mixing of fully and incompletely oxygenated blood, thereby improving
systemic oxygenation.
The additional presence of significant pulmonary or aortic valvar or sub-valvar stenosis can prohibit a
straightforward arterial switch procedure. Severe pulmonary valve stenosis generally occurs with a large
ventricular septal defect, and a Rastelli procedure or variation of it is done by placing a patch from the crest of
the ventricular septum into the upper right ventricle to direct left ventricular blood to the transposed aorta and
directing blood flow from the lower right ventricle to the pulmonary artery by interposition of a conduit
(generally homograft) from a right ventriculotomy to the distal transected main pulmonary artery. The presence
of native aortic valve stenosis can be dealt with by a modification of the Damus-Kaye-Stansel type in which left
ventricular blood flow to the body is through anastomosis of the transected proximal main pulmonary artery to
the ascending aorta, and pulmonary blood flow is through a conduit interposed between the right ventricle and
distal main pulmonary artery. These procedures require revision of the surgically implanted conduits with growth
and in the not infrequent occurrence of extrinsic compression. Without therapy, 95% of infants born with
transpositions die within 1 year. With aggressive medical and surgical treatment, mortality is less than 5% to
10%.
Anomalies with Cyanosisfrom Decreased Pulmonary Blood Flow

Tetralogy of Fallot
The tetralogy of Fallot is characterized by a large ventricular septal defect and infundibular pulmonary stenosis
or pulmonary atresia. Both environmental factors and a number of genetic disorders are associated with
tetralogy of Fallot. Many have microdeletions in a critical region of chromosome 22q11 (25,30). The primary
anatomic pathologic abnormality appears to be anterior deviation of the upper part of the ventricular septum,
known as the conal septum that separates the anterior pulmonary outflow of the right ventricle from the subaortic left ventricular outflow. This results in the sub-aortic anterior malalignment ventricular septal defect and
in the right ventricular infundibulum being hypoplastic and narrow. There is often considerable valvar pulmonary
stenosis, hypoplasia of the pulmonary arteries, right ventricular hypertrophy, a relatively large ascending aorta,
and a right aortic arch (25%). In infants with pulmonary atresia, pulmonary perfusion occurs by way of a patent
ductus arteriosus or by systemic-to-pulmonary arterial collateral vessels. Five percent of patients have abnormal
coronary distribution that may influence surgical correction. Tetralogy is one of the most common cyanotic
congenital heart lesions presenting in the newborn period (see Tables 33-1 and 33-2) and is occasionally (10%)
associated with severe extra-cardiac malformations.
Pathophysiology
Depending on the severity of right ventricular outflow obstruction, there may be intracardiac left-to-right flow or
right-to-left shunt and hypoxemia. Pressures equalize between the ventricles through the large septal defect.
The peripheral arterial oxygen saturation depends on the amount of systemic venous admixture and the
absolute pulmonary flow (Fig. 33-20). The extent of systemic venous admixture, that is, right-to-left shunting of
systemic venous blood away from the pulmonary outflow through the ventricular septal defect to the aorta, is
directly related to the severity of the pulmonary stenosis and inversely related to the systemic vascular
resistance. The amount of pulmonary blood flow depends on the amount of antegrade flow through the right
ventricle outflow and the existence of alternative sources of flow (through a ductus arteriosus or systemic-topulmonary arterial collateral vessels). For example, in pulmonary atresia with a ventricular septal defect, the
entire right heart output passes right to left through the ventricular defect, and pulmonary flow is supplied by a
ductus arteriosus or collateral vessels and is usually less than normal. Cyanosis is the result. If the newborn has
large aortic-pulmonary collateral vessels perfusing the lung, pulmonary blood flow may be large; and the infant
may be barely cyanotic and some have congestive heart failure.
The pulmonary outflow stenosis in tetralogy of Fallot is progressive. Clinically significant cyanosis is present at
birth in 25%, by 1 year of age in 75%, and almost all have become cyanotic by 20 years of age. Infundibular
hypoplasia and stenosis are progressive in both absolute and relative terms. Progression of the stenosis during
the first 6 months of life is most often largely relative because of a lack of adequate infundibular expansion
during rapid somatic growth and need for proportionately greater pulmonary flow (59).
P.663
With time, complete atresia can occur. A patent ductus arteriosus usually closes within the first week of life,
resulting in severe and often sudden hypoxemia or cyanotic spells.
Figure 33-20 Diagram of the cardiac anatomy and physiology in a girl with mild cyanosis and a loud murmur
audible at birth. She was found to have tetralogy of Fallot and was followed without medication prior to primary
reparative surgery. The numbers below the chamber name are pressure measurements (mm Hg) determined at
cardiac catheterization; the percentages indicate oxygen saturation. LA, left atrium; LV, left ventricle; PA,
pulmonary artery; PV, pulmonary vein; RA, right atrium; RV, right ventricle; SVC, superior vena cava. (Adapted
from Mullins CE, Mayer DC. Congenital heart disease: a diagramatic atlas. New York: Alan R Liss, 1988, with
permission.
Hypoxemia of rapid onset is characteristic of a tet spell secondary to an infundibular spasm. This may be
associated with an increased adrenergic contractile state, systemic vasodilation associated with a meal, warm
bath, or certain types of anesthesia, or constriction of a ductus arteriosus. Hypoxemia may result in a fall in
systemic vascular resistance, metabolic acidosis, hyperpnea, and further hypoxemia. Hyperventilatory
compensation for the metabolic acidosis may be ineffective because of inadequate pulmonary blood flow. The
self-aggravating cycle of increasing hypoxemia and metabolic acidosis can progress to unconsciousness and
convulsions.
Clinical Findings
Cyanosis of various degree and mild tachypnea often develop soon after delivery. If hypoxemia becomes severe,
the infant may become hypotonic, hypotensive, and bradycardic. Tet spells characterized by a sudden onset of
irritability, hyperpnea, and increasing cyanosis may develop. Spells may end in a loss of consciousness, seizures,
cerebral injury, hemiparesis, or death. The disappearance of a previously heard right ventricular outflow systolic
murmur with increased cyanosis suggests a spell and constitutes an indication for immediate therapy.
The second heart sound is single. With typical tetralogy, systolic murmurs from the pulmonary stenosis and/or
ventricular septal defect are at the left upper and mid sternal border, respectively. With pulmonary atresia,
when the systolic murmur is absent; there may be a constant apical systolic ejection click and prominent
continuous murmurs of a patent ductus arteriosus or aortic-pulmonary collaterals, audible at the base, in the
axillae, and/or over the back. A patent ductus arteriosus usually does not cause a continuous murmur in the first
months of life; therefore, the presence of murmurs with cyanosis and a single S2 in a neonate strongly suggests
tetralogy of Fallot with pulmonary atresia. Delay in height, weight, and skeletal maturation is common,
particularly when DiGeorge syndrome is present, but some infants flourish despite severe hypoxemia.
Some young infants with the anatomic but acyanotic tetralogy of Fallot experience congestive heart failure from
left-to-right ventricular septal shunting, later develop increased pulmonary stenosis, recover from congestion,
and become cyanotic. Congestive heart failure also sometimes occurs in infants with pulmonary atresia when
numerous or very large aortic-pulmonary collaterals are present. Although sub-acute bacterial endocarditis and
brain abscess are common in older children with tetralogy of Fallot, these complications are extremely rare in
infancy. However, spontaneous cerebrovascular accidents are common, particularly in infants with severe
hypoxemia and relative anemia (<6-8 g/dL of oxyhemoglobin).
Chest radiograph shows a normal-sized heart, sometimes with right ventricular enlargement resulting in an
upturned apex and an absent or diminished main pulmonary artery segment (i.e., boot-shaped heart),
diminution of the pulmonary vasculature, and, in 25%, the aorta arching to the right.
ECG demonstrates right axis deviation, right atrial enlargement, and right ventricular hypertrophy, which at
birth is often difficult to differentiate from normally prominent right ventricular forces.
Echocardiographic examination shows anterior and leftward deviation of the infundibular septum, creating subpulmonary stenosis and a malalignment ventricular septal defect with a large overriding aortic root (Fig. 33-21).
Additional ventricular or atrial septal defects, central pulmonary artery hypoplasia, and coronary artery anatomy
can usually be delineated by echocardiography but may sometimes require cardiac catheterization for
clarification.
Angiography (Fig. 33-22), and often magnetic resonance imaging, may be used to determine presence and
anatomy of distal pulmonary artery stenoses and systemic-arterial-to-pulmonary-arterial collaterals.
The features of cyanosis, a harsh systolic ejection murmur, chest radiographic findings of diminished pulmonary
P.664
vasculature with a normal-sized heart, and ECG evidence of right ventricular hypertrophy are characteristic of
tetralogy of Fallot (see Table 33-12). The same findings with a continuous murmur suggest tetralogy of Fallot
with pulmonary atresia. A few infants with tetralogy of Fallot and an underdeveloped pulmonary valve present
with a characteristic to-and-fro murmur (i.e., steam engine sound) and severe respiratory distress caused by
bronchial or tracheal compression by aneurysmally dilated pulmonary arteries.
Figure 33-21 Echocardiographic subxiphoid (A) and parasternal short-axis (B) views of an infant with
tetralogy of Fallot. Anterior deviation of the conal septum (above the arrow) is associated with malalignment
ventricular septal defect and obstruction of the right ventricular outflow above the pulmonary artery; arrow,
ventricular septal defect and narrowed right ventricular outflow; ANT, anterior; AV, aortic valve; LA, left atrium;
LV, left ventricle; LVOT, left ventricular outflow tract; PA, pulmonary artery; RA, right atrium; RV, right venticle.
Treatment
Initial treatment and timing of surgery depend on the severity of the pulmonary stenosis. If severe cyanosis
develops in the newborn, prostaglandin E1 should be employed to reopen the ductus arteriosus, improve
pulmonary perfusion and stabilize the infant, and surgery should be done. Cyanotic tet spells are much more
likely to occur in the infant with a moderate or greater degree of preexisting cyanosis, and echocardiographic
findings of severe sub-valvar infundibular stenosis and hypoplasia. Later onset cyanotic tet spells should be
treated with oxygen, intramuscular or subcutaneous morphine sulfate (0.1 mg/kg), intravenous administration
of saline boluses and sodium bicarbonate (approximately 1 mmol/kg), and, if needed, phenylephrine (0.1 mg/kg
subcutaneously; 5 to 20 mg/kg by intravenous bolus; 0.1 to 0.5 mg/kg/min by intravenous infusion) titrated to
elevate systemic vascular resistance and pressure. Propranolol may be of some value in treating the infant with
a reactive infundibulum. The hemoglobin concentration should be maintained high enough to permit adequate
oxygen transport. The occurrence of a single tet spell is an indication for surgery, possibly as an emergency
procedure.
The newborn with tetralogy of Fallot and little or mild cyanosis can be carefully observed with repeated
measurement of transcutaneous systemic oxygen saturation until a stable level is apparent after ductal closure.
Many remain asymptomatic through the first 4 to 6 months of life, and can then undergo surgery with greater
likelihood of preserving adequate pulmonary valve function. However, because the right ventricular outflow
obstruction is often progressive, careful serial follow-up is prudent.
In the past, critically ill infants who required surgery underwent palliative procedures, usually a shunt between
the subclavian artery and branch pulmonary artery (Blalock-Taussig). The shunt was ligated during a reparative
operation when the child was older. Infants with uncomplicated tetralogy of Fallot requiring surgery now
undergo one-stage reparative procedures with excellent results (60). Using sternotomy and cardiopulmonary
bypass, the right ventricular outflow tract is enlarged with a pericardial patch, and the ventricular septal defect
is closed with a Dacron patch. There are several potential late sequlae, including late dysrhythmias and
problematic pulmonary regurgitation, but most patients with uncomplicated tetralogy of Fallot have an
asymptomatic course into young-mid adulthood (61). Those with very low birth weight, anomalous origin of the
left anterior descending coronary from the right coronary artery and those with pulmonary atresia and
hypoplastic-distorted pulmonary arteries may require a palliative shunt with a more definitive repair, entailing a
conduit from the right ventricle to the pulmonary artery, when the child is older.
Figure 33-22 Angiograms from a cyanotic infant with complex tetralogy of Fallot with atresia of the right
ventricular outflow, pulmonary valve and proximal main pulmonary artery demonstrate the anatomy of the
hypoplastic mediastinal pulmonary arteries and collateral vessels that perfuse them. A. Left ventricular
angiogram demonstrates tetralogy of Fallot with sub-aortic ventricular septal defect and pulmonary outflow
atresia. B. Balloon occlusion angiography in the descending thoracic aorta demonstrates collateral vessels from
the aorta to the right and left lung. C. Additional balloon occlusion aortography demonstates the collateral
vessel to the left lower pulmonary artery is in continuity with hypoplastic central mediastinal and right
pulmonary arteries. Left sublavian arteriography, not shown, demonstrated an additional collateral soley
supplying much of the left upper lung lobe. D. Right ventricular angiography after surgical implantation of a
homograft conduit from right ventricle to the central pulmonary artery, and catheter balloon angioplasties of
the left and right pulmonary arteries demonstrates substantial increases in sizes of the pulmonary arteries. LV,
left ventricle; LLL-PA, left lower lobe pulmonary artery; LPA, left pulmonary artery; LUL-PA, left upper lobe
pulmonary artery; PA, central pulmonary artery; prox, proximal; RLL-PA, right lower lobe pulmonary artery;
RPA, right pulmonary artery; RV, right ventricle; VSD, ventricular septal defect.
P.665
In general, the earlier severe hypoxemia develops, the more severe is the tetralogy of Fallot, and the poorer the
prognosis without surgery. The overall mortality rate without surgery was approximately 35% by 1 year of age.
Pulmonary Stenosis
Pulmonary Valve and Subvalvar Stenosis
Pulmonary valve stenosis is one of the more common intracardiac anomalies detected in the first month of life
(Table 33-1). It is usually mild and not progressive. Even mild obstruction (<10 mm Hg systolic pressure
gradient across the obstruction) produces a readily audible murmur. Moderate and severe pulmonary valve
stenosis detected in the first week of life often progresses for a limited time over the following weeks or months.
It generally presents with an isolated murmur radiating to the suprasternal notch, often but not always with an
early systolic click resembling a split S1, little or no cyanosis, and no signs of congestive heart failure. A
parasternal thrill, high-pitched systolic ejection murmur, and single second heart sound indicate severe
pulmonary valve stenosis. Severe valvar obstruction may produce cyanosis (Fig. 33-23)
P.666
because of right-to-left shunting through a foramen ovale or rarely may present with the findings of right-sided
congestive heart failure. In its most severe form, a very immobile, critically stenotic pulmonary valve requires
ductal patency for adequate pulmonary blood flow and systemic arterial oxygenation. The right ventricular
chamber may be small and noncompliant, resulting in some degree of right-to-left shunting although the
foramen ovale, even after removal of the stenosis. Infundibular (i.e., sub-valvar) obstruction as an isolated
lesion is rare; its presence usually indicates an associated ventricular defect.
Figure 33-23 Diagram of the cardiac anatomy and physiology in a 3-day-old baby with cyanosis from birth.
There was no murmur. The pulmonary valve was nearly atretic with only a tiny orifice. A valvotomy was done,
and the cyanosis resolved in 3 weeks. The numbers below the chamber name are pressure measurements (mm
Hg) determined at cardiac catheterization; the percentages indicate oxygen saturation. LA, left atrium; LV, left
ventricle; PA, pulmonary artery; PV, pulmonary vein; RA, right atrium; RV, right ventricle; SVC, superior vena
cava. (Adapted from Mullins CE, Mayer DC. Congenital heart disease: a diagramatic atlas. New York: Alan R
Liss, 1988, with permission.
Figure 33-24 A: Lateral view of the right ventriculogram of a neonate with mild cyanosis, loud murmur, and
severe pulmonary valve stenosis. Cineangiography during injection of contrast through a catheter (C) into the
right ventricle (RV) demonstrates a jet of contrast through the small orifice (arrow) of a thickened doming
stenotic pulmonary valve (PV). There is poststenotic dilation of the main pulmonary artery. B: Immediately
following the angiogram seen above, the pulmonary valve was crossed with a catheter, and balloon pulmonary
valvuloplasty performed over a guide wire, reducing the peak-to-peak systolic pressure gradient from 82 mm
Hg to 2 mm Hg.
associated with other cardiovascular or somatic malformations.
Chest radiography demonstrates normal heart size with fairly normal contour except for occasional upturning of
the apex and normal or decreased pulmonary vascular markings. The electrocardiogram is usually normal at
birth, although with severe and critical stenosis there may be relative mild left axis deviation for age (R axis 60
to 90 degrees) and increased or decreased right ventricular forces (see Table 33-12).
Echocardiography can determine the valve leaflet mobility, systolic pressure gradient, presence of right
ventricular or infundibular hypoplasia, ductal patency, and possible associated anomalies such as atrial septal
defect.
Only reassurance and observation are required for the mild stenoses because progressive obstruction is rare.
Because moderate obstructions in early infancy often become progressively worse with growth, the patients
should be periodically examined with this in mind. The severe and critical obstructions are relieved by catheter
balloon dilation (Fig. 33-24) (47,49). Recurrence is unusual. Bacterial endocarditis is rare, but the administration
of antibiotics
P.667
for prophylaxis before procedures that may be accompanied by bacteremia is recommended.
Pulmonary Artery Stenosis

Proximal pulmonary artery stenosis is probably the most common form of congenital heart disease discovered
by hearing a murmur. The abnormality most often appears to be acute angulation, sometimes with mild
narrowing of the proximal pulmonary artery branches. Minimal obstruction produces a readily audible murmur in
the left upper sternal border, radiating to the clavicular regions, axillae, and back. Echocardiography can detect
proximal pulmonary artery narrowing and pressure gradient and, when bilateral and severe, the presence of
right ventricular hypertension. Severe peripheral pulmonary stenosis is rare and is usually associated with other
problems such as tetralogy of Fallot, Williams syndrome, Alagille arteriohepatic dysplasia, and infants born of
mothers with rubella. Angiography or magnetic resonance imaging is required to visualize the more distal
pulmonary arteries. Pulmonary artery stenoses are usually amenable to catheter angioplasty and/or stenting.
Pulmonary Atresia with Intact Ventricular Septum

When the pulmonary valve is atretic and there is no ventricular septal defect, blood cannot pass through the
right ventricle in the fetus. Without normal flow, the right ventricle cavity cannot grow normally and changes
little in nature and size. If a severely stenotic valve becomes atretic late in gestation, the right ventricle can be
near normal and quite adequate in size. However, when atresia occurs earlier in gestation the right ventricle at
term is frequently coarsely trabeculated and pea sized. There is membranous pulmonary valvar atresia and, in
approximately one-third of the cases, associated infundibular hypoplasia or atresia. The tricuspid valve annulus
remains proportionately small and may be stenotic or incompetent. When the atresia occurs early in gestation
and the right ventricle is hypoplastic, fistulous tracts connecting the hypertensive right ventricular sinus and the
distal coronary arteries often persist, sometimes with proximal coronary artery stenosis. The right ventricle
without outflow usually generates supra-systemic pressures. The high pressure probably contributes to the
enlargement of fistulas that connect the right ventricle to the coronary arteries. The resultant severe coronary
hypertension in utero may cause the coronary arteriopathy, proximal coronary artery stenoses, and myocardial
fibrosis frequently seen. The pulmonary arteries are of adequate size and are perfused through a patent ductus
arteriosus or, rarely, through aortic-pulmonary collaterals. An unrestrictive interatrial communication is essential
to intrauterine survival and is present at birth. Pulmonary atresia with an intact ventricular septum is rarely
associated with other cardiovascular or somatic malformations.
Pathophysiology
The major hemodynamic consequence of pulmonary atresia with intact septum is the obligatory right-to-left
passage of the total systemic venous return through the foramen ovale to the left atrium. A patent ductus
arteriosus provides the only entrance to the pulmonary circulation. As it closes, pulmonary perfusion declines,
which results in progressively severe hypoxemia, metabolic acidosis, and death. In those infants with fistulas
from the right ventricle to the left anterior descending and/or right coronary arteries and proximal coronary
artery stenosis, perfusion of the coronary arteries distal to the stenoses occurs from the high pressure right
ventricle through the fistulas. Surgical establishment of continuity between the right ventricle and the pulmonary
artery may result in sufficient decrease in the right ventricle pressure to cause hypoperfusion of coronary beds
supplied solely by fistulas and produce myocardial infarction. The size of the infarct and its impact directly
correlate with the area of distribution of the coronary vessel involved (62).
Clinical Findings
Most infants with pulmonary atresia develop progressively severe cyanosis within the first week of life. As the
ductal arteriosus constricts, severe cyanosis, hypotension, bradycardia, hypotonia, and marked acidosis occur.
Signs of right-sided failure may develop but are usually absent. The precordium is quiet, and there is no thrill.
S2 is single, and there is often a pan-systolic murmur of tricuspid incompetence.
Chest radiography, generally shows normal or mildly enlarged heart size, reduced lung vascular markings, and a
left aortic arch (Fig. 33-25). ECG usually reveals a QRS axis in the frontal plane between zero and 80 degrees,
absent or diminished right ventricular forces, and a pattern of left ventricular dominance reflecting right
ventricular hypoplasia.
Figure 33-25 A chest radiograph shows decreased pulmonary vascularity and mild cardiomegaly in a 1-day-old
infant with pulmonary atresia and an intact ventricular septum.
Figure 33-26 Cineangiogram of the right ventricle (RV) in an infant who presented with severe cyanosis, no
murmur, single S2, and oligemic pulmonary vascular markings on chest radiograph. Echocardiogram confirmed
pulmonary valve atresia, intact interventricular septum, and a closing ductus arteriosus. The aortic origin of the
right coronary artery could not be visualized, and there was evidence for coronary fistula. Prostaglandin
infusion was started. Selective coronary arteriography and right ventriculography demonstrated atresia of the
proximal right coronary artery. A large fistula (FISTULA, black print) from the small right ventricle perfuses the
distal right coronary artery and, through collaterals, the circumflex (C). There was also fistula (FISTULA, white
print) from the right ventricle to the left anterior descending (LAD). The baby successfully underwent
implantation of a modified Blalock-Taussig shunt, with no attempt to connect the right ventricle to the
pulmonary artery.
P.668
Echocardiographic examination shows normal or somewhat enlarged left-sided structures, a small tricuspid valve
and right ventricle, and atresia of the pulmonary valve. If there is membranous atresia of the pulmonary valve,
the membrane can move like a critically stenotic valve, and the diagnosis cannot be made with certainty without
a careful Doppler flow examination. Limitations in the size of the tricuspid valve and right ventricle and the
presence of tricuspid stenosis can be quantified. High-resolution two-dimensional imaging and color Doppler can
detect fistulas between the right ventricle and the coronary arteries. After the presumptive clinical diagnosis is
made, prostaglandin should be given to dilate and maintain patency of the ductus arteriosus and increase
pulmonary blood flow. In those babies diagnosed after ductal constriction, there should be rapid improvement in
oxygenation and relief of acidosis.
Cardiac catheterization with selective right ventriculography and ascending aortography is done to determine the
presence of fistulas and stenosis in the coronary arteries, essential for planning surgery (Fig. 33-26).
Pulmonary atresia with an intact septum needs to be differentiated from other cardiac causes of severe cyanosis
in the first week of life (see Table 33-12). The combination of peaceful cyanosis, little murmur, single S2, chest
radiograph with normal heart size and diminished pulmonary markings, and ECG with R axis of 0 to 80 degrees
and diminished V1 R wave amplitude is characteristic. Critical valvar pulmonary stenosis alone, or with tetralogy
of Fallot, are accompanied by a systolic ejection murmur, and right ventricular hypertrophy. Tetralogy of Fallot
with pulmonary atresia has a continuous murmur and ECG with normal axis for age and prominent right
ventricular forces. Transposition of the great arteries with an intact ventricular septum causes peaceful cyanosis
and little murmur but is associated with a spilt S2, chest radiograph with cardiomegaly and increased vascular
markings, and ECG with normal axis for age and evidence of prominent right ventricular forces. Infants with
tricuspid atresia often have a prominent murmur and have a superior frontal plane axis on the ECG. In Ebstein
anomaly of the tricuspid valve, the ECG has tall wide P waves and an rsR8S8 pattern, and chest radiograph
demonstrates severe cardiac enlargement (Fig. 33-29). Echocardiographic examination differentiates all of these.
Treatment
After stabilization with prostaglandins, the treatment of choice is surgery, which is indicated as soon as feasible
after the anatomy is established by catheterization. For infants with an adequately sized right ventricle,
pulmonary valvotomy or valvectomy is performed. Some infants may require placement of a patch across the
right ventricular outflow tract and valve annulus for adequate relief of obstruction. Often, even when of
adequate size, the right ventricle is noncompliant, resulting in persistence of marked right to left atrial shunting.
In infants with severe right ventricular noncompliance or hypoplasia a systemic-to-pulmonary shunt also may be
necessary for relief of hypoxemia. After a period during which the right ventricular capacity and compliance
improve and the pulmonary vascular resistance decreases, these infants may no longer require shunts for
maintenance of adequate pulmonary blood flow. Even infants with initially diminutive right ventricular chambers
may sometimes demonstrate adequate growth of the chamber if flow through the ventricle is provided by
establishing continuity between the right ventricle and pulmonary artery, and the tricuspid valve is adequate.
The right ventricle cannot be decompressed in infants who have fistulas between the right ventricle and
coronary arteries and stenoses involving at least two separate proximal coronary vessels, without risk of fatal
infarction (62). Infants with two or more major coronary stenoses, and those with persistent severe right
ventricular hypoplasia, require a staged approach with an initial neonatal shunt and a later Fontan-type cavo-
pulmonary anastomosis or transplantation. Without therapy, the malformation is usually fatal, and with surgical
treatment, more than 80% to 95% of these newborns survive to 1 year of age.
P.669
Tricuspid Atresia
Tricuspid atresia is a relatively uncommon disease that is characterized by absence of the tricuspid valve. Except
in rare cases, no valve exists, and tricuspid agenesis is therefore a more precise description of this anomaly.
Pathophysiology
The entire systemic venous return (i.e., cardiac output) enters the right atrium and exits through the foramen
ovale to the left heart. The systemic and pulmonary venous streams mix in the left atrium. After passage to the
left ventricle, the cardiac output passes to the aorta, and a variable amount gains access to the pulmonary
artery through a ventricular septal defect, diminutive right ventricle, and a variable degree of pulmonary
stenosis. Flow to the pulmonary artery is limited by the size of the ventricular defect and the amount of
infundibular and valvar pulmonary stenosis. The level of cyanosis is determined by the amount of pulmonary
blood flow. If no ventricular septal defect is present there is also pulmonary valve atresia and like infants with
isolated pulmonary atresia, all pulmonary flow is dependent on the ductus arteriosus. If the ventricular septal
defect and right ventricular outflow are small and the infant is very cyanotic in early infancy, the pulmonary
blood flow can be increased by maintaining a patent ductus arteriosus and subsequently surgically implanting of
a Blalock-Taussig shunt. Some infants have naturally balanced circulations with sufficient flow to the pulmonary
arteries to allow an adequate arterial oxygen saturation of 75% to 88%, although not so much regarding cause
pulmonary hypertension or congestive heart failure. Infants with a large ventricular septal defect and
unobstructed right ventricular outflow are minimally cyanotic and develop congestive heart failure and
pulmonary artery hypertension. Occasionally, the great vessels are transposed, with the aorta arising from the
right ventricle, and the systemic output may be limited by the size of the ventricular defect. A left superior vena
cava is a common associated anomaly of importance to later surgery.
Figure 33-27 Electrocardiogram of a cyanotic newborn with tricuspid atresia shows the characteristic left axis
deviation and low right precordial forces.
Clinical Findings
Babies are usually discovered to have tricuspid atresia during the first days or weeks of life because of cyanosis.
Some have a harsh mixed frequency systolic murmur of left to right flow through the ventricular septal defect or
ductus arteriosus, and/or a harsh high-pitched systolic ejection murmur from pulmonary stenosis. The S2 is
most often single.
ECG characteristically has features distinguishing it from most other cyanotic lesions. There is a leftward
superior axis similar to endocardial cushion defects, but usually with diminished right precordial forces (see Fig.
33-27 and 33-10).
On chest radiograph, the heart is normal size or minimally enlarged and the pulmonary vasculature is
diminished when the ventricular septal defect is restrictive, although heart size and pulmonary vascularity are
increased when the ventricular septal defect is large and unrestrictive.
Figure 33-28 Echocardiographic apical four-chamber view in a baby with tricuspid atresia. Systemic venous
blood flows from the right atrium (RA) across a widely patent foramen ovale (double arrows) to the left atrium
(LA) and into the left ventricle (LV). Some of the outflow of the left ventricle passes through a ventricular septal
defect (single arrow) and hypoplastic right ventricle (RV) to the pulmonary arteries (not shown).
P.670
The diagnosis is readily confirmed by echocardiographic identification of a diminutive right ventricle, absent
tricuspid valve, and right-to-left flow through the foramen ovale (Fig. 33-28). The size of the ventricular septal
defect and the degree of sub-valvar and valvar pulmonary stenosis can be accurately determined.
Treatment
Infants with small or no ventricular septal defect and severe obstruction of blood flow through the right ventricle
may require infusion of prostaglandin E1 before palliative surgery. Surgical implantation of a modified or classic
Blalock-Taussig type shunt from a subclavian or innominate artery to a pulmonary artery provides the means for
adequate arterial oxygenation, survival, and growth in those infants with severe intracardiac obstruction to
pulmonary blood flow. After several months of age the pulmonary vascular resistance usually decreases
sufficiently to permit success of bidirectional Glenn surgery, connecting the upper superior vena cava carrying
systemic venous blood return from the upper half of the body directly into the pulmonary artery. Variations on
the Fontan operation, in which the systemic venous return from the remainder of the body is directed into the
pulmonary arteries, are undertaken at or beyond the age of 1 year (63). More than 75% of infants with tricuspid
atresia survive with Fontan physiology (64).
Rarely, in an infant with a large ventricular defect and no pulmonary stenosis, the pulmonary blood flow may be
excessive enough to cause congestive heart failure. Anticongestive medications are usually sufficient to allow
growth, although a pulmonary artery banding or other procedure may be necessary to diminish pulmonary
vascular pressure and resistance.
Ebstein's Anomaly of the Tricuspid Valve

Ebstein's anomaly of the tricuspid valve is encountered rarely in the newborn period. The septal leaflet of the
tricuspid valve is displaced downward and adheres to the ventricular septum to various degrees. The result is a
dysfunction of the tricuspid valve with regurgitation, and sometimes stenosis, to varying degrees. The
dysfunction is compounded by the fact that the right atrium and right ventricle contract at different times, and
there is an area of atrialized ventricle or ventricularized atrium, depending on the point of view. This discordant
pumping of parts of each chamber contributes to the dysfunction. Those with severe prenatal tricuspid
regurgitation often have massive cardiomegaly and may have pulmonary hypoplasia that is rapidly fatal after
birth (Fig. 33-4). The effective right ventricular volume is reduced, and there is limited passage of blood through
the right ventricle. Some right atrial blood courses through the patent foramen ovale, causing cyanosis. The
severity of the defect can be described by the degree of cyanosis that, in the newborn period, can be severe
because of the concomitant unresolved elevation of the pulmonary vascular resistance left over from fetal life.
As the newborn's pulmonary vascular resistance regresses, the cyanosis often improves, sometimes markedly,
although babies with severe regurgitation and pulmonary hypoplasia have a high mortality rate. After surviving
the newborn period, the infant has a course determined by the degree of abnormality; some patients survive
into late adulthood without important limitation, but others remain cyanotic and prone to supraventricular
tachycardia.
Ebstein's disease is recognized because of minimal or prominent systolic low-pitched murmur of tricuspid
regurgitation, multiple clicks, and minimal to severe cyanosis.
Chest radiography demonstrates cardiomegaly that may vary from minimal to some of the largest hearts
encountered in the newborn period (see Fig. 33-29).
ECG has tall wide P waves and an rsR'S' pattern (see Table 33-12).
Figure 33-29 A chest radiograph of a 1-day-old cyanotic infant with Ebstein anomaly of the tricuspid valve
shows marked cardiomegaly.
P.671
Echocardiographic examination confirms the diagnosis and delineates the extent and precise nature of the
tricuspid anomaly, the amount of tricuspid regurgitation, the effective right ventricular volume, the atrial septal
defect and associated abnormalities. Treatment is usually supportive. A surgical Blalock-Taussig shunt may be
required in infants who remain severely cyanotic. Treatment for episodic supraventricular tachycardia may be
needed.
Anomalies with Cyanosis as a Result of Complete Mixing of

Systemic and Pulmonary Circulations
Hypoplastic Left Heart Syndrome
The hypoplastic left heart syndrome encompasses a variety of specific cardiovascular malformations producing
similar hemodynamic and clinical manifestations, including aortic atresia, mitral atresia, premature closure of
the foramen ovale, hypoplastic left ventricle with critical mitral stenosis, and aortic stenosis. Some cases of
severe complex coarctation, critical aortic valve stenosis, and malaligned atrioventricular canal defects are also
included in this category. The left heart chamber is usually very small, and endocardial fibroelastosis in it is
common. Hypoplastic left heart syndrome occurs in 10.2% of infants with serious heart disease and is one of the
most common lesions presenting in the first week of life (see Tables 33-1, 33-2, and 33-7). It is less common in
very premature infants (<1.85 kg). It is usually an isolated lesion, although it has been described in association
with trisomy syndromes and in infants of diabetic mothers. Familial cases occur.
Pathophysiology
Obstruction or atresia of the mitral or aortic valves limits or prevents flow through the left heart. The systemic
venous return enters the right heart and is ejected into the pulmonary artery. The systemic circulation is largely
or totally supplied by right-to-left flow through the ductus arteriosus. Blood traversing the lung enters the left
atrium, flows through an atrial septal defect or dilated foramen ovale, and returns to the right atrium to join the
incoming systemic venous return. Complete mixing takes place in the right atrium, with similar oxygen
saturation measured in the right ventricle, pulmonary artery, and aorta. With little or no egress through the left
heart, pulmonary flow must pass left to right through an interatrial communication. Any limitation of flow
through the atrial septum produces pulmonary venous hypertension.
The maintenance of adequate systemic circulation requires patency of the ductus arteriosus. In aortic atresia,
the ascending aorta, brachiocephalic vessels, and coronary arteries are perfused in a retrograde fashion with
blood originating from the patent ductus arteriosus. Spontaneous constriction of the ductus results in flooding of
the pulmonary circulation simultaneous with low systemic blood flow, poor coronary perfusion, congestive
failure, and shock with metabolic acidosis, electrolyte imbalance, and coagulation abnormalities. Closure of the
ductus stops blood flow to the body and causes immediate death.
Clinical Findings
These infants usually become symptomatic within the first week of life. Congestive failure and a shock-like
picture may develop precipitously. The baby becomes ashen gray with poor peripheral perfusion, and all pulses
are weak. Ductal constriction or flow may appear intermittent, with femoral pulses intermittently palpable.
Symptoms and signs of congestive failure are associated with hypotension and, terminally, with bradycardia. S2
is single, and a gallop may be heard.
Chest radiograph shows cardiac enlargement and pulmonary plethora.
ECG can be near-normal, but usually has abnormalities including markedly diminished or absent left ventricular
forces, left precordial T wave flattening or inversion, right atrial enlargement, right ventricular hypertrophy, and
sometimes right axis deviation (see Fig. 33-30).
Echocardiographic examination demonstrates a small or tiny left ventricle. The ascending aorta is small with
retrograde flow in cases of aortic atresia, and there is frequently aortic arch hypoplasia and juxta-ductal
coarctation (Fig. 33-31).
The clinical picture of the hypoplastic left heart syndrome may be simulated by respiratory distress syndrome,
interrupted aortic arch, severe complex coarctation, early neonatal myocarditis, isolated critical valvar aortic
stenosis, sepsis or some inherited metabolic disorders (see Table 33-12).
Treatment
Without surgery, the mortality rate is 98% by 1 year of age. Surgical therapy consists of converting the
circulation to single-ventricle Fontan-type physiology and/or to transplant a new heart. Previously the survival
with the first stage of surgical palliation or with attempt to neonatal transplantation was only 1 in 2, or less, and
comfort care was often presented as an option. Outcomes have significantly improved; currently survival with
stage 1 surgery is 80% to 90% at specialized surgical centers.
Preoperatively the infant must be stabilized with prostaglandin E1, inotropic agents, volume infusion, and
bicarbonate. Hyperventilation (i.e., arterial partial pressure of carbon dioxide, PaCO2 < 40 mm Hg) and
unnecessary supplemental oxygen administration are avoided. These measures decrease pulmonary vascular
resistance and increase preferential flow of the right ventricular output into the pulmonary vascular bed instead
of across the ductus to the systemic vasculature, worsening the shock. Often for preoperative stabilization,
mechanical hypoventilation with muscle relaxation to maintain the PaCO2 45 to
P.672
55 mm Hg, and sometimes FiO2 less than 0.21 to maintain arterial O2 saturation 70% to 75%, are required to
elicit sufficient pulmonary vasoconstriction to induce adequate systemic blood flow.
Figure 33-30 Electrocardiogram shows low left precordial forces in a cyanotic newborn with hypoplastic left
heart syndrome.
Palliative surgical therapy consists of staged surgical procedures that convert the circulation to a systemic-rightventricle perfusing the aorta, with pulmonary arterial Fontan-type physiology. The first-stage, modified Norwood
procedure connects the right ventricle to the aorta, reconstructs the hypoplastic aortic arch, and provides an
interim source of pulmonary blood flow with arterial pressure sufficient to overcome physiologically relatively
elevated neonatal pulmonary vascular resistance. Cardiopulmonary bypass is used during cooling of the infant's
body temperature to 15 to 18 degrees centigrade and usually temporarily discontinued (i.e., complete deep
hypothermic circulatory arrest), the ductus arteriosus is ligated, the obliquely transected main pulmonary artery
is anastomosed to the underside of the aortic arch and further augmented with homograft patch, and an atrial
septectomy done to allow unimpeded pulmonary venous return to the right atrium. Following re-establishment
of cardiopulmonary bypass, blood flow to the pulmonary artery is provided by insertion of a modified BlalockTaussig, or direct central systemic-to-pulmonary artery shunt, or anastomosis of a small conduit from a superior
right ventriculotomy (65).
Figure 33-31 Echocardiographic parasternal long-axis view in an infant with hypoplastic left heart syndrome.
The right ventricle (RV) is near normal in size. The left atrium (LA), left ventricle (LV) and ascending aorta (AO)
are tiny. Both the mitral and aortic valves are atretic.
Several months later, when the pulmonary resistance has dropped from relatively high neonatal values,
pulmonary circulation can be supplied relatively directly with systemic venous return. A second-stage
bidirectional Glenn procedure is done. The systemic venous return from the upper body is placed directly into
the lung through a anastomosis of the upper superior vena cava to the right pulmonary artery, and the systemicarterial-to-pulmonary-arterial shunt (with its obligate ventricular volume overload) is taken down. Finally, at a
third Fontan-type procedure, systemic venous return from the lower body is also diverted directly into the
pulmonary artery through a lateral tunnel within or adjacent to the right atrium, directing flow to a second,
lower anastomosis of the lower superior vena cava to the

P.673
pulmonary artery (65). The perinatal and first-stage surgical survival is 85-90%, and with the second and third
stage procedures are greater than 98% and 95% to 99%, respectively (66,67). Adult survival is not yet
determined.
Cardiac transplantation has been used as an alternative approach with good survival and childhood functional
capacity, but there are critical limitations in the timely availability of neonatal donors and long-term morbidities
and risks (68,69 and 70).
Neither reconstructive surgery leading to a Fontan procedure nor cardiac transplantation can be viewed as
curative. Both methods of treatment have high fiscal and emotional costs. However, many survivors eventually
do remarkably well in childhood.
Total Anomalous Pulmonary Veins

In the event of failure to connect the common pulmonary vein to the left atrium in the embryo, communications
are established with available systemic venous channels that then drain the pulmonary veins. Anatomically
abnormal drainage may be supra-cardiac (i.e., into the right or left superior vena cava), intracardiac (i.e., into
the coronary sinus, right atrium), or sub-diaphragmatic (i.e., through the inferior vena cava or porta hepatis).
Mixed sites of drainage occur in approximately 10% of these patients. A patent foramen ovale or atrial septal
defect is invariably present, permitting venous return to the left heart. Anomalous pulmonary venous return is
often associated with heterotaxy.
Although isolated total anomalous venous return accounts for only 2% of newborns with serious cardiac disease
(see Table 33-2), it is an important lesion because it is potentially curable and often misdiagnosed as pulmonary
disease.
Figure 33-32 A: A chest radiograph of a 3-day-old term infant with obstructed total anomalous pulmonary
venous drainage shows a ground-glass appearance of the lungs with normal heart size, similar to the
radiographic appearance of respiratory distress syndrome in preterm neonates. B: Postmortem angiography
shows obstruction of the common pulmonary venous channel below the diaphragm (arrow). C:
Echocardiographic supra-sternal notch view of another neonate with similar anatomy demonstrated the left
lobar pulmonary veins (>) and right lobar pulmonary veins (<) connecting to retro-cardiac posterior
confluence, that drained via a vertical vein inferiorly across the diaphragm. Additional images not shown
demonstrated intrahepatic connection of the vertical vein through a constricted ductus venosus to the inferior
vena cava. The baby did well after emergent surgical anastomosis of the pulmonary venous confluence to the
posterior left atrium, ligation of the vertical vein inferiorly, and closure of the formaen ovale.
Pathophysiology
Infants with anomalous pulmonary venous drainage can be divided into two major categories on the basis of the
hemodynamic changes produced: those with unobstructed veins and those with obstructed veins. Unobstructed
pulmonary veins entering the systemic venous circulation or directly into the right heart result in a large left-toright shunt, congestive heart failure, and pulmonary artery hypertension. Systemic output is maintained through
right-to-left flow across an interatrial communication. Despite the obligatory right-to-left shunt through the
atrium, the large pulmonary blood flow mixing with the systemic venous return at the right atrium allows a
reasonable peripheral oxygen tension and produces only mild or moderate cyanosis.
If pulmonary venous return is obstructed, the circulatory effects are drastically different. The obstruction may
take the form of increased resistance to flow produced by a long, common, pulmonary venous channel or
localized intrinsic or extrinsic obstruction. Sub-diaphragmatic anomalous pulmonary venous return is almost
always obstructed by constriction of the ductus venosus, obstructing flow into the inferior vena cava (Fig. 3332). Obstruction to supra-cardiac pulmonary venous return may occur because of compression of the common
pulmonary venous channel between the left primary bronchus and left pulmonary artery or because of narrowing
at the entry of the common pulmonary vein into the right superior vena cava. Obstruction at the foramen ovale
is uncommon. After birth, significant resistance to flow through the pulmonary veins becomes evident, causing
pulmonary venous hypertension, pulmonary edema, marked pulmonary artery hypertension and diminished
flow, and severe cyanosis. The arterial
P.674
oxygen tension is low because pulmonary blood flow is markedly reduced, and the relative contribution of fully
oxygenated blood to the venous return to the heart mixing in the right atrium is less than in the unobstructed
form of this disorder.
Clinical Findings
Infants with total anomalous pulmonary venous return without significant obstruction usually become
symptomatic after the neonatal period, when the pulmonary vascular resistance decreases and a large left-toright shunt and congestive heart failure develop. They are mildly cyanotic with increased respiratory rate and
work, often have frank congestive heart failure, and have a large heart revealed on chest radiographs.
Infants with obstructed pulmonary venous return are usually critically ill, severely cyanotic, and tachypneic
within the first week of life. There are congestive heart failure and poor peripheral perfusion. In the absence of
associated anomalies such as malposition with pulmonary valve or sub-valvar stenosis, there is generally only a
relatively soft murmur of a patent ductus arteriosus and/or tricuspid valve regurgitation.
As seen on the chest radiograph, heart size is often normal, and there is evidence of pulmonary edema (Fig. 3332). The clinical and the radiographic pictures may resemble hyaline membrane disease or diffuse pneumonia
complicated by persistent pulmonary hypertension.
ECG at birth may be normal or may show right axis deviation, right atrial hypertrophy, and right ventricular
hypertrophy.
Echocardiographic findings include absence of pulmonary venous connections to the left atrium, right-to-left
bulging of the atrial septum, right-to-left atrial shunt, and pulmonary venous confluence posterior to the left
atrium connecting to a systemic venous channel.
The findings of severe cyanosis, little murmur, and a roentgenographic picture of a normal heart size associated
with pulmonary edema are characteristic (see Fig. 33-28A and Table 33-11).
Respiratory distress syndrome and interstitial pneumonia can be clinically indistinguishable from obstructed total
anomalous pulmonary venous connection. Any suggestion of an atypical course mandates echocardiography,
particularly in term neonates with equally diffuse involvement of both lungs, a failed or inconclusive hyperoxia
test, or a murmur. Two-dimensional echocardiography shows the anomalous common venous connections and
obstructions and the presence or absence of other cardiac anomaly.
Treatment
The treatment of total anomalous pulmonary veins is surgical, and success is related to the anatomy (e.g.,
results are poorer in the mixed variety) and to the age of onset of symptoms, as in patients with the infradiaphragmatic type. The success rate is better for infants with intracardiac drainage. Using cardiopulmonary
bypass and deep hypothermic circulatory arrest, continuity or redirection of the pulmonary venous drainage into
the left atrium is established, usually by anastomosis of the retro-atrial common pulmonary venous confluence
to a parallel posterior left atriotomy. Although inotropic agents, diuretics, and supportive medical treatment may
help temporarily and partially with stabilization, infants with severely obstructed anomalous venous return
require immediate surgical intervention. Prostaglandin E1 therapy is not beneficial and can lead to dramatic
worsening of the pulmonary edema. In most patients, surgery can be done on the basis of information from
echocardiography without the delays involved with cardiac catheterization or magnetic resonance imaging. If
there are associated complex congenital anomalies (e.g., heterotaxy syndrome) or intrinsic pulmonary vein
stenosis is suspected, preoperative catheterization may sometimes help determine the best management. Those
with unobstructed pulmonary veins and congestive heart failure with relatively normal pulmonary artery
pressure may be improved with medical treatment, and corrective surgery may be delayed for a few weeks.
Among those without additional lesions, the postoperative survivors have an excellent prognosis for a relatively
normal life. A few infants develop atrial dysrhythmias postoperatively. Because life threatening, progressive and
recurring pulmonary vein obstruction develops in approximately 5% to 10% of infants 1 to 12 months
postoperatively, particularly in those with associated heterotaxy syndrome and preoperative pulmonary vein
stenosis, initial close follow-up is mandatory.
Truncus Arteriosus
Failure of the cono-truncus to septate into the aorta and main pulmonary artery results in the clinical problem
described as truncus arteriosus. It is often associated with microdeletions in a critical region of chromosome
22q11 containing genes responsible for conotruncal development, and with related extra-cardiac anomalies (e.
g., DiGeorge syndrome). Other factors can be responsible (28,29). The only artery arising from the heart is the
common truncus arteriosus. There is one semilunar valve that may have extra valve leaflets (e.g., four or five)
and is often incompetent and rarely stenotic. The pulmonary arteries arise from the left anterior aspect of the
truncal root as a single main pulmonary artery, at their bifurcation, or with separate right and left branches.
Classifications in vogue are based on the level at which the pulmonary vessels take off, but they are not
especially pertinent to the physiology and the clinical picture. Almost universally, there is a ventricular septal
defect, usually in the sub-aortic septum, similar to that seen in tetralogy of Fallot. Interrupted aortic arch is an
infrequent associated anomaly that should be borne in mind for timely recognition and management.
P.675
Pathophysiology
Because of the large ventricular defect and the common arterial trunk, the systemic and pulmonary venous
returns are mixed, and the patient is cyanotic. The degree of cyanosis is determined by the pulmonary flow,
which is a function of obstruction in the proximal pulmonary arteries. These obstructions are common, rarely
severe, and located at the junction of the pulmonary artery and the trunk. If there is no obstruction, which is
likely, pulmonary flow exceeds systemic output several-fold, obligating a high-output state and resulting in
congestive heart failure and poor survival without surgery. Without surgery, irreversible pulmonary vascular
disease is likely to develop as early as the patient's first birthday. Congestive heart failure is less of a problem if
there is branch pulmonary artery stenosis, although the degree of cyanosis is greater. Proximal pulmonary
artery obstructions frequently develop after surgery. When there is also interrupted aortic arch, a ductus is
present, and its untreated closure results in lower body hypotension and hypo-perfusion, followed by death.
Clinical Findings
Infants with truncus arteriosus resemble those with ventricular defect more than the other cyanotic defects.
Except cyanosis, which may be mild, development of symptoms is delayed until the pulmonary vascular
resistance has resolved enough to allow a large pulmonary flow and the features of a left-to-right shunt.
Tachypnea and the other signs of congestion predominate, although cyanosis may be recognized and
documented in the first days of life. There is usually a murmur that sounds like a ventricular defect, the
peripheral pulses are bounding, and other signs of an aortic runoff are present. The S2 is loud and single, and
systolic clicks may be heard. If interrupted aortic arch is also present, diminished pulses, pulse pressure in the
lower body, azotemia, and metabolic acidosis develop as the ductus constricts
On chest radiographs, the heart is enlarged, the pulmonary vasculature is engorged, and the aortic arch may be
rightward (33%).
ECG inexplicably varies, showing right, left, or combined ventricular hypertrophy (see Table 33-11).
Echocardiographic examination demonstrates the anatomy (Fig. 33-33).
Treatment
Anticongestive measures to control congestion and promote growth have limited effect. Surgical correction is
usually undertaken within the first 1 to 6 weeks of life. The later the surgery the more likely that the early
postoperative course and survival will be threatened by severe elevation in pulmonary vascular resistance.
Although banding of the pulmonary arteries on both sides is possible and is sometimes the only choice, it is easy
to understand the difficulty of applying bands on each side equally. For this reason, and because of empirically
poor results, most centers perform one-stage repair in infancy (71). This surgery consists of separating the
pulmonary arteries from the trunk, establishing a conduit from the right ventricle to the pulmonary arteries, and
closing the ventricular defect. Postoperative management and outcome are aided by the early use of potent
pulmonary vasodilators such as nitric oxide. The long-range outcome of this type of repair involves two or three
surgical revisions with larger pulmonary artery conduits to accommodate growth during childhood. Severe
compression of the conduits and natural or surgery-related proximal pulmonary artery stenoses can develop and
require catheter-based interventions in early childhood. The truncal semilunar valve may rarely be sufficiently
incompetent to influence the outcome. However, with staged surgical intervention, the long-term survival and
outcome are good (72).
Figure 33-33 Echocardiographic subxiphoid view in a neonate with truncus arteriosus. The right ventricle (RV)
and left ventricle (LV) both pump blood through a subaortic ventricular septal defect (arrow) and common
truncal valve into the ascending aorta (AO). The main pulmonary artery (PA) arises from the side of the
ascending aorta.
Anomalies with Cyanosis with Variable and Mixed Anatomy and

Physiology
Single Ventricle
There are few patients who have only one ventricle, with either two atrioventricular valves or a common
atrioventricular valve. Most patients characterized as having a single ventricle have one dominant ventricle and
a second diminutive structure that is often characterized as an outflow chamber. Because of the similarity of this
syndrome to tricuspid atresia, some physicians have preferred the term univentricular heart to describe all
patients with a functionally single ventricle, whether they have two atrioventricular valves, tricuspid atresia, or
mitral atresia.
Seventy percent of these patients have a single left ventricle with a rudimentary anatomic right ventricular
P.676
outflow chamber that is left sided (l-transposed position). The ventricular outflow chamber leads to an anterior
and leftward aorta, similar to that seen in corrected transposition. Most of the remainder of patients have a
single left ventricle with a rudimentary right ventricular outflow chamber that is right-sided (d-transposed
position). Pulmonary stenosis coexists in 50% of these patients. In those without pulmonary stenosis, sub-aortic
stenosis and coarctation can occur. Almost any other cardiac anomaly may be associated with single ventricle.
Single right ventricles also occur, particularly with heterotaxy syndrome.
Pathophysiology
Depending on the presence or absence of pulmonary stenosis, the clinical picture may be dominated by
diminished pulmonary flow and cyanosis or by excessive pulmonary flow and high-output congestive heart
failure, respectively. Occasionally the pulmonary and systemic flow are balanced, and the patient is only mildly
cyanotic and otherwise asymptomatic. The problems peculiar to corrected transposition, such as the tendency to
develop complete heart block or develop an incompetent atrioventricular valve (usually the left), are risks. The
connection between the single ventricle and the outflow chamber (i.e., ventricular defect) tends to get smaller
with time in approximately 50% of patients. This has the physiologic effect of sub-aortic stenosis and must be
considered in any management program.
Clinical Findings
The patient usually is visibly cyanotic during the neonatal period. Sometimes those with excessive pulmonary
blood flow, minimal cyanosis, and congestive heart failure present later because of growth failure or tachypnea.
Most have systolic murmurs from pulmonary stenosis, atrioventricular valve regurgitation, or from other
associated defects. The detailed diagnosis is obtained by echocardiographic examination. Catheterization or
magnetic resonance imaging are sometimes used preoperatively to delineate anatomic or physiologic details that
may influence surgical management.
Treatment
The ultimate goal of management is to separate the pulmonary and systemic circulations by directing all
systemic venous return through a cava-pulmonary or atrial-pulmonary anastomosis (e.g., Fontan procedure)
and utilizing the single ventricle solely for pumping oxygenated blood to the aorta. Rare patients who have
perfectly balanced pulmonary and systemic circulations, and are only mildly cyanotic and virtually
asymptomatic, fare well with no surgery for years. Most require a pulmonary artery band to limit pulmonary flow
or a shunt procedure to increase pulmonary flow and arterial oxygenation until a modified Fontan procedure (i.
e., connection of systemic venous return to the pulmonary arteries) can be performed.
Double-Outlet Right Ventricle

Double-outlet right ventricle is a rare, anatomically and physiologically heterogeneous group of anomalies. The
diagnosis is applied when imaging studies demonstrate that both great arteries arise principally from the
morphologic right ventricle, invariably in the presence of a ventricular septal defect. A muscular conus
commonly underlies both of the great artery outflows. Although various types of double-outlet right ventricle
share a common ventricular-arterial relationship, the anatomy of the ventricles, ventricular conotruncal
outflows, the relationship of the great arteries, and the presence of additional anomalies are variable. Of primary
importance is the relationship of the great arteries to each other and, as a consequence, to the ventricular septal
defect. This determines the basic physiology and the surgical management options (73).
Pathophysiology and Clinical Findings

The most common great artery arrangement is a normal relationship, with the aortic valve posterior and
rightward and the pulmonary valve leftward and anterior. The ventricular septal defect usually relates to the subaortic outflow. Many will have sub-valvar and valvar pulmonary stenosis. Depending on the presence or absence
of pulmonary stenosis, the physiology, clinical findings, clinical course, and therapeutic approach are,
respectively, like those with tetralogy of Fallot with diminished pulmonary flow or like those with an isolated
large sub-aortic ventricular septal defect and pulmonary hyperemia and hypertension.
The second most frequent great artery relationship in double-outlet right ventricle is with the aortic valve to the
right, and anterior to or alongside the pulmonary valve, and the great arteries side-by-side. The ventricular
septal defect is below the pulmonary outflow. The hemodynamics, clinical findings, and course are as in
transposition.
Less common types of double-outlet right ventricle include those with the aorta anterior and leftward, those with
a remote uncommitted ventricular septal defect that does relate to either great artery outflow (e.g.,
atrioventricular inlet type defect), and those with a doubly committed ventricular septal defect that relates to
both great artery outflows. Important associated anomalies in those without pulmonary stenosis, with either
normally related or side-by-side great arteries, include mitral stenosis or atresia, sub-aortic outflow obstruction,
and aortic arch hypoplasia and coarctation. Other less common associated anomalies include heterotaxy with its
many related defects (see below and Table 33-13), pulmonary or aortic atresia, multiple ventricular septal
defects, and superior-inferior ventricles. These associated anomalies can profoundly influence the
hemodynamics, clinical course, and management.
Echocardiographic examination determines the diagnosis by showing the right ventricular-great arteries
relationship and can usually determine the great artery and ventricular septal defect relationships and delineate
the various associated defects. The need and timing of
P.677
catheterization depend on the anatomy demonstrated by echocardiography and the therapeutic management.
TABLE 33-13 COMMON ABNORMALITIES IN ASPLENIA ANDPOLYSPLENIA SYNDROMES
Asplenia
Isolated levocardia or dextrocardia

Visceral isomerism (midline liver and stomach)
Polysplenia
Transposition of the great arteries

Double-outlet right ventricle
Dextrocardia or levocardia
Absent inferior vena cava (renal to hepatic
segment)
Total anomalous pulmonary venous return
Total anomalous pulmonary venous return

Coronary sinus rhythm

Bilateral epiarterial bronchi
Pulmonary atresia or pulmonary stenosis

Single ventricle
Bilateral superior vena cava
Absent coronary sinus
Ipsilateral inferior vena cava and abdominal aorta
P-wave axis of atrial inversion
Bilateral eparterial bronchi
Bilateral trilobed lung
Treatment
Management depends on the specific anatomic abnormalities. Medical and surgical treatment in those with
normally related great arteries and no pulmonary stenosis is as with a large ventricular septal defect and in
those with pulmonary stenosis as with tetralogy of Fallot. Those infants with side-by-side great arteries and
transposition-like physiology may be managed by intracardiac repair or arterial switch procedure. Associated
anomalies may require other approaches including systemic-to-pulmonary arterial shunts, Norwood-like
procedure, aortic arch repair, and later Fontan-like approach.
l-Transposition of the Great Arteries

In l-transposition of the great arteries with situs solitus, also called corrected transposition, the circulation is
often physiologically corrected. The terminology refers to the position of the aorta, which is abnormally
positioned, anterior and usually to the left of the pulmonary artery, and to the position of the ventricles. lTransposition of the great arteries is most commonly associated with ventricular inversion as well, i.e., the
ventricles undergo l-looping instead of the usual d-looping. The systemic venous blood enters the right atrium
and flows into a right-sided, morphologically left, ventricle and out to the pulmonary artery. Pulmonary venous
blood returns to the left atrium and by way of the tricuspid valve into the left-sided, morphologically right,
ventricle to the aorta. The aorta is abnormally positioned, anterior and usually to the left of the pulmonary
artery. The hemodynamic changes in patients with l-transposition of the great arteries are caused by the
commonly associated cardiac abnormalities that include ventricular septal defect (50%), single ventricle (42%),
and pulmonary stenosis or atresia (45%). The latter three result in cyanosis. Left atrioventricular valve
regurgitation (23%), conduction disturbances, and arrhythmias, particularly supraventricular tachycardia and
complete heart block, are common. This diagnosis should be considered for newborns with complete heart block
or supraventricular tachycardia. Medical management and surgery are directed toward correction or palliation of
the associated cardiovascular malformations, such as closure of the ventricular septal defect, pulmonary
valvotomy, and a cardiac pacemaker if needed.
Malpositions
The term cardiac malposition describes abnormal position of the heart within the thorax or relative to the
abdominal viscera. Cardiac position, or malposition, is relatively independent of the intracardiac segmental
anatomy or interrelationships. For instance, dextrocardia, positioning of the heart in the right chest, and
mesocardia, midline cardiac positioning, may occur with normal positioning of the abdominal viscera and atrial
situs as a result of displacement from pulmonary disease or diaphragmatic hernia. Dextrocardia may also occur
from genetically orchestrated complete inversion of thoracic and abdominal sidedness (i.e., situs inversus
totalis), and with heterotaxic, ambiguous malpositioning of the viscera resulting from loss of control of
sidedness during embryonic development, (e.g., asplenia and polysplenia syndromes). Cardiac malposition is
also present when the heart is in the left thorax (levocardia) when there is heterotaxy or discordant sidedness of
the viscera. The term malposition also describes misplacement of the heart as in ectopia cordis.
Fixed cardiac malposition most commonly occurs with heterotaxic malposition of the viscera, as occurs with the
P.678
asplenia and polysplenia syndromes (Fig. 33-34), and is characterized by discordant sidedness (e.g.,
dextrocardia with rightward or midline liver, or levocardia with leftward or midline liver). If the heart is displaced
into the right chest or if there is total situs inversus, there is no heterotaxy. Situs inversus totalis is rare, and
the heart may be anatomically normal, unlike the more common forms of malposition.
Pragmatically, finding cardiac or abdominal malposition is important as evidence that the patient has a high
likelihood of complex combinations of congenital cardiac anomalies. Heterotaxy may be discovered by physical
examination or chest radiographs because of abnormal sidedness of the abdominal contents or because the
heart is located in the right chest. The cardiac anomalies are remarkably variable, often multiple and complex,
and often life-threatening. Table 33-13 shows the anatomic abnormalities encountered in asplenic and
polysplenic patients. Incredibly, often all of these occur in combination. For example, a baby with asplenia may
have the combination of inferior vena cava crossing from one side of the midline to the other, single left or
bilateral superior vena cava, dextrocardia, total and sometimes mixed pulmonary venous connections, single
atrium, common atrioventricular valve, single right ventricle with double outlet, and pulmonary stenosis or
atresia. Clinically, the initial objective is to determine the details of anatomy and anomaly in each central
systemic and pulmonary vein, cardiac chamber, valve, and great artery on a segment-by-segment basis,
delineate the interconnections of the various segments of the heart and plan the procedure or sequence of
procedures required to repair or optimally palliate the baby's heart. When evidence of cardiac or abdominal
malposition are detected, and regardless of the clinical appearance of the baby, it is safest to promptly proceed
with a detailed diagnostic evaluation by a cardiac team that has the facilities and experience to accurately
determine the anomalies and management. Most patients can be palliated. The overall mortality rate tends to be
higher because of the combination of lesions, particularly in those with obstruction of anomalous pulmonary
veins (see Table 33-2). Those with asplenia should receive lifetime prophylactic antibiotics because of their
propensity for sepsis.
Figure 33-34 A chest radiograph of a neonate with asplenia syndrome, tricuspid atresia, transposition of the
great arteries, and a right aortic arch. The liver and stomach are on the right side.
Acyanotic Lesions
Acyanotic diseases associated with normal pulmonary flow include those with systemic outflow obstruction such
as coarctation of the aorta and aortic stenosis, mild and moderate pulmonary outflow stenosis, myocardial
diseases, and arrhythmias. Acyanotic lesions usually associated with increased pulmonary blood flow include
ventricular septal defect, atrial septal defect, endocardial cushion defects, patent ductus arteriosus,
aortopulmonary window, and arteriovenous malformations (see Tables 33-14 and 33-15).
Acyanotic Anomalies with Systemic Outflow Obstruction

Coarctation and Interruption of the Aorta
For clinical, prognostic, and probably etiologic reasons, coarctation of the aorta is best considered in two
separate categories: simple and complex. Simple coarctation is usually a discrete constriction of the aortic
isthmus area, occasionally associated with a patent ductus arteriosus inserting just at or below it. Complex
coarctation involves tubular hypoplasia of the aortic arch, with or without discrete aortic narrowing and one or
more of the following lesions: patent ductus arteriosus, ventricular septal defect, endocardial cushion defect,
aortic stenosis, sub-aortic stenosis, mitral stenosis or regurgitation, hypoplasia of the left ventricle and
ascending aorta, other cyanotic anomalies, and endocardial fibroelastosis. In its most severe form the aortic
arch may be atretic and completely interrupted as in DiGeorge syndrome. In complex coarctation and
interruption of the aorta, the amalgam of left-sided involvement may be secondary to reduced intrauterine flow
through the left heart, with consequent underdevelopment and hypoplasia extending from the left atrium to the
aortic isthmus. In simple and complex coarctation, there are great variations possible in the extent and location
of the coarctation.
Coarctation occurs in 6% to 7% of newborns with heart disease (see Tables 33-1 and 33-2). It is one of the
common causes of congestive failure in the neonate (see Tables 33-7). Among symptomatic infants, 82% have
complex coarctation, and 18% have simple coarctation. It is more common in male and premature infants. Girls
sometimes have Turner
P.679
syndrome. Severe extra-cardiac anomalies, usually renal or gastrointestinal, occur in 6% to 9% of these
patients (see Table 33-4).
TABLE 33-14 FINDINGS IN ACYANOTIC 0- TO 2-WEEK-OLD NEONATES WITH CONGESTIVE HEART
FAILUREa
Diagnosis
Coarctation
Physical
Examination
Radiographic
Findings
Leg pulses and leg Heart size,

BP, soft SEM in
pulmonary edema
back, +/- SRM, +/click, S3, +/-
+/- RVH, develops LVH, BVH
differential cyanosis,
shocklike sepsis
picture
Critical aortic stenosis
Shock, pulses and Heart size,

perfusion, SEM,
pulmonary edema
click, S3, single S2

in premature infant
Heave, pulses,
pulse pressure,
continuous or SRM
LVH, T-wave abnormalities
Heart size (LV, LA), Develops RVH, LVH, BVH

pulmonary arterial
markings
Cardiomyopathy
Large globular heart, or voltage, T-wave changes,

pulmonary edema
Q waves in ALCA
Pulses,
perfusion, pulse
pressure, HR, SRM
Critical pulmonary
stenosis
SEM, click, single

S2, most have
cyanosis
Systemic arteriovenous
fistula
Normal or
pulmonary arterial
markings, RAE
Heave, pulses,
Heart size,
wide pulse pressure, pulmonary arterial
soft SEM or SRM,
markings
bruit, shock, +/cyanosis
QRS axis 0-90, +/- LVH,

develops RVH
Develops RVH, LVH, BVH
Congestive heart failure with cyanosis may be caused by hypoplastic left heart syndrome,
transposition of the great arteries, truncus arteriosus, total anomalous pulmonary venous connection,
pulmonary atresia with tetralogy, tricuspid atresia, Ebstein malformation, or persistent pulmonary
hypertension.
+/-, may or may not be present; , decreased; , increased; ALCA, anamolous left coronary artery;
BVH, biventricular hypertrophy; HR, heart rate; LA, left atrium; LV, left ventricle; LVH, left ventricular
hypertrophy; RAE, right atrial enlargement; RVH, right ventricular hypertrophy; S2, second heart
sound; S3, third heart sound; SEM, systolic ejection murmur; SRM, systolic regurgitant murmur.
Pathophysiology
Simple Coarctation. The isthmus is normally smaller than the ascending or descending aorta in newborn infants
because only 10% of the combined ventricular output during fetal life passes through the isthmus into the
descending aorta, whereas approximately 60% passes through the ductus arteriosus to the descending aorta.
After birth, the isthmus gradually grows, but in simple coarctation, there is a constricting band just above the
point of connection to the ductus arteriosus. Aortic coarctation may acutely further constrict in the neonatal
period, as constriction of the adjacent ductal tissue occurs. During infancy, there may be relative progression of
the coarctation from inadequate,
P.680
disproportionately small aortic isthmus growth, and perhaps aortic wall hypertrophy and endothelial thickening
at the coarctation site. Collateral circulation may be present at birth. In simple coarctation, the increased
resistance to flow results in a pressure overload on the left ventricle. If the coarctation is not severe and there is
a patent ductus arteriosus, with fall in pulmonary vascular resistance after birth, there is a reversal of flow
through the ductus arteriosus from the aorta to the pulmonary artery, and a considerable left-to-right shunt may
develop. If the increased pressure and volume load exceed the ability of the heart to compensate by
hypertrophy or dilation, congestive failure with diminution of systemic output ensues. Left ventricular enddiastolic pressure is elevated, resulting in increased pulmonary venous pressure and development of pulmonary
edema. The increased pulmonary venous pressure also produces pulmonary artery hypertension and right heart
failure.
TABLE 33-15 FINDINGS IN ACYANOTIC 2- TO 8-WEEK-OLD NEONATES WITH CONGESTIVE HEART
FAILUREa
Diagnosis
Ventricular septal defect Heave, harsh SRM,

+/- S3, +/- diastolic
Endocardial cushion
defect
Radiographic
Findings
LA), pulmonary
rumble, normal pulses arterial markings
Same as ventricular
Same as ventricular
septal defect
septal defect, fixed
split S2
Heart size (RV, LV, Develops RAE, RVH, LVH, BVH
Left axis deviation, develops

RAE, RVH, LVH, BVH
Atrial septal defect
Hyperdynamic
Develops RAD, RVH
Heart size (RV,
precordium, soft SEM, normal LA and LV),
fixed split S2, +/pulmonary arterial
markings
diastolic rumble
Patent ductus arteriosus Same as presentation

in full-term infants
Cardiomyopathy
at 0-2 weeks of age

Same as presentation
at 0-2 weeks of age
Murmur may be present earlier than 2 weeks of age, and congestive heart failure may occur earlier in
premature infants.
+/-, may or may not be present; , decreased; , increased; BVH, biventricular hypertrophy; LA, left
atrium; LV, left ventricle; LVH, left ventricular hypertrophy; RAD, right axis deviation; RAE, right atrial
enlargement; RV, right ventricle; RVH, right ventricular hypertrophy; S2, second heart sound; S3, third
heart sound; SRM, systolic regurgitant murmur.
Complex Coarctation and Aortic Interruption. Complex coarctation and aortic interruption are characterized by
pulmonary artery hypertension with a ductus arteriosus supplying the descending aorta, usually a large
intracardiac left-to-right shunt, and increased pulmonary flow. The right-sided structures are dilated and
hypertrophied. There is a pressure and volume overload on both ventricles and congestive heart failure. In those
with a large ventricular septal defect and patent ductus arteriosus, the systolic pressures in the pulmonary
artery, descending aorta, ascending aorta, and right ventricle are identical. Peripheral pulse pressure is normal,
and the pulses are equal throughout. With ductal constriction, the femoral arterial pulsations diminish. If the
aortic arch obstruction is severe or complete, perfusion to the lower one-half of the body, previously supplied by
the open ductus, is reduced. Manifestations of shock, renal and mesenteric hypoperfusion, and metabolic
acidosis develop. Ductus closure causes death.
Clinical Findings
Simple Coarctation. Infants with isolated discrete coarctation may be asymptomatic, although some develop
congestive heart failure, often after the age of 1 month. The femoral and pedal pulses are absent or diminished
compared with brachial or carotid pulses. Radial and brachial pulses may be decreased if the subclavian artery
on that side arises at or below the coarctation. Systolic blood pressure in the upper extremities is higher than in
the lower extremities, but marked hypertension is uncommon. Pulse pressure in the lower extremities is
narrowed, often 10 to 15 mm Hg. S3 is often present, and there may be an apical systolic ejection click (~ half
have a bicuspid aortic valve). A systolic ejection murmur from the coarctation may be heard at the left
interscapular area over the back, and at the left upper sternal border. In the neonate a continuous murmur,
when present, is usually from a left-to-right shunt across the ductus arteriosus, as collateral vessel flow is
generally not audible until older age. Manifestations of congestive heart failure are those of combined left and
right heart failure.
Chest radiograph can be nearly normal, but usually shows cardiac enlargement and pulmonary venous
congestion.
ECG can be normal, but usually reveals right ventricular hypertrophy in the early months and left ventricular
hypertrophy later.
Echocardiographic visualization of the aortic arch usually shows the site, length, and severity of coarctation and
the aortic arch branching pattern. There is characteristically a constriction from the outer posterior curvature of
the aortic wall, and an anterior peri-ductal shelf may be identified. An instantaneous systolic gradient may be
derived from the velocities across the coarctation but may underestimate the severity of the lesion if cardiac
output is depressed. The descending aortic flow has a characteristically diminished systolic upstroke velocity and
prolonged antegrade flow.
Magnetic resonance and CT imaging can be used to delineate anatomic features not evident with
echocardiography (Fig. 33-35).
Complex Coarctation and Aortic Interruption

Infants with severe isolated and complex coarctation usually present with congestive heart failure in the early
neonatal period. Generally, the younger the infant, the more severe and complex are the combined
malformations. Complete interruption of the aortic arch is usually associated with a ventricular septal defect and
a systemic patent ductus arteriosus and is clinically indistinguishable from complicated coarctation. It is
frequently seen as part of DiGeorge syndrome, which may have additional manifestations of hypocalcemia,
absent thymic shadow on the initial chest radiograph, and possible impaired immune response to transfused
viable nonirradiated leukocytes. Besides the findings described for simple coarctation, there is evidence of a
large left-to-right shunt and pulmonary artery hypertension. Femoral pulsations may wax and wane, depending
on ductal
P.681
caliber. A pan-systolic murmur of a septal defect or mitral regurgitation may be found. Ductal closure may result
in a critically ill baby with poor perfusion, metabolic acidosis, and possibly disseminated intravascular
coagulation, necrotizing enterocolitis, and renal and hepatic dysfunction.
Figure 33-35 Magnetic resonance image of aortic coarctation in a 1 day old. Arrow indicates focal coarctation
in the distal end of a hypoplastic aortic arch.
Figure 33-36 A chest radiograph of a 1-day-old infant with complex coarctation of the aorta shows marked
cardiac enlargement and pulmonary vascular engorgement.
Chest radiographs show considerable cardiac enlargement, pulmonary plethora, and edema (Fig. 33-36).
ECG can be normal, but often has right axis deviation, right atrial hypertrophy, right ventricular hypertrophy,
and sometimes diminished left ventricular forces.
Echocardiographic examination delineates the aortic arch anatomy, and reveals associated lesions, including
mitral and aortic stenosis, ventricular septal defect, sub-aortic obstruction, and conotruncal abnormalities.
Aortic arch obstruction should be suspected in any critically ill term baby with a septic-like shock. It should also
be suspected as an associated anomaly in young babies with intracardiac anomalies such as ventricular septal
defect, single ventricle, truncus arteriosus, and aortic or mitral valve disease who develop signs of poor systemic
output. A thorough examination, with careful pulse palpation and blood pressure measurement in all 4
extremities, should lead to the correct diagnosis (see Tables 33-14 and 33-15). Infants presenting before 1
month of age usually have severe or complex coarctation. The presence of a ductus arteriosus supplying the
descending aorta may be demonstrated by the finding of a lower arterial PO2 in the legs than in the arms. The
hypoplastic left heart syndrome produces a similar shock-like picture or congestive failure in the first week as
the ductus arteriosus closes. In these patients, there is cyanosis (See Color Plate), the peripheral pulses are
diminished throughout, and the ECG shows marked diminution in left ventricular forces.
Treatment
All neonates with congestive heart failure thought to have coarctation of the aorta should be promptly
hospitalized, treated, and examined by echocardiography. Infants with complex coarctation and aortic
interruption become symptomatic because of constriction of the ductus arteriosus. Prostaglandin E1 infusion can
dilate the ductus, restore systemic perfusion, improve metabolic abnormalities, and support life during the time
needed to study the anatomy and arrange for surgery. Inotropic support with intravenous dopamine or
adrenergic agents is often needed. In critically ill babies, there may be adverse ischemic consequences for the
gastrointestinal, renal, hepatic, and coagulation systems. Echocardiographic examination usually provides the
anatomic detail needed for surgery. If needed, cardiac catheterization, digital subtraction angiography, or
magnetic resonance imaging may be useful for additional delineation of the aortic arch and intracardiac anatomy.
If after initial clinical management to effect stabilization, improvement or deterioration occurs, surgery should
not be unduly delayed. The surgical procedures employed depend on the severity of the lesion and include
resection of the coarctation with primary anastomosis, subclavian or prosthetic patch aortoplasty, or
construction of a conduit from ascending to descending aorta; division of the patent ductus arteriosus; and, if
needed, intracardiac repair of additional defects such as a large ventricular septal defect. The mortality rate for
infants with complicated coarctation is 85% without surgery. Surgery increases the survival rate to 85%.
Regardless of the type of coarctation, the mortality is related to age of presentation and is higher for those with
duct-dependent descending aortic flow. In some infants with simple and milder coarctation who respond well to
medical therapy, surgery may be delayed. Those who undergo surgical coarctation repair early in infancy may
develop re-stenosis later, which may require re-operation or catheter balloon dilation. The survivors need close
medical supervision throughout childhood and may require other operations for various associated abnormalities
later. Catheter balloon dilation of un-operated primary discrete coarctation can offer palliation in the complex
critically ill infant, but only in selected infants with an otherwise good size aortic arch does it appear to it provide
long-lasting relief (50).
Aortic Stenosis
Fusion of the right-left or right-non commissures of the aortic valve, resulting in a bicommissural, functional
bicuspid aortic valve, is one of the most common congenital anomalies, occurring in 1.5% of the population.
The resulting valve orifice may be diminished, but the effective stenosis is usually mild in neonates. Only very
severe aortic valve narrowing produces symptoms or requires intervention in early infancy (Fig. 33-37). Critical
isolated aortic valve stenosis is rare, and usually the result of fusion of both the right-non and right-left
commissures, with the valve orifice too small to allow adequate cardiac output at physiologically obtainable left
ventricular pressures. In this situation, adequate systemic blood flow (before relief of the aortic stenosis) is
dependent on a patent ductus arteriosus that
P.682
allows right-to-left blood flow from the pulmonary artery into the aorta.
Figure 33-37 Diagram of the cardiac anatomy and physiology in a 1-month-old infant with valvar aortic
stenosis had a systolic pressure gradient of 70 mmHg across the aortic valve. The blood passing from left to
right through the ductus must return again through the aortic valve, with the excess flow compounding the
obstruction. The large atrial shunt, whether a true anomaly or a sprung foramen ovale, elevates left atrial
pressure. The numbers below the chamber name are pressure measurements in mmHg determined at cardiac
catheterization; the percentages indicate oxygen saturation data. LA, left atrium; LV, left ventricle; PA,
pulmonary artery; PV, pulmonary vein; RA, right atrium; RV, right ventricle; SVC, superior vena cava. Adapted
permission.
The symptoms are those of congestive heart failure, pulmonary edema, and sometimes peripheral vascular
collapse. The baby may appear ashen and cyanotic if the pulmonary edema is severe. The cardinal features are
tachypnea, a blowing systolic murmur at the upper right or middle left sternal border and an apical early systolic
click resembling a split S1. If the valve orifice is very small, forward valve flow is small, and the murmur may be
soft.
Chest radiographs show cardiac enlargement and pulmonary venous congestion.
ECG usually has biventricular hypertrophy with T-wave changes (see Table 33-14).
Echocardiographic examination demonstrates a deformed immobile aortic valve with commissural fusion. In
severe aortic obstruction, trans-valvar flow is diminished, and the systolic murmur and the Doppler-derived
pressure gradient are of low amplitude and do not reflect the severity of the lesion. The left ventricle appears
hypertrophied and may have decreased or dilated internal dimensions and poor or hyperdynamic systolic
function. Some patients may have coarctation and mitral valve abnormalities. Echocardiography can also identify
those neonates with associated hypoplasia of the left ventricular chamber, mitral and aortic annuli and aortic
root that do not benefit adequately from valvuloplasty and require a staged hypoplastic left-heart-type surgical
approach for survival (51).
Treatment of critical stenosis consists of stabilization with administration of inotropic support, oxygen, and
frequently prostaglandin E1 to allow right ventricular support to the systemic circulation; to be followed as soon
as feasible by balloon valvuloplasty or surgical valvotomy (52). This provides effective palliation in infancy.
Although many require repeat valvuloplasty during childhood, this is usually accomplished with success.
Ultimately, and hopefully after childhood growth, many of the worse valves will require surgical replacement
with a prosthetic valve or pulmonary autograft (Ross procedure). The asymptomatic infant with auscultatory
findings of aortic stenosis and those after valvuloplasty require serial evaluation because, over the long term,
valvar aortic stenosis very frequently progresses and recurs to some degree.
Acyanotic Anomalies with Left-to-Right Shunt

Ventricular septal defects may be small or large, single or multiple, and isolated or associated with other
cardiovascular malformations. They are an integral part of complex congenital heart disease lesions, such as
tetralogy of Fallot, truncus arteriosus, double-outlet right ventricle, and atrioventricular canal, and they have
been associated with virtually every other known congenital cardiac malformation. Small, isolated self-closing
muscular ventricular septal defects, frequently detectable only by echocardiography, are the most common
congenital cardiac anomaly, occurring in 2% to 3% of term newborns (2) and more frequently in infants born
prematurely (see Tables 33-1 and 33-2). Large defects occur most commonly singly in the membranous septum,
less often in the low portion of the muscular septum, infrequently beneath the pulmonary valve, or posterior
next to the tricuspid valve. Even though only 10% of ventricular septal defects cause symptoms, they remain
the most common cause of congestive heart failure after the second week of life (see Tables 33-7). Although
small defects are usually isolated anomalies, extra-cardiac malformations occur in up to 24% of the patients
with large defects. Recognition of a large ventricular septal defect remains paramount because without closure a
large defect with pulmonary hypertension can lead to irreversible Eisenmenger-type pulmonary vascular disease
by as early as the first birthday.
Pathophysiology
The common small ventricular septal defect has minimal hemodynamic functional effect and does not produce
symptoms, but a moderate or large defect in a neonate may
P.683
cause significant hemodynamic alterations. If the defect is large, right and left ventricular pressures equilibrate,
and pulmonary hypertension results (Fig. 33-38). The decreasing pulmonary resistance after birth allows an
increasing left-to-right shunt through the defect. The normal regression of pulmonary resistance in the first
week of life is usually delayed in these babies. Nonetheless, sufficient reduction in pulmonary resistance occurs
by the second week of life to cause symptoms in many patients. Others, with smaller defects or further delay in
the reduction of pulmonary vascular resistance, develop symptoms as late as 3 to 4 months of age.
Symptoms are the result of congestive heart failure, sometimes presenting with superimposed pulmonary
problems such as bronchiolitis, pneumonia and atelectasis. Congestive heart failure is caused by the obligate
high resting cardiac output associated with recirculation of large amounts of blood through the heart and lungs
although simultaneously attempting to meet the demand for systemic flow. Cardiac pump reserve for exertions
such as feeding is therefore diminished. Excessive pulmonary vascular flow and pressure decrease lung
compliance, resulting in more rapid and labored, but shallower, breaths. The fixed tachypnea impairs feeding
and increases caloric expenditure, resulting in diminished growth. Pulmonary congestion may not only decrease
the tolerance for but increase susceptibility to recurrent respiratory infections. Enlarged left pulmonary artery
and atrium may compress bronchi and result in pulmonary atelectasis. Because pulmonary vascular resistance is
lower in premature infants at birth, the development of symptoms from a ventricular septal defect occurs earlier.
Figure 33-38 Diagram of the anatomy and physiology with a large ventricular septal defect in a 1-month-old
baby. The defect allows equilibration of pressure between the two ventricles. With pulmonary resistance much
less than systemic resistance, there is a very large left-to-right shunt that caused congestive heart failure, as
evidenced by the elevated atrial pressures and the reduced pulmonary venous oxygen saturation because of
pulmonary edema. The numbers below the chamber name are pressure measurements in mmHg determined at
cardiac catheterization; the percentages indicate oxygen saturation data. LA, left atrium; LV, left ventricle; PA,
pulmonary artery; PV, pulmonary vein; RA, right atrium; RV, right ventricle; SVC, superior vena cava. Adapted
permission.
Gradual improvement and diminution in pulmonary blood flow in an infant with a moderate or large ventricular
defect may occur if there is an anatomic decrease in the size of the defect. Many defects spontaneously close,
and mostparticularly muscular and membranous defectsbecome smaller with time. During childhood, but
rarely in infancy, there may be progressive and irreversible development of anatomic obstructive changes in the
pulmonary arterioles.
Clinical Findings
A small ventricular septal defect is characterized by an isolated mid-higher pitched blowing systolic murmur,
which is fairly localized along the left sternal border or at the apex. Infants with large septal defects develop
congestive failure in the first few months of life with symptoms of tachypnea (i.e., rate consistently > 60/min),
fatigue with feeding, decreased oral intake. Increased respiratory work and retractions, excessive diaphoresis,
and recurrent respiratory infections are later manifestations. Weight gain lags considerably behind height
maturation. The infant often presents with a respiratory infection that may precipitate or mask underlying
congestive failure.
On examination, the infant is scrawny and tachypneic. The cardiac impulse is hyperdynamic. If pulmonary artery
hypertension exists, the second heart sound may appear single and loud from early accentuated pulmonary
closure. A gallop sound may be heard and is often associated with a mid-diastolic rumble. A mixed frequency
holo-systolic murmur, heard best at the lower left sternal border, is soft at birth and usually becomes loud,
rough and well transmitted throughout the precordium below the suprasternal notch. There is hepatomegaly and
infrequently pulmonary wheezing and rales. Peripheral pulses can be rapid and peripheral edema is rare.
Chest radiograph shows considerable cardiac enlargement, increased pulmonary blood flow, and sometimes
pulmonary edema. The main pulmonary artery segment and left atrium are often enlarged. Atelectasis and
parenchymal infiltrates are common. ECG usually reveals left ventricular hypertrophy, and if the lesion is
associated with pulmonary artery hypertension, right ventricular hypertrophy is detected.
Echocardiographic exam can demonstrate the size, location, and number of ventricular septal defects (Fig. 3339). Associated lesions not appreciated on physical examination, including atrial septal defect, patent ductus
arteriosus,
P.684
coarctation of the aorta, and left and right ventricular outflow obstructions, are revealed by echocardiography.
Right ventricular and pulmonary hypertension can be assessed both from the curvature of the ventricular
septum and by comparison of Doppler measurement of the instantaneous systolic pressure gradient across the
defect with simultaneous blood pressure. Often there is at least a trivial degree of tricuspid regurgitation
providing means to also estimate right ventricular pressure from the pressure gradient between the right
ventricle and right atrium. Defects with a large amount of shunt across them also show evidence of left
ventricular volume overload, with large left atrial and left ventricular dimensions and hyperdynamic left
ventricular function. Occasionally, an infant with moderate congestive symptoms or findings, or evidence of
borderline pulmonary hypertension, or of elevated pulmonary vascular resistance may require cardiac
catheterization to delineate the hemodynamics.
Figure 33-39 Echocardiogram 4 chamber view with color Doppler analysis demonstrates an apical musular
ventricular septal defect. (LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle; VSD,
ventricular septal defect.) (See color plate)
In the neonate, the murmur of a small ventricular septal defect is generally characteristic, but sometimes can be
difficult to differentiate from that caused by a small patent ductus arteriosus or tricuspid regurgitation (see Table
33-15). Some infants with large sub-aortic defects have or develop progressive pulmonary stenosis that
prohibits left-to-right shunting, cardiomegaly, and congestive heart failure. These babies may develop the
features of the tetralogy of Fallot. Later increasing right ventricular hypertrophy on the ECG suggests the
development of pulmonary stenosis or increasing pulmonary vascular resistance and the need for careful
reevaluation. The coexistence with a large ventricular septal defect of additional malformations resulting in a
large left-to-right shunt (e.g., truncus arteriosus) can be difficult to differentiate clinically from isolated large
ventricular septal defects; ascertaining their presence often requires echocardiographic examination.
Treatment
An infant with a small ventricular septal defect requires no specific treatment but should be followed. In infants
with congestive heart failure, administration of digitalis and diuretics may produce considerable improvement in
respiratory symptoms and growth. Systemic afterload reduction with ACE inhibitors (e.g., captopril) moderately
decreases the pulmonary/systemic flow ratio and often has additional moderate symptomatic benefit in
refractory patients (see Table 33-10). The use of high-caloric formulas, made by supplementation of standard
formulas with additional carbohydrate (e.g., corn syrup or polycose) and oil (e.g., corn oil or MCT oil) up to a
total of 30 kcal per ounce, is often needed for growth in babies with large defects. The total ad libitum oral
intake should not be restricted, because growth failure and small size are a common issue in these infants.
Timely surgical closure of the defect is indicated if the infant does not begin to grow with timely escalation of
optimal medical therapy, or has persistent significant pulmonary artery hypertension after 6 months of age.
Primary repair in infants entails cardiopulmonary bypass, and, in most patients, atriotomy with patch closure
through the tricuspid valve. Some patients require closure through the pulmonary valve or right ventriculotomy.
Very low birth weight premature infants and those with multiple large muscular defects may require an initial
pulmonary artery banding procedure with corrective surgery done at a later age. Of infants born with isolated
large defects requiring closure in the first year of life, mortality is now 1% to 2% or less. Mortality is greater if
there are associated severe extra-cardiac congenital anomalies, pulmonary complications, or prematurity. The
long-term prognosis after trans-atrial surgical closure of an isolated ventricular septal defect in the first year of
life is excellent, with essentially normal hemodynamics and a small risk for symptomatic dysrhythmias for most
patients.
Secundum and Sinus Venosus Atrial Septal Defects

Virtually all babies have a patent foramen ovale at birth. Many foramen ovale functionally close within hours of
birth, but many others remain at least partly open for several months and in about 20% the foramen has some
blood flow across it throughout life. This is important to the neonatologist, because umbilical vein catheters tend
to follow the course of the circulation for the preceding 9 months and may pass through the foramen ovale into
the left heart,
P.685
providing erroneous measures of oxygen levels and allowing passage of intravenously injected air or thrombus
to the brain, occasionally with disastrous results.
Pathophysiology and Clinical Features

Secundum atrial septal defects are a common anomaly resulting from an opening in the primum atrial septum
around the region of the foramen ovale that rarely causes symptoms or a murmur loud enough to attract
attention in infants, and often close spontaneously (2). Much more rare sinus venosus atrial septal defects occur
in the atrial septum in which the cavae enter, usually the superior vena cava, often with partial anomalous
connection of one or more right pulmonary veins, and are clinically indistinguishable from secundum atrial septal
defects. As systolic blood pressure, arterial stiffness, and left ventricular stiffness increase with age, left to right
shunting develops across the defect, resulting in increased flow across right heart valves, systolic pulmonary
flow murmur at the left upper sternal border and suprasternal notch, diastolic tricuspid flow rumble at the left
lower sternal border, persistent delayed closure of the pulmonary valve (and relatively wide fixed splitting of the
second heart sound), increased right ventricular workload and increased pulmonary blood flow. Most are found
because a murmur of mild pulmonary stenosis is heard and an echocardiographic examination is performed (Fig.
33-40). Others are discovered during evaluation of coexisting cardiac anomaly, or evaluation initiated by extracardiac anomalies or failure to thrive. Most small secundum defects (<5 mm diameter), many moderate-size
ones (5-8 mm diameter), and some large defects spontaneously close, or nearly close in the first several years
of life (2,74). Sinus venosus defects rarely if ever close spontaneously. Because there is rarely significant
congestive heart failure, initial management consists of observation and, if the defect persists, it may be closed
by surgical or catheter techniques. If a large defect remains open, approximately 10% of those develop
Eisenmenger-type pulmonary vascular disease later in life.
Figure 33-40 Echocardiogram 4 subcostal view with color Doppler analysis demonstrates an secundum atrial
septal defect. (ASD, ventricular septal defect; LA, left atrium; RA, right atrium.) (See color plate)
Rarely, a large atrial septal defect is associated with an early decrease in the pulmonary resistance and a large
left-to-right shunt in the first months of life. Growth failure and congestive heart failure may raise the question
of early cardiac surgery. This is a potentially more complicated and treacherous situation because occult leftsided heart disease (e.g., myocardial disease, aortic stenosis, coarctation) is often responsible for the unusually
large imbalance of ventricular compliances, causing a large left-to-right atrial shunt in infancy. Surgical closure
of the defect without dealing with the additional problem may have a poor outcome. The simple rule of thumb is
to search diligently for associated anomalies and proceed to surgery for isolated atrial septal in early infancy
with caution.
Endocardial Cushion Defects

Defects of endocardial cushion development may be partial, resulting in an ostium primum atrial septal defect;
complete, resulting in additional total deficiency of the posterior inlet ventricular septum and a common
atrioventricular valve (i.e., complete atrioventricular canal); or transitional with a combination of a smaller
restrictive defect of the inlet ventricular septum and primum atrial septal defect, (i.e., transitional
atrioventricular canal defect). The atrioventricular valves, particularly the anterior mitral valve leaflet, are
usually malformed, deficient, or abnormally attached to the ventricular septum. With ostium primum defect,
there is usually a cleft in the mitral valve and frequently mitral regurgitation. In complete atrioventricular canal,
the primitive atrioventricular valve floats like a sail over both ventricles. This malformation results in a large
communication between the right and left atria and the right and left ventricles. Significant atrioventricular valve
regurgitation is less common than in those with only an ostium primum defect. Occasionally, but more often in
those without trisomy 21, the mitral valve has abnormal chordal attachments and is stenotic. Rarely the large
atrioventricular valve is laterally shifted and primarily centered over one ventricle, and the contralateral ventricle
is much smaller than normal. Endocardial cushion defects as primary lesions account for 4% of all newborns
with serious heart disease (see Tables 33-1 and 33-2).
P.686
Pathophysiology
The hemodynamic consequence of an ostium primum atrial septal defect is right ventricular volume overload,
which is caused by a left-to-right shunt across the atrial septal defect, and variable biventricular volume
overload from regurgitation from the left ventricle through the cleft mitral valve to the atria. The volume load, if
aggravated by significant mitral regurgitation, can be large and result in congestive heart failure. Streaming of
inferior vena cava blood across the large, low-lying defect and cleft common valve leads to mild systemic arterial
oxygen desaturation. In complete atrioventricular canal, there is an additional left-to-right shunt through a
ventricular septal defect and right ventricular and pulmonary artery hypertension at a systemic level. Infants
with pulmonary artery hypertension are particularly susceptible to the development of pulmonary vascular
obstructive disease and its complications in later childhood.
Clinical Findings
Isolated primum atrial septal defects without mitral regurgitation results in physical findings similar to secundum
and sinus venosus defects with pulmonary flow murmur, fixed second heart sound splitting and sometimes
tricuspid diastolic rumble, and is distinguished by ECG left axis deviation. Infants with ostium primum atrial
septal defects who are symptomatic in the neonatal period usually have severe mitral regurgitation. Growth
retardation may be marked, and weight lags considerably behind height maturation. Recurrent pulmonary
infections are common.
Figure 33-41 Diagram of the anatomy and physiology of an atrioventricular canal in an asymptomatic girl with
Down syndrome. There was no murmur. A: Although she was breathing room air, the pulmonary resistance
was high; there was no left-to-right shunt, and she had arterial oxygen unsaturation. B: When breathing
oxygen, a large left-to-right shunt developed and the estimated pulmonary resistance fell sharply. The
percentages indicate oxygen saturation; the numbers in italics are pressure measurements. The numbers below
the chamber name are pressure measurements (mm Hg) determined at cardiac catheterization; the
percentages indicate oxygen saturation data. LA, left atrium; LV, left ventricle; PA, pulmonary artery; PV,
pulmonary vein; RA, right atrium; RV, right ventricle; SVC, superior vena cava. Adapted from Mullins CE, Mayer
DC. Congenital heart disease: a diagramatic atlas. New York: Alan R Liss, 1988, with permission.
With complete atrioventricular canal, there is frequently mild cyanosis. The cardiac impulse is hyperdynamic,
and S1 is obscured by a loud pan-systolic murmur audible at the apex or left sternal border. There is usually
pulmonary hypertension, and the S2 is accentuated. A S3 and an apical mid-diastolic rumble are often heard.
Occasionally, particularly among neonates with Down syndrome, there may be no perceptible auscultatory
abnormality. When the murmur is subdued, cardinal features of cardiac anomaly are a hyperdynamic
precordium, abnormal second heart sound, and ECG leftward superior axis.
Approximately half of those with isolated complete atrioventricular canal defects have trisomy 21 (75). Forty
percent of infants with Down syndrome have congenital heart disease, complete atrioventricular canals being
most common. Because babies who have Down syndrome have a tendency to under-ventilate, causing
pulmonary venous oxygen desaturation, they may have pulmonary hypertension that, associated with a
common atrioventricular canal, may limit left-to-right shunting to amounts that do not produce a murmur (Fig.
33-41). All infants with Down
P.687
syndrome should be evaluated for congenital heart disease by a cardiologist.
Figure 33-42 Electrocardiogram of infant with atrio-ventricular cushion septal defect shows the characteristic
left axis deviation.
Chest radiograph shows cardiac enlargement, sometimes out of proportion to the increased pulmonary
vasculature, attributable to the large atria. The main pulmonary artery segment is prominent, and there is
pulmonary vascular engorgement.
ECG characteristically shows a left superior QRS axis in the frontal plane, commonly 0 to -60 in primum
defects and -60 to -100 in complete canal with a small Q wave in lead aVL (see Fig. 33-42). Significant right
ventricular hypertrophy usually indicates right ventricular hypertension (Fig. 33-41).
Echocardiographic examination demonstrates the anatomic features relevant to surgical repair, including the
anatomy of the atrioventricular valve with its chordal attachments, papillary muscles, ventricular relationships,
possible regurgitation or stenosis of the atrioventricular valves (Fig. 33-43), and possible associated anomalies
including systemic and pulmonary venous anomalies, secundum atrial septal defect, muscular ventricular septal
defects, ventricular hypoplasia, left or right ventricular outflow stenosis. Preoperative cardiac catheterization is
generally not required except to evaluate pulmonary vascular resistance if there is evidence of pulmonary
vascular disease.
Treatment
In many patients, palliative or corrective surgery has to be performed in infancy because of refractory
congestive heart failure or pulmonary hypertension. Timely treatment with digitalis, diuretics, afterload
reduction and caloric supplementation of feedings may result in sufficient improvement in respiratory work and
growth to allow substantial increase in the baby's size and well being for several months before operative repair
(see Table 33-10). Infants with refractory symptoms of congestive heart failure should proceed to surgery. In
infants with complete atrioventricular canals, there is pulmonary artery hypertension, and surgery is mandatory
within the first year to prevent irreversible pulmonary vascular changes.
Primary complete repair is the preferred treatment. This entails cardiopulmonary bypass, atriotomy, patch
closure of the atrial and ventricular septal defects, and attachment of the common valve leaflet to the patch or
patches. In infants with refractory congestive heart failure weighing less than approximately 2 kg or those with
serious
P.688
confounding noncardiac illness (e.g., duodenal atresia), pulmonary artery banding may be helpful, with complete
repair accomplished later. Children with isolated uncomplicated ostium primum atrial septal defects and few
symptoms can undergo complete repair at several years of age. The long-term prognosis after surgery in infancy
is excellent (76). Late dysrhythmias occasionally occur. There is often some postoperative regurgitation of the
atrioventricular valve, but in most infants this is not a significant problem. Systemic vasodilators can reduce the
volume of regurgitation and may help preserve ventricular function in patients with significant postoperative
mitral regurgitation. When there is severe mitral regurgitation, closure of the septal defect with valvuloplasty
may result in clinical improvement, but residual mitral regurgitation may later require valve palliation or
replacement. Without surgery, the prognosis is poor. Only 50% of patients with endocardial cushion defects who
become symptomatic in the first month of life survive beyond 1 year of age without surgical treatment, and
many of these have significant growth failure.
Figure 33-43 Echocardiographic apical four-chamber view of an infant with trisomy 21 and complete
atrioventricular canal defect. >, primum atrial septal defect; <, posterior inlet ventricular septal defect; LA, left
atrium; LV, left ventricle; RA, right atrium; RV, right ventricle.
Patent Ductus Arteriosus

The ductus arteriosus, arising from the distal dorsal sixth aortic arch, is well developed by the sixth week of
gestation and forms a bridge between the pulmonary artery and the dorsal aorta, inserting at the aortic isthmus.
At term, it is a muscular contractile structure. In the full-term infant, functional closure occurs during the first
day of life. Persistence of patency of the ductus arteriosus alone or in association with other cardiovascular
lesions may produce no symptoms or severe hemodynamic changes, depending on its size.
Isolated patent ductus arteriosus, without hyaline membrane disease, is common, accounting for 4% of all
newborns symptomatic with heart disease (4). It is more prevalent in female than male infants. Patent ductus
arteriosus is a frequent complication of hyaline membrane disease in the premature infant, in surviving
premature infants, and in infants born at high altitudes; in these patients, there is no sex difference. It is
common in combination with other congenital heart lesions (e.g., coarctation of the aorta, ventricular septal
defect, vascular ring). It occurs in 60% to 70% of infants born with congenital rubella.
Pathophysiology
Ductal closure occurs by constriction and then remodeling with apoptosis (77). At birth, ductal constriction is
caused by multiple factors, the most important of which appear to be increased oxygen tension, the levels of
prostaglandins, and ductus muscle mass. Prostaglandin E1 is used to dilate a closing ductus in several forms of
congenital heart disease in which patency of the ductus arteriosus is necessary to support pulmonary or
systemic blood flow (53,54). Delayed closure frequently occurs in premature infants with respiratory distress
syndrome. An inhibitor of prostaglandin synthesis (e.g., indomethacin) is used to promote closure of the ductus
in this situation (78,79). The ductus arteriosus that remains patent in the term infant is abnormal and is rarely
susceptible to pharmacologic closure.
Within the first hours after birth a small right-to-left or bidirectional shunt may occur. With fall in pulmonary
vascular resistance and a rise in systemic resistance, a left-to-right shunt develops through the ductus
arteriosus. If spontaneous closure does not occur and the ductus is small, the left-to-right shunt remains small.
However, a moderate-sized patent ductus arteriosus is usually associated with a significant left-to-right shunt,
increased pulmonary blood flow, left ventricular volume overload, increased left ventricular end-diastolic volume
and pressure, elevation of left atrial pressure, and the development of congestive heart failure. The run off of
flow from the aorta into the ductus produces wide pulse pressure, and generates bounding peripheral pulsations.
A large patent ductus arteriosus produces pulmonary artery hypertension because the pressure is transmitted
directly from the aorta to the pulmonary artery through the large defect. Those with moderate and large ducts
are prone to the development of pulmonary vascular obstructive disease by 1 year of age or beyond.
The premature infant may develop congestive heart failure earlier because of incomplete development of the
medial musculature in the small pulmonary arterioles. The contractile function of the heart, required to handle
the increased volume load, may be incompletely developed. Among those with respiratory distress syndrome,
there may be an initial period of improvement as the pulmonary status improves, followed by clinical
deterioration as left-to-right shunting through the ductus arteriosus increases.
Clinical Findings
Term Infants
In the neonate with a patent ductus arteriosus, as in all left-to-right shunts, the elevated but decreasing
pulmonary vascular resistance determines the clinical manifestations. There is usually a crescendo systolic
murmur, often with clicks, sometimes detectably spilling into diastole. Often, S2 is not clearly audible. A
continuous murmur develops later. The infant with a large patent ductus arteriosus has bounding peripheral
pulses, wide pulse pressure (defined as the difference between systolic and diastolic pressure exceeding half of
the systolic blood pressure), and hyperactive cardiac impulse at the apex (see Table 33-15). There may be an
apical diastolic rumble and symptoms and signs of congestive heart failure, poor weight gain, and recurrent
pulmonary infections. In a full-term infant with a large patent ductus arteriosus, overt failure usually does not
develop until 3 to 6 weeks of age. If the pulmonary resistance remains high, there may be little murmur.
Chest radiograph shows cardiac enlargement, pulmonary plethora, a prominent main pulmonary artery, and left
atrial enlargement. ECG develops left ventricular hypertrophy, occasionally left atrial hypertrophy, and in severe
failure, ST-T wave changes.
P.689
Echocardiography demonstrates the ductus arteriosus, its size, and the direction of the flow across the defect.
Disturbed flow in the pulmonary artery, seen best with color Doppler techniques, is particularly helpful in
identifying a patent ductus arteriosus. Continuous-wave Doppler allows measurement of the pressure gradient
across the defect and, thereby, estimation of pulmonary pressure (Fig. 33-14). Right ventricular hypertension is
also indicated by flattening of the ventricular septum curvature. A large left-to-right shunt is indicated by left
heart volume overload and a large left atrium and left ventricle. If there is associated pulmonary disease, the
pulmonary resistance may be high, allowing only right-to-left shunting. Right-to-left ductal shunting also occurs
with left heart obstructive lesions and coarctation of the aorta.
Preterm Infants
Preterm infants with a patent ductus arteriosus often have the same clinical findings as term babies. Many have
a systolic murmur and some have a classic continuous murmur. However, many premature neonates with a
large ductus arteriosus have no murmur. Most will have an increase in pulse pressure, at least intermittently.
Because arterial pressure varies with age, gestational age, and illness, a rule of thumb for elevation of pulse
pressure is when it exceeds half the systolic arterial pressure. Although preterm infants with a large patent
ductus arteriosus may develop circulatory overload within the first week of life, some have no specific clinical or
radiographic signs discernible from respiratory illness. Unlike term infants, there is no substantial increased
incidence of additional cardiac anomalies. However, if examination raises the likelihood of other cardiac or aortic
arch anomalies, echocardiographic examination should be done before pharmacologic treatment.
Echocardiographic examination of ductal diameter and length, ductal pressure gradient to estimate pulmonary
artery pressure, aortic arch anatomy, and possible associated cardiovascular anomalies is indicated prior to
surgical closure (Fig. 33-14).
The infant with congestive failure and a large left-to-right shunt caused by a ventricular septal defect may be
clinically indistinguishable from the one with a large patent ductus arteriosus. Other lesions that may result in a
large aortic runoff and mimic a patent ductus arteriosus include truncus arteriosus, hemitruncus (i.e., right
pulmonary artery from the ascending aorta), aortopulmonary window, aneurysm of the sinus of Valsalva, and
large arteriovenous malformations (see Table 33-15). In the sick neonate, clinical differentiation from other
lesions is possible using echocardiography.
Treatment
Term Infants. The full-term baby with a persistent patent ductus arteriosus and no evidence of cardiovascular
embarrassment should be followed, and catheter closure or thoracoscopic or surgical division of the ductus
performed later. The choice of method and timing of closure depend on a number of factors, including ductal
size. Before therapeutic closure, term infants with congestive heart failure often have symptomatic improvement
from treatment with digoxin and diuretics (Table 33-10).
Preterm Infants. Among preterm infants with significant patent ductus arteriosus, indomethacin treatment
produces closure in approximately 85% of patients. In symptomatic babies there is a corresponding resolution of
findings of congestive heart failure, reduction in required respiratory support and improvement in survival (80).
Its use in premature infants with a patent ductus arteriosus who do not yet manifest obvious symptoms from
circulatory overload appears to improve many outcome variables including development of congestive failure
symptoms, duration of ventilatory and oxygen treatment, and growth (80). Prophylactic administration of
indomethacin early after birth in very premature infants decreases the incidence of patent ductus arteriosus,
congestive symptoms, cerebral intraventricular hemorrhage, and possibly mortality (81,82,83 and 84).
However, until there are data demonstrating an acceptable effect on neurologic function long term, there is
uncertainty about the routine prophylactic early use of indomethacin because of the demonstrated negative
effects of the drug on neonatal vasoregulation and cerebral blood flow and a theoretical increased risk for
cerebral leukomalacia (83). Indomethacin can also cause deterioration of renal and platelet function, and its
prophylactic use increases the incidence of oliguria and necrotizing enterocolitis (80,82). It should be avoided if
there is significant renal dysfunction, thrombocytopenia, or bleeding. When available, newer prostaglandin
synthesis inhibitors such as ibuprofen may have fewer side effects (85). The ductus arteriosus occasionally
reopens after initially successful indomethacin treatment and may respond to a second course of treatment.
Failure of indomethacin does not adversely affect subsequent surgery (86,87). Avoiding or correcting anemia
diminishes the left ventricular volume overload and increases the arterial oxygen content. Surgical interruption
of the ductus arteriosus is indicated, regardless of age or weight, in any infant with a persistent
hemodynamically significant left-to-right shunt, particularly if there is pulmonary artery hypertension. Surgical
mortality is low, and dramatic improvement often occurs. The procedure is performed using a left thoracotomy,
or thoracoscope, in the intensive care nursery or the operating room under intravenous or inhalation general
anesthesia. Catheter closure is not yet readily technically achievable in small preterm neonates.
Aortopulmonary Window
Defects in the aortopulmonary septum are a rare anomaly resulting in a communication, usually large, between
the ascending aorta and main pulmonary artery. Unlike truncus arteriosus, there are usually two normal
semilunarvalves, and most do not have a ventricular septal defect.
P.690
In the approximately half without other cardiovascular anomalies, the physiology and clinical course are similar
to truncus arteriosus with large left-to-right shunt, congestive symptoms, and pulmonary hypertension. The half
with other cardiovascular anomalies most often have interrupted aortic arch and present with signs of aortic arch
obstruction. Anomalous origin of the right pulmonary artery from the aortic trunk (right hemitruncus),
anomalous origin of the coronary arteries from the pulmonary trunk, and other anomalies also occur with it. The
diagnosis is established by echocardiography. Angiography is sometimes needed to delineate details of the
anatomy needed for management. Treatment is surgical (88).
Arteriovenous Malformations
Malformation of the developing peripheral vascular system can result in abnormal connections of arteries,
arterioles, and capillaries to the venous system (i.e., arteriovenous fistulae) that create a large shunt. These
fistulas can involve vessels of any size and location. Large malformations presenting soon after birth with
congestive heart failure occur more often in the liver and head. Capillary hemangiomas involve ongoing
abnormal neovascularization. Rarely, infants with prolonged respiratory disease complicated by pneumothorax
requiring multiple chest tubes may develop collateral vessels from systemic arteries in the chest wall to the
pulmonary arteries. Although most infants with arteriovenous malformations have no other cardiovascular
anomaly, abnormal congenital systemic-to-pulmonary vascular corrections can occur with tetralogy of Fallot with
pulmonary atresia, partial anomalous pulmonary venous connection (i.e., scimitar syndrome), and
bronchopulmonary sequestration.
Pathophysiology
Although most infants do not develop cardiovascular symptoms, a large systemic arteriovenous malformation
can result in significant left-to-right shunt and congestive heart failure. Symptomatic babies usually have
connections of relatively large arteries and veins in the cerebral or hepatic vasculature. Pulmonary arteriovenous
malformations result in an intrapulmonary right-to-left shunt and cyanosis, but they do not produce congestive
heart failure.
Clinical Findings
Arteriovenous fistula is one of the few cardiovascular defects that may produce severe congestive heart failure in
the first day of life. Cardiovascular shock may be the predominant clinical picture. There may be a hyperdynamic
precordium and pulses, flow murmur, severe congestive heart failure, and cyanosis. Bruits over the fontanelle,
posterior neck, or abdomen may be audible, and there may be an enlarged head or liver. Echocardiography can
demonstrate biventricular dilation and sometimes an enlarged cava with increased flow. Arterial contrast
injection demonstrates systemic arteriovenous fistulas. Systemic venous or pulmonary artery injection of
contrast demonstrates pulmonary arteriovenous malformations. Ultrasonography, computed tomography, MRI,
and angiography may be useful in finding and delineating the lesion.
Treatment
Malformations causing congestive heart failure usually do not spontaneously improve, except for capillary
malformations that may respond to steroid or antiangiogenic drugs such as interferon. Large vessel
malformations require mechanical occlusion. Surgery carries a considerable risk, and transcatheter occlusion
with a variety of devices, including coils and detachable balloons, has been successful in many, usually older,
patients.
Vascular Rings and Slings

A variety of intrathoracic vascular anomalies may encircle the trachea and esophagus and result in symptoms in
the neonatal period. Depending on the degree of compression of the trachea or esophagus, several may present
in infants with stridor, wheezing, cough, recurrent infections, or feeding difficulties. All of these symptoms are
more commonly caused by other abnormalities, such as choanal atresia, tracheomalacia, laryngeal web,
hemangioma, or gastroesophageal reflux.
Although uncommon, vascular anomalies may result in serious or life-threatening symptoms and therefore
should be considered in any infant with persistent unexplained respiratory symptoms.
Right Aortic Arch with Anomalous Left Subclavian Artery

The spectrum of aortic arch anomalies is most commonly explained by the double arch model first proposed by
Edwards and subsequently modified by others (89). This hypothesis explains all observed arch variants by
abnormal persistence or regression of portions of a double arch present in embryologic development. The most
common of the arch anomalies that has been associated with symptoms in the neonate is the right aortic arch
with anomalous left subclavian artery. In this malformation, the aortic arch passes to the right of the trachea
over the right mainstem bronchus, then giving rise to the left subclavian artery as the last brachiocephalic
branch. A left-sided remnant of the ductus arteriosus connects the pulmonary artery with the descending aorta,
resulting in a vascular ring encircling the trachea and esophagus.
Symptoms, when present, are usually not severe and commonly occur beyond the newborn period. The
diagnosis may be suspected on plain anterior-posterior chest x-ray by leftward shifting of the trachea from the
right-sided arch.
Barium swallow may demonstrate a posterior, oblique indentation of the esophagus from the left subclavian
artery.
Echocardiography can demonstrate the position of the arch relative to the trachea and the branching pattern.
P.691
Associated cardiac malformations, if present, can also be determined at the time of echocardiographic evaluation.
Figure 33-44 Aortogram in the anteroposterior projection in a 2-month-old infant with a history of sudden
apnea requiring resuscitation. The anatomy of a typical double aortic arch is displayed. The ascending aorta
(AAO) connects to bilateral aortic arches that encircle the airways. The larger and more cephalad right aortic
arch (RAA) passes over the right mainstem bronchus. The smaller, lower left aortic arch (LAA) encircles the
trachea and passes over the left mainstem bronchus before rejoining the right arch to form the descending
aorta (DAO). Solid white lines outline the trachea and mainstem bronchi. Black arrows show the course of the
aortic arches.
Magnetic resonance imaging has proven to be a useful tool for determining vascular anatomy, and may display
evidence of tracheal compression.
Double Aortic Arch

Failure of the normal regression of the embryologic right arch between the right subclavian artery and the
descending aorta results in a double aortic arch. The right-sided arch is the larger and more cephalad in
approximately 75% of patients. The resulting vascular structure completely encircles the trachea and esophagus.
Symptoms may be dramatic, even life-threatening, and usually occur within the first few months of life.
The plain chest radiograph is generally nondiagnostic. The anatomy can be imaged by echocardiographic
examination, magnetic resonance imaging, or angiography (Fig. 33-44).
Barium swallow demonstrates bilateral indentation of the esophagus. If done, bronchoscopy will show the
pulsatile compression of the trachea by the vascular ring.
Surgery is indicated in symptomatic infants and consists of division of the smaller arch, usually the left. The
postoperative relief of symptoms and prognosis are usually good.
Anomalous Origin of the Left Pulmonary Artery (Pulmonary Sling)

Anomalous origin of the left pulmonary artery is a rare but serious vascular defect, in which the left pulmonary
artery arises from the proximal right pulmonary artery and passes between the trachea and the esophagus
before supplying the left lung.
Respiratory symptoms are often severe, and there may be associated hypoplasia or stenosis of the trachea or
right mainstem bronchus. Swallowing difficulties are uncommon.
Barium esophagram demonstrates anterior compression from the aberrant vessel. Hyperinflation of the right
lung as a result of selective compression of the right bronchus may be noted on plain chest radiograph.
Echocardiography will generally demonstrate the key features of the anatomy, although MRI, angiography, and
bronchoscopy have been recommended by some to delineate additional details (90,91).
Surgery is indicated and consists of division of the left pulmonary artery with reanastomosis to the main
pulmonary artery in a more normal location. Outcome after surgical correction of the pulmonary artery origin
has been fair, limited in part by associated lesions and by persistent pulmonary complications related to residual
airway abnormalities. Residual left pulmonary artery stenosis is common after surgery.
Anomalous Origin of the Innominate Artery

Compression of the trachea by the innominate artery is a controversial source of respiratory distress in neonates
and infants. It has been contended that airway compromise may result if the origin of an otherwise normal
innominate artery is more distal than usual, resulting in indentation of the trachea as the vessel courses
anteriorly and rightward before splitting into the right subclavian artery and right carotid artery.
Bronchoscopy may reveal a pulsatile mass compressing the distal trachea.
Surgical intervention to suspend the innominate from the anterior chest wall has been advocated, with reported
relief of respiratory symptoms in some patients. In some series, this situation has represented a significant
portion of infants undergoing surgery for airway compression (92). However, others have disputed this
mechanism as a common cause of respiratory distress and suggest conservative management (93).
Acyanotic Anomalies with Abnormal

Cardiac Function or Structure
Cardiomyopathy
Sick neonates without cardiac anatomic anomaly often develop symptoms of
cardiac and myocardial dysfunction. Systolic and/or diastolic dysfunction most
often occur secondary to another abnormality such as sepsis or hypothermia, but
occasionally from disorders fundamentally involving myocyte biochemistry.
Cardiomyopathy is abnormal myocardial cellular function or structure that occurs
with
P.692
many abnormalities and disorders (see Tables 33-16 and 33-17). Myocardial
dysfunction may be grouped by clinical and echocardiographic determination of
the cardiovascular pathophysiology, without regard to etiology, as dilated,
hypertrophic, and restrictive cardiomyopathy. The nature of appropriate
supportive cardiac treatment depends on this cardiovascular physiologic
classification. However, the outcome of supportive therapies alone is limited.
Additional improvement in outcome may result from determining causation and
directed treatment based on etiology.
TABLE 33-16 DILATED CARDIOMYOPATHIES
Infectious causes
Viral (coxsackie, adenovirus, echo, CMV)
Bacterial sepsis (endotoxemia, exotoxemia)
Myocardial ischemia
Asphyxia
Anomalous origin of left coronary artery
Reversible electrolyte and metabolic causes of myocardial
dysfunction
Hypoglycemia
Hypocalcemia
Hypophosphatemia
Hypothermia
Polycythemia
Work-overload cardiomyopathies
Tachycardia induced (incessant SVT or VT)
Severe pulmonary hypertension (H&D)
Critical aortic valve stenosis (H&D)
Genetic isolated cardiomyopathy

Dilated cardiomyopathy
Familial dilated cardiomyopathy (1p, AD)
Familial dilated cardiomyopathy (AD, AR)
Familial dilated cardiomyopathy (dystrophin promoter, X-linked)
Arrhythmogenic right ventricular dysplasia (AD)
Noncompaction of the left ventricle (AD)
Mitochondrial transfer RNA (tRNA) mutations
T9997C tRNA Gly
C3303T tRNA leu
Restrictive cardiomyopathy
Familial restrictive cardiomyopathy (AD) (R)
Neuromuscular diseases
Duchenne muscular dystrophy (X-linked)
Becker-type muscular dystrophy (X-linked)
Myotubular myopathy (X-linked, also AR, AD)
Nemaline rod myopathy (AD, AR) (H&D)
Multicore myopathy (AR, also possible AD) (D, H, R)
Friedreich ataxia (AD) (H&D)
Phytanic acid oxidation disorder (Refsum disease, AR) (H&D)
Metabolic disorders
Decreased Energy Production
Disorders of mitochondrial fat oxidation
Primary carnitine transport protein deficiency (AR) (H&D)
Primary carnitine palmitoyltransferase II deficiency (AR) (H&D)
Secondary carnitine deficiencies
(many causes, e.g. methylmalonic and isovaleric acidemias,
multiple ORT & fat acyl Co-A dehydrogenase deficiencies, Kearns-Sayre
syndrome, etc.) (H&D)
Disorders of pyruvate metabolism
Pyruvate dehydrogenase deficiency
(Leigh necrotizing encephalopathy syndrome, AR) (H&D)
Pyruvate carboxylase deficiency (Leigh syndrome, AR) (H&D)
Disorders of oxidative phosphorylation
Complex I (NADH-coQ reductase) (AR, mtDNA)
Complex III (reduced coQ-cytochrome c reductase, cytochrome b)
(AR, mtDNA) (H&D)
Complex IV (cytochrome c oxidase)
(Leigh syndrome variants, AR, mtDNA) (H&D)
Complex V (ATP synthetase)
(Leigh syndrome variants, AR, mtDNA) (H&D)
Combined respiratory chain deficiencies (H&D)

Lethal infantile histiocytoid cardiomyopathy (AR, mtDNA) (WPW)
Lethal infantile mitochondrial disease (mtDNA)
Mitochondrial transfer RNA (tRNA) mutations
(H&D) cardiomyopathy and myopathy syndromes (multiple
mutations in tRNAs for leu, iso, gly, glu, pro) (WPW)
MELAS syndrome (multiple tRNA leu mutations) (D, H, WPW)
MERRF syndrome (multiple tRNA lys mutations) (D, H)
Kearns-Sayre syndrome (multiple tRNAs leu, asp, cys) (HB)
Mitochondrial DNA (mtDNA) mutations and delelions
Kearns-Sayre syndrome (multiple mtDNA mutations, AD) (HB)
others (e.g., 5 kb deletion, 7.4 kb deletion)
Barth syndrome (3-methylglutaconic aciduria type II, X-linked)
(H&D)
Infiltrative storage disorders
Glycogen storage disease
Type IV (Andersen disease, branching enzyme deficiency) (D)
Mucopolysaccharidosis
Type I (Hurler syndrome) (AR) (H&D)
Type VI (Maroteaux-Lamy syndrome) (AR)
Ganglioside degradation disorders
GM2 gangliosidosis (Sandhoff disease, AR) (H&D)
Amino acid and organic acid disorders with toxic metabolites
Propionic acidemia (AR)
Ketothiolase deficiency (AR)
D, dilated cardiomyopathy; H, hypertrophic cardiomyopathy; SVT,
supraventricular tachycardia; VT, ventricular tachycardia; WPW, WolfParkinson-White syndrome; HB, heart block; AD, autosomal dominant; AR,
autosomal recessive.
From refs. 33, 34.
Dilated Cardiomyopathies
Dilated cardiomyopathies are characterized by diminished cardiac contractility,
with ventricular enlargement, abnormal diastolic function, and congestive heart
failure. Neonates with dilated cardiomyopathy more frequently have an
identifiable cause than is currently achievable in older children and adults (see
Table 33-16). These include identifiable infection (e.g., bacterial or viral sepsis,
Coxsackie or adenovirus myocarditis, toxoplasmosis), ischemia (e.g.,
P.693
birth asphyxia, anomalous left coronary artery origin), hemodynamic work

overload (e.g., incessant tachyarrhythmia) and electrolyte or metabolic imbalance
(e.g., hypothermia, polycythemia, hypoglycemia, hypocalcemia). Increasingly
infants are being recognized with primary biochemical disorders of energy
production and metabolism that result in isolated cardiomyopathy or generalized
myopathy and encephalopathy (33,34 and 35). These infants often have
significant deterioration with stress, including that with birth. Caution should be
used in attributing permanent or temporary cardiac dysfunction and
encephalopathy entirely to birth asphyxia in a baby with low Apgar scores
without identifiable perinatal cause for asphyxia. Additionally, myocardial diseases
that more commonly
P.694
present in older children, such as those associated with neuromuscular disorders;
sometimes have unusually early presentations in infancy.
TABLE 33-17 HYPERTROPHIC CARDIOMYOPATHY
Hormonal causes
Maternal diabetes mellitus
In utero sympathomimetic exposure
Pheochromocytoma
Hyperthyroidism
Work overload
Severe pulmonary hypertension
Critical aortic valve stenosis
Genetic isolated cardiomyopathy
Contractile protein mutation
HCM-1 (myosin heavy chain, AD)
HCM-2 (troponin T2, AD)
HCM-3 (alpha-tropomyosin, AD)
HCM-4 (myosin binding protein C, AD)
HCM-5 (AD)
HCM-6 (AD) (WPW)
HCM (AR)
Cardiac phosphorylase kinase deficiency (AR)
Restrictive cardiomyopathy
Familial restrictive cardiomyopathy (AD) (R)
Genetic syndromes
Noonan syndrome (AD)
Cardio-facio-cutaneous syndrome (AD)
LEOPARD syndrome (AD)
Neurofibromatosis (AD)
Beckwith-Wiedmann syndrome (AD)

Cutis laxa (X-linked)
Neuromuscular diseases
Nemaline rod myopathy (AD, AR) (D&H)
Multicore myopathy (AR, also possible AD) (D, H, R)
Friedreich ataxia (AD) (D&H)
Refsum disease (D&H)
Metabolic disorders
Decreased energy production
Disorders of mitochondrial fat oxidation
Primary carnitine transport protein deficiency (AR) (H&D)
Primary carnitine/acylcarnitine translocase deficiency (AR)
Primary carnitine palmitoyl-transferase II (AR) (D&H)
Secondary carnitine deficiencies (H&D) many causes, e.g., organic
acidemias, multiple ORT & fat acyl Co-A dehydrogenase deficiencies, etc.
Very-long-chain acyl-CoA dehydrogenase (VLCAD) deficiency
Long-chain-acyl-CoA dehydrogenase (LCAD) deficiency
Long-chain 3-hydoxyacyl-CoA dehydrogenase (LCHAD) deficiency
Disorders of pyruvate metabolism
Pyruvate dehydrogenase deficiency (Leigh necrotizing
encephalopathy syndrome, AR) (H&D)
Pyruvate carboxylase deficiency (Leigh syndrome, AR) (H&D)
Disorders of oxidative phosphorylation
Complex II (succinate-coQ reductase) (AR) (WPW)
Complex III (reduced coQ-cytochrome c reductase, cytochrome b)
(AR, mtDNA) (H&D)
Complex IV (cytochrome c oxidase) (Leigh syndrome variants, AR,
mtDNA) (H&D)
Complex V (ATP synthetase) (Leigh syndrome variants, AR, mtDNA)
(H&D)
Combined respiratory chain deficiencies (D&H&WPW)
Lethal infantile histiocytoid cardiomyopathy (AR, mtDNA)
Lethal infantile mitochondrial disease (mtDNA)
Mitochondrial transfer RNA (tRNA) mutations cardiomyopathy and
myopathy syndromes (multiple mutations in tRNAs for leu, iso, gly, glu,
pro) (H&D&WPW)
MELAS syndrome (muliple tRNA leu mutations) (D&H&WPW)
MERRF syndrome (multiple tRNA lys mutations) (D&H)
Kearns-Sayre syndrome (multiple tRNAs leu, asp, cys) (HB)
Mitochondrial DNA (mtDNA) mutations and deletions
Kearns-Sayre syndrome (multiple mtDNA mutations, AD) (HB)
Barth syndrome (3-methylglutaconic aciduria type II, X-linked)

(H&D)
Sengers cardiomyopathy with cataracts syndrome (AR)
Infiltrative storage disorders
Glycogen storage disease
Type II (Pompe disease, acid maltase deficiency, AR)
Type III (Cori disease, debranching enzyme deficiency)
Type IX (Cardiac phosphorylase kinase deficiency)
Mucopolysaccharidosis
Type I (Hurler syndrome, AR) (H&D)
Type II (Hunter syndrome, X-linked)
Type III (Sanfilippo syndrome, AR)
Type IV (Morquio syndrome, AR)
Type VII (Sly syndrome, AR)
Ganglioside degradation disorders
GM1 gangliosidosis (AR)
GM2 gangliosidosis (Sandhoff disease, AR)
Glycoprotein metabolic disorder
Carbohydrate-deficient glycoprotein syndrome (AR)
Others
Glycosphingolipid degradation disorder (Fabry, X-linked)
Globoside degradation disorder (Gaucher disease, AR)
Phytanic acid oxidation disorder (Refsum disease, AR) (D&H)
Tyrosinemia (AR)
D, dilated cardiomyopathy; H, hypertrophic cardiomyopathy; SVT,
supraventricular tachycardia; VT, ventricular tachycardia; WPW, WolfParkinson-White syndrome; HB, heart block; AD, autosomal dominant; AR,
autosomal recessive.
From refs. 33, 34.
Pathophysiology
Although the etiologies are diverse, in most dilated cardiomyopathies the clinical
course, pathophysiology, and some molecular mechanisms are similar. Myocyte
damage, from infection, cytokines, toxic metabolite, or energy deprivation from
metabolic block or ischemia, results in myocardial injury. This results in a
sequence of molecular and cellular changes with myocardial dysfunction, stunning,
apoptosis, necrosis, and interstitial fibrosis, leading to impaired systolic
contractility and diastolic compliance. Ventricular enlargement, because of FrankStarling phenomenon, and tachycardia partially compensate for diminished
systolic shortening fraction and support resting cardiac output, but use up reserve
in pump function. The impairment in diastolic compliance results in generalized
edema and in pulmonary venous engorgement with tachypnea. If cardiac function
worsens, resting cardiac output diminishes, and multi-system dysfunction results.
Clinical Findings
The primary manifestations are those of combined right and left heart congestive
failure, including diminished activity and feeding, hepatomegaly, tachypnea,
retractions, S3, variable systolic regurgitant murmur, and variable signs of low
cardiac stroke volume, including tachycardia, narrowed pulse pressure, diminished
brachial radial and pedal pulses, systolic hypotension, diminished perfusion, and
oliguria.
Chest radiograph shows cardiomegaly and pulmonary edema.
ECG has resting tachycardia, often diffusely diminished voltage amplitude,
sometimes diffusely increased voltage amplitude, and often diffuse repolarization
changes.
Echocardiographic examination demonstrates ventricular enlargement, generally
affecting the left ventricle more than right, and often very severe. Mitral and
tricuspid regurgitation, atrial enlargement, and pulmonary hypertension occur
frequently. Coronary anomalies, or other cardiac structural anomalies with similar
presentation, should be detected.
Because the best hope for successful treatment depends on treating the primary
cause, diagnostic evaluation should seek causation. Detailed family history can
provide information not elicited by cursory generalized questions. Obstetrical
history may provide information about possible infectious causes and events with
asphyxia. Physical examination may demonstrate malformations consistent with
genetic syndromes, dysmorphic features and organomegaly consistent with
peroxisomal or infiltrative storage disorders, encephalopathy and hypotonia
consistent with various metabolic disorders and less often with neuromuscular
disorders. The absence of noncardiac findings also provides diagnostic
information. Although the history and exam may help point the direction, the
diagnosis depends on laboratory studies. Initial evaluation should usually include
blood electrolytes with measurement of total CO2 or bicarbonate, glucose, blood
urea nitrogen, creatinine, and complete blood count. If infection is suspected,
appropriate bacterial cultures (blood, endotracheal tube aspirate, urine,
cerebrospinal fluid), viral cultures (nasopharyngeal, perirectal, cerebrospinal fluid)
and serology should be obtained.
Specific Dilated Cardiomyopathies

Although rare, anomalous origin of the left coronary artery from the pulmonary
artery should be considered in all children with dilated cardiomyopathy,

particularly if there is an ECG pattern of anterolateral myocardial infarction. The
anomalous origin of the left coronary artery from the pulmonary artery and
retrograde flow in the left anterior descending and left main coronary arteries can
usually be seen on echocardiographic examination, although angiography may be
needed in some cases. Recognition is key; treatment is surgical, and is usually
successful if done promptly after ventricular dysfunction and symptoms appear.
Metabolic disorders can present with either dilated or hypertrophic
cardiomyopathy, and often clinically deteriorate with intercurrent infection. If
metabolic disease is suspected, additional tests should be done; including
measurement of blood ammonia, arterial blood gases, total and free carnitine,
lactate, pyruvate, liver function tests, creatine kinase and urine quantitative
amino acids, organic acids and, if appropriate, mucopolysaccharide and
oligosaccharide levels. Expert consultation from a metabolic disease specialist
should be obtained. Chromosomal analysis and skeletal x-ray analysis may be
helpful if dysmorphic features are present. Fundoscopic examination for retinal
disease and cataracts may help in evaluation for disorders accompanied by these
findings. Biopsy of skeletal muscle and/or cardiac muscle for light and electron
microscopic examination, mitochondrial genomic studies and biochemical studies
are often needed in the evaluation of metabolic disease (33).
Neonatal viral myocarditis is an often fulminant disease, frequently associated
with hepatitis and encephalitis. The most commonly identified causes are
echovirus, Coxsackie virus, particularly type B, and in some locations rubella
virus. In individual cases, the cause may not determined with culturing of
nasopharyngeal, tracheal, and stool swabs and serologic tests and PCR analysis
may be required for diagnosis. The infection may be acquired perinatally or
postnatally. Treatments including immunoglobulin, interferon, steroids, and
ribavirin have been under investigation in biopsy-proven myocarditis, but
supportive measures are the mainstay of treatment. Several other viruses,
bacteria, mycoplasma, rickettsiae, spirochetes, and fungi rarely cause
myocarditis. Myocarditis, because of an autoimmune reaction, may occur with
maternal lupus erythematosus.
P.695
Maternal IgG Ro antibodies cross the placenta, bind to the fetal myocardium, and
may block conduction or cause cardiomyopathy. Steroids may be beneficial in this
disease.
Chronic supraventricular and ventricular tachycardia can lead to persistent
myocardial dysfunction with the picture of cardiomyopathy. Some incessant
supraventricular arrhythmias, such as ectopic atrial tachycardia and permanent
junctional reciprocating tachycardia can be relatively occult with heart rates of
180-210 per minute, but distinguished from the sinus tachycardia of other dilated
cardiomyopathies by abnormal P waves. Again, recognition is key; effective
treatment of the cardiomyopathy depends on treatment of the arrhythmia.
Treatment
General acute supportive treatment consists of correction of coexistent electrolyte,
calcium, and acid-base abnormalities, providing abundant dextrose intravenously
to support potentially jeopardized energy production, judiciously providing fluids
to maintain cardiac output although minimizing edema, supporting the myocardial
function with intravenous inotropic agents (e.g., dopamine, dobutamine,
epinephrine, amrinone) (Table 33-11), and using antiarrhythmic medications as
needed. (see Tables 33-18 and 33-19). Additionally, antibiotics, hyperventilation,
paralysis, sedation, and vasodilators may be employed. In cases with severe, but
presumably self-limited, cardiopulmonary failure refractory to conventional
therapy, venoarterial extracorporeal membrane oxygenation (ECMO) has been
used with success. Although serious complications continue to exist, ECMO has
become a standard treatment for critically ill neonates with self-limited
cardiopulmonary failure.
Chronic supportive therapy is aimed at general measures to maximize the

strength and longevity of cardiovascular performance, control symptoms
of congestive heart failure and control arrhythmias (see Tables 33-10 and
33-19). Optimizing outcome of those more severely effected is often best
when done with experience and close monitoring and adjustment.
Spironolactone, carefully adjusted doses of angiotensin-converting
enzyme (ACE) inhibitors, and judicious use of beta adrenergic receptor
blockers (e.g., Carvedilol) can produce significant benefit in hemodynamic
function and significantly improve survival. Renal function and potassium
levels should be monitored carefully with use of spironolactone and ACE
inhibitors, and their dose adjusted for gestational age, postnatal age, and
concomitant use. Beta-adrenergic receptor blockers, such as Carvedilol,
should be started at very low dose and increased slowly as tolerated
using standard regimens. ACE inhibitors and Carvedilol are orally
administered, and currently require pharmacy compounding from pills.
Digitalization should be carried out with caution and orally if possible in
infants with myocarditis, because they may be unduly susceptible to druginduced arrhythmias. Diuretics can help with symptoms related to
pulmonary and systemic edema, but do not appear to influence survival.
Cardiac transplantation should be considered if the course appears likely
to be fatal despite optimal use of stabilizing measures and treatment of
the primary disorder, if there is no irreversible severe dysfunction of other
organs, and the underlying primary disease is not likely to recur in the
transplanted heart or other organs.
TABLE 33-18 TACHYARRHYTHMIA DIAGNOSIS AND TREATMENT

AV Reciprocating
Type of Arrhythmia
Tachycardia
Ectopic Atrial
Tachycardia
Atrial Flutter
Atrial
Fibrillation
Ventricular
Tachycardia
Usual QRS in
Unchanged
Unchanged
Unchanged
Unchanged
Abnormal
arrhythmia
Onset and
termination
Sudden
Gradual
Sudden
Sudden
Sudden or
gradual
Fixed-rate
Yes
No
V varies, A
No
yes or no
A > V or 1:1
Automaticity
fixed
A > V or 1:1
Reentry
A>V
Reentry
V > A or 1:1
Reentry or
tachycardia
A:V relationship
Mechanism
May respond to
vagal maneuvers
May respond to
adenosine
May respond to
1:1
Reentry
Yes
Rarely
Rarely
No
automaticity
Rarely
Yes
Rarely
No
No
Rarely
Yes
No
Yes
No
No
Yes
Yes
Yes, if reentry
esophageal pacing
May respond to DC Yes
No
countershock
Antiarrhythmic
Dig, Es, Pro, Proc, Es, Flec
Dig, Pro, Flec, Dig, Pro, Flec, Lido, Proc, Es,
agents for acute

Aden
Sota, Aden
Sota, Aden
Bret, Phen
therapy
Antiarrhythmic
Dig, Pro, Sota,
Pro, Flec, Sota, Dig, Proc, Q,
Dig, Pro, Q, y Pro, Proc, Q,
agents for chronic Flec, Proc, Q, oV, Amio
Flec, Sota,
Mex, Sota, Amio
therapy
Amio
Amio
A, atrial; V, ventricular; Dig, digoxin; Es, esmolol; Pro, propranolol; Proc, procainamide; Aden,
adenosine; Flec, flecainide; Sota, sotalol; Lido, lidocaine; Bret, bretylium; Phen, phenytoin; Q,
quinidine; oV, oral verapamil; Mex, mexiletine; Amio, amiodarone.
TABLE 33-19 NEONATAL ANTIARRHYTHMIC DRUGS
Drug (Class)
Adenosine
Currently
Commonly
Used
Yes
Route of
Metabolism/
Excretion
Oral Dose
Red blood cells

and vascular
endothelium
IV Dose
Therapeutic Level
0.0750.10 mg/
kg rapid IV push,
Indications
Contraindications
Rx-reentry SVT, DX
atrial flutter
Toxicity
Transient AV block,
HR, BP, and
flushing, rare atrial
fibrillation
to 0.150.25 mg/
kg after 1 min if
not effective
Digoxin
Yes
Renal
Load: 2030 g/ 80% of oral dose 0.82.2 ng/mL

kg divided in 3
doses,
Maintenance: 3
5 g/kg/12
SVT, atrial flutter
AV block, VT, many

WPW
hours; w/renal
hepatic
Quinidine (IA)
Yes
Hepatic
AV block, HR,
tachyarrhythmias,
vomiting, use w/
caution w/renal
failure, toxicity with
hypocalcemia
dysfunction
315 mg/kg/6
25 g/mL
SVT, WPW w/
Long QT, known
propranolol, PVC, VT sensitivity, IV use,

conduction block,
myasthenia gravis
hours; w/renal
or hepatic
dysfunction
Contractility, QT,
VT, conduction block,
vomiting, diarrhea,
rash, blood HR w/
atrial flutter w/o
digoxin level, need to
Procainamide
(IA)
Yes
Disopyramide
(IA)
No
Renal hepatic
Hepatic, renal
2.58 mg/kg/4
hours
a
7 mg/kg over 1
Procainamide 4 SVT, WPW, PVC, VT Conduction block,
myasthenia gravis
hour Infusion: 20 10 g/mL
60 g/kg/min
3.57.5 mg/
kg/6 hours
25 g/mL
SVT, WPW, PVC, VT Conduction block

myasthenia gravis
dose by
Similar to quinidine,
BP, lupus like
reaction, no effect on
level
Similar to quinidine,
CNS reactions,
seizure, BP,
respiratory drive
Lidocaine (IB)
Yes
Hepatic
Bolus: 1 mg/kg/
510 min
Infusion: 2050
g/kg/min; w/
cyanosis, hepatic
dysfunction
25 g/mL
PVC, VT
Conduction block
junctional and
ventricular escape
rate
Phenytoin (IB)
No
Renal, hepatic
23 mg/kg/12
Load: 10 mg/kg
hours; w/
over 3060 min

Maintenance:
hepatic
dysfunction
1020 g/mL
PVC, VT, digitalis

intoxication
Not FDA approved for CNS reactions, ?BP,

VT
blood dyscrasias,
hepatic
hypertrichosis,
same as oral
gingival hyperplasia,
coarse facies, rash
Flecainide (IC)
No
Renal, hepatic
0.32 mg/kg/8
0.21.0 g/mL
hours; w/
renal, hepatic
dysfunction
Esmolol (II)
Yes
RBC esterases
Refractory lifethreatening SVT,

PJRT, PVC, VT
Conduction block,
hepatic dysfunction,
myocardial
Occasional SVT
dysfunction
frequency w/WPW,
pacing threshold
conduction block, VT,
Recurrent,
sustained SVT,
WPW, VT
As per propranolol
nausea, contractility
As per propranolol
0.31. mg/
SVT, WPW, PVC,
Use w/verapamil,
kg/6 hours; w/
chronic
VT, hypertrophic
cardiomyopathy,
long QT
IV only
Load: 0.5 mg/kg

over 1 min
Infusion: 50100
g/kg/min 0.02
0.10 mg/kg over
20 min.
Propranolol (II)
Yes
Hepatic
cyanosis, renal,
hepatic
dysfunction
Sotalol (II/III)
No
Renal
?HR, conduction
bronchospasm,
block, bronchospasm,
conduction block, CHF BP, hypoglycemia,
depression, cardiac
reflexes w/anesthesia
2570 mg/m2
BSA/ 8 hours
SVT, WPW & VT

with structurally
Use with verapamil,

bronchospasm
As per propranolol
plus ventricular
normal heart
conduction block, ?
structural heart
arrhythemias
disease
Amiodarone (III) No
Hepatic
5 mg/kg/12
hours for 1
week; then 5
Load: 5 mg/kg
over 1530 min
Infusion: 1020
mg/kg/day; w/ g/kg/min
hepatic
12 mg/L
Refractory, lifethreatening, SVT,

VT, recurrent VF
Conduction block
Extremely long halflife, corneal deposits,

thyroid and hepatic
dysfunction,
pulmonary fibrosis,
may conduction
block and digoxin and
dysfunction
quinidine levels, need

to digoxin by,
hypotension w/IV
Bretylium (III)
No
Renal
Load: 5 mg/kg
over 15 min
Infusion:
Refractory VT, VF
Transient BP,
arrhythmia, then BP
Verapamil (IV)
No
Hepatic
24 mg/kg/8
Refractory SVT,
IV use, conduction
BP, HR,
hours; w/
hypertrophic
cardiomyopathy
dysfunction, CHF,
many WPW,
conduction block,
hepatic, renal
dysfunction,
neuromuscular
disease
some PVC, VT
myocardial
propranolol muscular depression,
dystrophy, use w/
constipation, may
quinidine, generally digoxin level, need to
avoid in infants
digoxin dose 1/3 to
To differentiate from metabolite measure by some laboratories.

Note: Continuous ECG monitoring should be done during initiation of antiarrhythmic therapy and with IV administration because of potential proarrhythmia and
conduction block.
AV arteriovenous; BP, blood pressure; BSA, body surface area; CHF congestive heart failure; CNS, central nervous system; DX, diagnose; FDA, Federal Drug
Administration; HR, heart rate; PJRT, permanent junctional reciprocating tachycardia; PVC, premature ventricular contractions; RX, treatment; SVT, supraventricular
tachycardia; VF, ventricular fibrillation; VT, ventricular tachycardia; w/, with; WPW, WolffParkinsonWhite syndrome.
P.696
P.697
P.698
Hypertrophic Cardiomyopathies
Hypertrophic cardiomyopathy is a disorder characterized by inappropriate
thickening of the ventricular walls, with normal, hyperdynamic or diminished
systolic performance and normal or diminished ventricular chamber size. There
are many causes (see Table 33-17). Most commonly, transient cardiomyopathy
occurs as secondary disorder in infants of diabetic mother, or with in utero
sympathomimetic exposure. Permanent myocardial diseases that more commonly
present in older children and adults, such as isolated hypertrophic cardiomyopathy
associated with contractile gene mutations, also present in infancy. Hypertrophic
cardiomyopathy may occur in infants in association with genetic syndromes (e.g.,
Noonan syndrome) and infiltrative storage diseases (e.g., Pompe and other
glycogen storage diseases, mucopolysaccharidosis). Primary metabolic disorders
of energy production (disorders of fatty acid oxidation, nuclear and mitochondrial
genome abnormalities in oxidative phosphorylation) cause isolated hypertrophic or
dilated cardiomyopathy, or multi-system dysfunction with cardiac and skeletal
myopathy and encephalopathy (33,34).
Pathophysiology
Although resulting from diverse etiologies, most hypertrophic cardiomyopathies
share certain clinical features and pathophysiology. Specific endocrine, genetic,
mitochondrial or metabolite abnormalities elicit a sequence of molecular and
cellular changes with myocardial sarcomeric production and often dysfunction,
resulting in myocyte thickening, often with interstitial fibrosis, leading to
ventricular wall thickening and impaired diastolic compliance. The ventricle walls
often become so thick regarding severely narrow ventricular outflow, particularly
dynamic with systolic ventricular wall thickening and result in the physiology of
aortic stenosis. Sometimes severe thickening of the ventricular walls significantly
encroaches and limits ventricular chamber size. Impaired diastolic compliance
results in generalized edema and in pulmonary venous engorgement. Metabolic
storage disorders, such as glycogen storage disease, have ventricular wall
thickening from infiltration of the myocardium with accumulated metabolic
products, not from deranged myocyte hypertrophy, but present with clinical
features of metabolic disease with hypertrophic cardiomyopathy.
Clinical Findings
There may be a history of maternal diabetes, exposure to sympathomimetic or
steroid medications, or family history of hypertrophic cardiomyopathy or metabolic
disease. A prominent systolic murmur from ventricular outflow stenosis and/or
mitral regurgitation are usually present. Signs of right and left heart congestive
failure, including diminished activity, hepatomegaly, tachypnea, retractions, and
S3, are often present. Signs of low cardiac stroke volume, including tachycardia,
narrowed pulse pressure, diffusely diminished pulses, and hypotension may be
present. Associated primary conditions may cause dysmorphic features,
hypotonia, encephalopathy, and organomegaly seemingly out of proportion to the
hemodynamic derangement.
Chest radiograph, in contrast to that seen at older age, often shows cardiomegaly
and pulmonary edema.
ECG usually has diffusely increased QRS voltage amplitude and repolarization
changes. Additionally, ECG in Pompe's disease characteristically shows a short PR
interval.
Echocardiographic examination demonstrates ventricular wall thickening, often
extraordinarily severe, variable left and sometimes right ventricular sub-valvar
outflow stenosis, and mitral and tricuspid regurgitation. Fixed anatomic sub-aortic
stenosis and other structural anomalies that can result in secondary ventricular
hypertrophy should be detected.
If there is no family history of known hypertrophic cardiomyopathy and initial
laboratory studies screening for metabolic disease are unremarkable, first-degree
relatives should be screened with electrocardiography and echocardiography for
asymptomatic hypertrophic cardiomyopathy. If physical examination or laboratory
studies suggest metabolic disorder, or if echocardiographic evaluation of first
degree relatives is negative or unobtainable, then additional laboratory evaluation
for metabolic disorder may be helpful (33).
Specific Hypertrophic Cardiomyopathies

Infants of diabetic mothers develop a hypertrophic cardiomyopathy that is
generally self-limited, although sometimes is a severe disorder. It results from the
myocardial trophic response to fetal hyperinsulinemia provoked by trans-placental
passage of high maternal glucose loads. Clinical findings include a systolic ejection
murmur, sometimes mild increase in respiratory rate and work and, rarely,
evidence of frank congestive heart failure. There is an increased risk of structural
heart disease in infants of diabetic mothers.
Echocardiographic examination reveals left ventricular hypertrophy that is
sometimes severe, usually with involvement of the septum, and occasionally with
outflow obstruction (Fig. 33-45). Treatment is supportive. Inotropes may worsen
outflow obstruction and are contraindicated if the obstruction is severe.

The most common permanent hypertrophic cardiomyopathies are genetic
disorders in one of the cardiac contractile proteins, most often myosin heavy
chain. In 50% of cases it is inherited as an autosomal condition with variable
penetrance, and in the others it occurs as a new mutation. This isolated cardiac
disorder is characterized by myocyte hypertrophy and disarray, propensity for
development of left ventricular outflow or intracavitary systolic pressure gradients,
symptoms of congestive heart failure caused by poor diastolic chamber
compliance, ventricular arrhythmias, and sudden death. There can be significant
progression with time, and a normal echocardiogram at birth does not exclude the
possibility for phenotypic expression later in life. Because it can be sub-clinical and
cause sudden death in older children and adults, echocardiograms
P.699
and electrocardiograms should be obtained in all first-degree relatives and
symptomatic relatives. Those presenting at birth appear to have the poorest
prognosis. Inotropic agents and diuretics are potentially harmful and generally not
used. Calcium channel blockers decrease the systolic pressure gradient, improve
diastolic compliance, and may improve survival in adults. Because of hazards
associated with calcium channel blockers in infants, their use in infants less than 1
year of age remains investigational. adrenergic receptor blockers can improve
symptoms, but do not appear to affect the progression of hypertrophy or survival.
Ventricular septal myomectomy may improve symptoms in those refractory to
medical treatment. Holter monitoring for ventricular arrhythmias should be
routinely performed, and amiodarone considered in those with ventricular
tachycardia. Cardiac transplantation may be required for survival in severely
affected, refractory patients.
Metabolic disorders of cellular energy production involving pyruvate metabolism,
fatty acid oxidation, and oxidative phosphorylation, and metabolic storage
disorders, such as glycogen storage disease, may present in early infancy with
clinical findings of isolated hypertrophic cardiomyopathy. However, these
disorders are particularly likely to be present when there is one or more
associated findings such as hypotonia, encephalopathy, cataracts, liver
enlargement or dysfunction disproportionate to the hemodynamics, hypoglycemia,
metabolic acidosis, elevated ketones, lactate or pyruvate. Treatment options
depend on the enzyme affected.
Figure 33-45 Echocardiographic parasternal long-axis view of a diabetic

mother's infant with severe hypertrophic cardiomyopathy and subaortic
ventricular septal defect. >, ventricular septal defect; AV, aortic valve, IVS,
interventricular septum; LA, left atrium; LV, left ventricle; RV, right ventricle.
Neonatal Tricuspid Valve Regurgitation

Regurgitation of an anatomically normal tricuspid valve is a common finding in
neonates. The regurgitation may be an isolated finding and mild, and cause only a
temporary soft regurgitant murmur at the left lower sternal border. Greater
degrees of regurgitation of an anatomically normal tricuspid valve often develop
with right ventricular myocardial disease, such as cardiomyopathy from asphyxia,
and with right ventricular systolic hypertension, as with persistent pulmonary
hypertension or pulmonary parenchymal disease. If there is right ventricular
diastolic dysfunction, there may be right-to-left shunting through a patent
foramen ovale. The explanation for a lower sternal murmur can be documented by
echocardiography and managed with observation, because tricuspid regurgitation
of this type tends to regress as the underlying right ventricular hypertension or
dysfunction resolves (see Fig. 33-15).
Chronic Pulmonary Hypertension and Cor

Pulmonale
Cor pulmonale is pulmonary hypertension secondary to ventilatory or pulmonary
disease, with right ventricular hypertrophy; it frequently presents as symptoms of
congestive heart failure with pulmonary disease. Patients with cor pulmonale have
a structurally normal heart, and must be differentiated from those with pulmonary
hypertension from intracardiac shunts or pulmonary venous obstruction. The most
frequent cause of cor pulmonale in infants has been bronchopulmonary dysplasia,

although its incidence has drastically declined in recent years and there are a
large number of other causes, including airway obstruction, central
hypoventilation, pulmonary hypoplasia, and diaphragmatic abnormalities.
Pathophysiology
Pulmonary artery pressure is controlled by vascular resistance, which is a function
of the number of small pulmonary arteries and the average luminal size. The
number of vessels may be diminished by congenital pulmonary hypoplasia or
acquired parenchymal damage. Additionally, abundant pulmonary artery
musculature sufficient to prevent more than minimal pulmonary flow before birth,
is present at birth, and can persist and re-constrict in response to certain
circumstances, further aggravating pulmonary hypertension. Factors resulting in
persisting postnatal pulmonary hypertension include alveolar hypoxia, acidemia,
increased pulmonary venous pressure, polycythemia and bacteremia. When
pulmonary resistance is fixed, increases in cardiac output with infection or anemia
also elevate pulmonary artery pressure. Pulmonary hypertension induces right
ventricular hypertrophy and, if severe, causes right ventricular diastolic
dysfunction and dilation. In addition to pulmonary hypertension,
bronchopulmonary dysplasia also is frequently associated with systemic
hypertension and left ventricular hypertrophy. Shifting of the ventricular septal
position, elevated ventricular pressure and biventricular hypertrophy, impair
biventricular diastolic compliance and increase sensitivity to intravascular volume.
The pulmonary hypertension and cardiac sequelae seen with chronic lung disease
generally resolve when the
P.700
pulmonary parenchymal disease resolves. In contrast, permanent pulmonary
vascular change (i.e., Eisenmenger syndrome) occurs in patients with congenital
heart disease after 1 year or more of exposure to a large left-to-right shunt with
pulmonary artery hypertension.
Clinical Findings
Variable hepatomegaly and systemic venous congestion secondary to right atrial
hypertension are predominant findings. Elevated left atrial pressure, especially in
the presence of pulmonary parenchymal disease, may predispose to pulmonary
symptoms and rales. The right ventricular impulse may be increased and the
second heart sound loud and appear single. There may be a relatively soft
murmur from tricuspid regurgitation, but a prominent murmur is not common and
suggests possible congenital heart disease. Cyanosis (See Color Plate) may result
from alveolar dysfunction with intrapulmonary shunting and from right-to-left
shunting across a patent foramen ovale. Infants with bronchopulmonary dysplasia
appear to be at increased risk for sudden death (10), similar to older patients with
primary pulmonary hypertension. The degree of pulmonary artery hypertension
and cardiac symptomatology may vary with labile ventilatory or pulmonary
conditions. Several echocardiographic methods, some fairly accurate, some not so
accurate, and most applicable to particular circumstances, can be used to assess
the right ventricular pressure (see Fig. 33-14). Sedation, sometimes necessary for
technically acceptable studies in older, active babies, may depress ventilation and
should be used with caution.
In patients with pulmonary hypertension, even those with lung disease, the
presence of congenital heart disease should be sought. The absence of a loud
murmur does not exclude a septal defect or patent ductus arteriosus in the
presence of elevated pulmonary vascular resistance. Because many of the
symptoms of cor pulmonale overlap those of congenital heart disease,
echocardiographic examination has been useful for excluding occult cardiovascular
lesions that could lead to irreversible Eisenmenger syndrome. However, with lung
disease ultrasound windows are often poor. These limitations should be realized,
and the results correlated with other findings.
Treatment
The treatment of pulmonary hypertension secondary to pulmonary or ventilatory
disease is directed primarily at the underlying disorder. Bronchodilators and
diuretics may lessen exacerbating abnormalities in alveolar gas exchange and
symptoms. Oxygen may be useful not only to prevent cyanosis and exacerbations
of pulmonary hypertension, but also as a pulmonary vasodilator. Digitalis has not
been found to be beneficial in most infants with congestive symptoms. With
resolution of the underlying ventilatory disease, the pulmonary hypertension also
resolves.
Cardiac Tumors
Intracardiac tumors in neonates are rare. Rhabdomyoma accounts for most.
Fibromas occur much less frequently, and other types occur rarely. Most babies
with cardiac rhabdomyoma have tuberous sclerosis (see Table 33-3), and vice
versa (94). The presence of one should prompt an investigation for the other.
Cardiac rhabdomyoma may be the only manifestation of tuberous sclerosis in
neonates. Neonatal cardiac rhabdomyoma are generally multiple and usually
regress, often completely. Cardiac tumors 2 mm or more in diameter are readily
demonstrated by echocardiography, even in the fetus. Many babies are
asymptomatic, even when the tumors are large and multiple, although peri-valvar
masses can obstruct valve flow and development. Serious arrhythmias also
sometimes occur.
Arrhythmias
All forms of cardiac arrhythmias can occur in the fetus or newborn. Those most
commonly encountered include sinus tachycardia and bradycardia, premature
atrial depolarizations, supraventricular tachycardia; and less commonly, atrial
flutter, ventricular arrhythmias, and complete heart block. Many arrhythmias are
benign, occurring in otherwise normal hearts, and are of no hemodynamic
consequence. Others may result in significant acute cardiovascular compromise,
particularly if they are very rapid or there is coexistent structural or functional
heart disease. Sustained tacharrhythmia can lead to reversible dilated
cardiomyopathy (95). Thus, in evaluating patients with arrhythmias it is important
to assess the hemodynamic status and cardiac structure and function. Rarely,
arrhythmias are the presenting sign of underlying cardiac abnormality such as
cardiomyopathy, Ebstein anomaly, or l-transposition of the great arteries.
Arrhythmias may also result from noncardiac disease; in neonates, ventricular
tachycardia, ventricular fibrillation, sinus arrest, and extreme bradycardia usually
occur in association with preceding severe hypoxemia, hypotension, acidosis,
electrolyte disturbance, or drug toxicity (e.g., digitalis).
Benign Arrhythmias
Sinus Bradycardia
Many infants have transient bradycardia associated with specific activities such as
crying, straining or micturition. Some healthy infants persistently have a heart
rate near 80 beats per minute. Sustained bradycardia at less than 70 beats per
minute in neonates is abnormal. Noncardiac causes, such as gastroesophageal
reflux leading to vagal stimulation, are common. Bradycardia can also be
produced by stimulation of the vagus nerve during procedures such as
endotracheal, nasogastric and orogastric intubation. Less commonly electrolyte
abnormalities, hypothyroidism and exposure to medications (e.g., prenatal betaadrenergic blockers) are the
P.701
cause. Isolated persistent sinus bradycardia should prompt careful inspection of
an ECG for nonconducted P waves as can occur with nonconducted atrial
premature depolarizations, second-degree atrioventricular conduction block, and
congenital long QT syndrome (96).
Sinus Arrhythmia
Sinus arrhythmia is the most common cause of an irregular heart rate
and rhythm, that is more prominent at slower heart rates, such as in
older infants. It represents normal physiologic variability in the sinus rate,
in phase with respiration and other variables. The P wave morphology and
axis do not change. Once recognized, no further evaluation or treatment
is necessary.
Sinus Tachycardia
Sinus tachycardia occurs with serious illness, fever, hypovolemia, anemia,
anxiety or pain and sympathomimetic medications (e.g., dopamine,
dobutamine, isoproterenol, epinephrine, caffeine, theophylline, and
aminophylline), at heart rates up to 230 beats per minute in infants. With
a sinus tachycardia, there is more rate variability than seen in many
tachyarrhythmias and normal P waves precede the QRS complex, often
merged into the preceding T wave at heart rates greater than 180 beats
per minute (P waves positive in leads I, II, and aVF; negative in lead
aVR). Pathologic supraventricular tachyarrhythmias can be differentiated
from sinus tachycardia by usually faster rates, abnormal P wave axis or
PR interval, and (when present) by an abrupt onset and termination or
wide QRS complexes.
Atrial Premature Depolarizations

Premature depolarizations can originate from any conducting tissue. Atrial
premature depolarizations occur in up to 30% of newborns (97). The
diagnosis is reliably assigned when there is an identifiable, early,
nonsinus P wave. However, the P wave may be hidden within the
preceding T wave. Atrial premature depolarizations may be conducted to
the ventricles normally, or with a bundle branch block pattern resulting in
a wide QRS complex (if a bundle branch is refractory from the preceding
beat), or may be blocked and not conducted to the ventricles (when
very early and occurring when the AV node or proximal His bundle is
refractory). When frequent, blocked atrial premature depolarizations
result in ventricular bradycardia as a result of resetting of the sinus node
with each premature atrial depolarization. In neonates with central

venous catheters, frequent atrial premature depolarizations may be as a
result of contact of the catheter with an atrial wall, and constitute an
indication to withdraw the catheter from the atrium. Atrial premature
depolarizations might also be secondary to electrolyte abnormalities or
drugs (e.g., dopamine, dobutamine, isoproterenol, epinephrine, caffeine,
theophylline, and aminophylline). Very rarely are atrial premature
depolarizations secondary to myocarditis or cardiac tumors.
In hemodynamically stable otherwise healthy infants, premature atrial
depolarizations generally do not warrant further evaluation. Isolated
ectopic atrial depolarizations and atrial bigeminy are only rarely
associated with tachycardia, and usually are of no serious consequence.
In most infants, these arrhythmias resolve over a few months.
Ventricular Premature Depolarizations

Premature ventricular depolarizations are early QRS complexes with a
morphology different from sinus beats and without an identifiable
preceding P wave. In newborns, ventricular premature depolarizations
may not be much wider than normal QRS complexes. Premature
ventricular complexes occur in less than 1% of healthy newborns (98).
Although premature ventricular depolarizations are usually benign, their
identification should prompt evaluation for possible structural or
functional heart disease, electrolyte abnormalities (e.g., hypokalemia,
hyperkalemia, and hypocalcemia), hypoglycemia, hypoxia, or the
congenital long-QT syndrome. Premature ventricular complexes may also
be related to administration of drugs (e.g., dopamine, dobutamine,
epinephrine, caffeine, aminophylline, theophylline, digoxin, or other
antiarrhythmic medications). In the absence of these problems, they
often resolve over a number of months. There are no data to suggest that
the daily number or morphology of these complexes influences the
prognosis.
Accelerated Ventricular Rhythm

An accelerated ventricular rhythm is a less common arrhythmia. There is a wide QRS rhythm, generally at a rate
not more than 10% greater than the underlying sinus rate. This probably represents enhanced automaticity of a
ventricular focus. There may be mild accelerations and decelerations of the rate. Atrioventricular dissociation is
usually seen. The duration of episodes is variable. Although this arrhythmia generally occurs in otherwise
healthy infants, it has been associated with structural heart disease, electrolyte abnormalities, cardiac tumors,
intracardiac catheters, maternal heroine and cocaine use, and respiratory distress. Patients are usually
asymptomatic and do not need treatment. This arrhythmia generally resolves within months (99).
Tachyarrhythmias
Heart rate alone is not always enough to establish the diagnosis of a pathologic tachycardia. Infants can have a
sinus tachycardia with rates up to at least 230 beats/min in response to serious illnesses, fever, hypovolemia,
anemia, pain or infusion of inotropic/chronotropic agents. Additionally, some unusual pathologic supraventricular
tachycardias can have rates less than 180 beats/min. In assessing a child with a fast heart rate, one should
determine if the
P.702
QRS complex is narrow or wide during tachycardia, if the rate is fixed or variable, if there is a visible P wave and
if so, the P wave axis. Supraventricular tachycardias typically have a narrow (normal) QRS complex. In general,
if the QRS complex in tachycardia remains wide, the rhythm should be considered ventricular tachycardia.
However, it is not uncommon for the first few beats of SVT to be wide because of aberrant conduction (right or
left bundle branch block) before changing to a narrow QRS complex.
A patient with an apparently fixed high heart rate should be carefully assessed. The fixed heart rate could
represent a sinus tachycardia secondary to a high-catecholamine state in an otherwise sick infant. An
electrocardiogram should be obtained, and the P wave morphology clearly established. If there is no clear P
wave, or the P wave does not have a sinus morphology (positive in leads I, II, and aVF; negative in lead aVR)
one should strongly consider a pathologic tachycardia or structural heart disease with heterotaxy (see Table 3317).
All neonates with documented tachyarrhythmias should have a complete cardiac evaluation, including 12-lead
ECG (during the tachycardia if hemodynamically stable, and later in sinus rhythm), and echocardiography, to
assess cardiac structure and function. It is estimated that between 8% and 25% of infants with SVT have
structural heart disease, most often Ebstein malformation of the tricuspid valve, corrected transposition of the
great arteries, or hypertrophic cardiomyopathy (100). Cardiac tumors and myocarditis rarely are predisposing
causes for ventricular arrhythmias.
The clinical status of infants with tachyarrhythmias depends on the ventricular rate, duration of tachycardia,
presence of underlying structural or functional heart disease, and other clinical problems. Patients may be
completely asymptomatic, with the arrhythmia noted during an otherwise routine evaluation or although
monitored for other reasons. The infant may not have been appearing well, with irritability, poor feeding,
restlessness, or respiratory difficulty with tachypnea, retractions and wheezing. With persistent tachyarrhythmia,
the child may develop signs and symptoms of congestive heart failure or acidosis, becoming pale and listless. If
the tachycardia persists for long enough, heart failure and a secondary dilated cardiomyopathy may develop. In
the fetus with persistent or recurrent tachycardia, this is manifest as nonimmune hydrops. Whether a
tachycardia related cardiomyopathy develops depends on the ventricular rate, whether the tachyarrhythmia is
intermittent or incessant, the frequency of recurrences if intermittent, and the presence of structural heart
disease.
Figure 33-46 Electrocardiogram during conversion of supraventricular tachycardia to sinus rhythm with
administration of adenosine. During tachycardia at a rate of 230 beats/min, there is a normal-appearing QRS
complex without a delta wave (no ventricular preexcitation), and there is no distinct P wave. After conversion to
sinus rhythm, there is a short PR interval (80 milliseconds) and wide up-sloping QRS complex (90 milliseconds)
representing ventricular preexcitation, indicative of the Wolff-Parkinson-White syndrome.
Supraventricular Tachycardias
Atrioventricular Reciprocating Tachycardia and Wolff-ParkinsonWhite Syndrome
The most common type of fetal and neonatal supraventricular tachycardia is atrioventricular reciprocating
tachycardia with a reentry circuit via an accessory atrioventricular conduction pathway (101). Most often these
accessory conduction pathways persist as isolated anomalies, but they can occur in association with cardiac
structural anomalies such as Ebstein anomaly. Accessory atrioventricular conduction pathways resulting in WPW
syndrome are rarely inherited as an autosomal dominant trait, sometimes associated with hypertrophic
cardiomyopathy (102).
During supraventricular tachycardia there is antegrade conduction from the atria to the ventricles over the AV
node and His-Purkinje system, with retrograde conduction over the accessory pathway from the ventricles to the
atria. Because antegrade conduction is normal over the AV node and His-Purkinje system, there usually is a
normal (narrow) QRS complex during tachycardia. Commonly the accessory atrioventricular conduction pathway
conducts only retrograde. Therefore, during sinus rhythm all antegrade conduction is through only the AV node,
and the electrocardiogram appears normal.
Atrioventricular reciprocating tachycardia is characterized by the abrupt onset and termination of a fairly fixed
heart rate of 230 to 300 beats/min, abnormal or unidentifiable P waves that may be superimposed on the T
waves, and usually a normal QRS morphology (see Fig. 33-46 and Table 33-18). The infant may be
asymptomatic initially, but then becomes irritable and fussy and refuses feeding, congestive heart failure
develops in approximately 20% after 36 hours and in 50% after 48 hours.
P.703
Some accessory pathways can conduct impulses in both directions, resulting in the Wolff-Parkinson-White
(WPW) syndrome. During sinus rhythm, antegrade atrial-to-ventricular conduction over the accessory pathway
results in a short PR interval and wide QRS complex with characteristic delta wave, resulting from rapid ectopic
depolarization of the ventricles, without the delay in conduction that occurs in the A-V node. Depending on the
location of the accessory pathway, and the ECG lead examined, a delta wave may be of variable size and
direction. Because antegrade conduction during supraven tricular tachycardia is usually through the AV node and
His-Purkinje system, the diagnosis of WPW syndrome cannot usually be made during tachycardia. Rarely, an
atrioventricular reciprocating tachycardia occurs in patients with WPW syndrome with antegrade conduction over
the accessory pathway, and retrograde conduction through the A-V node, resulting in a wide QRS tachycardia,
which may be electrocardiographically indistinguishable from ventricular tachycardia. Some patients with WPW
syndrome have the potential for very rapid antegrade conduction over their accessory pathway; if those
individuals develop a primary atrial tachycardia such as atrial fibrillation, it can lead to ventricular fibrillation
(103). Fortunately, atrial fibrillation is rare in children. Patients with WPW syndrome should generally not be
treated with digoxin or verapamil, as these medications sometimes enhance antegrade accessory pathway
conduction.
Persistent Junctional Reciprocating Tachycardia

One important, but unusual form of accessory pathway has slow retrograde conduction, resulting in a
supraventricular tachycardia referred to as persistent junctional reciprocating tachycardia. These tachycardias
are usually slower than other supraventricular tachycardias, often less than 200 beats/min in newborns. This
often causes an incessant or frequently recurring tachycardia that can result in a reversible dilated
cardiomyopathy. Initially this arrhythmia is often well tolerated because of the slower heart rate. It can be
recognized by having a fixed rapid heart rate in which there is an abnormal (nonsinus) P-wave axis (usually
negative in leads II and aVF, and positive in aVR and aVL). There may be a normal appearing PR interval as a
result of slow retrograde conduction, but the abnormal P-wave axis differentiates it from sinus tachycardia. This
type of supraventricular tachycardia might terminate with vagal maneuvers, adenosine, atrial pacing or
cardioversion, but frequently the arrhythmia rapidly recurs.
Ectopic Atrial Tachycardia

Another potential cause of a dilated cardiomyopathy is an ectopic atrial tachycardia (104). This type of
supraventricular tachycardia results from enhanced automaticity of a single cell or small cluster of cells in either
atrium. Generally, the P-wave morphology is not normal in at least one lead. The arrhythmia mechanism does
not involve the ventricles. Depending on the atrial rate and possible AV block, there may be a variable A:V
relationship (with more atrial than ventricular complexes). This type of SVT is characterized by a variable rate
and gradual onset and termination (see Table 33-17). Transiently blocking A-V conduction, with vagal
maneuvers or adenosine administration, will not terminate the atrial tachycardia but will demonstrate the
diagnosis by the presence of unabated rapid P waves. Ectopic atrial tachycardia also does not terminate with
pacing maneuvers or cardioversion, but often improves with beta-adrenergic receptor blockers.
Junctional Ectopic Tachycardia

Junctional ectopic tachycardia is a rare, often incessant, tachyarrhythmia due do to enhanced
automaticity in the A-V node or bundle of His. There is generally a normal (narrow) QRS
morphology with variable ventricular rates. If there are visible P-waves, the ventricular rate is
greater than the atrial rate and it does not respond to pacing maneuvers or cardioversion, and is
often difficult to control with milder antiarrhythmic medications.
Isolated junctional ectopic tachycardia is often familial. Some patients eventually develop
complete heart block (105).
Most often junctional ectopic tachycardia is seen early, and temporarily following cardiac surgery
in infants, and can result in significant hemodynamic compromise until it is controlled and abates.
Intravenous amiodarone may be the best treatment; but cooling the patient down to 33 to 35 C
with administration of intravenous procainamide is often helpful.
Atrial Flutter and Fibrillation

Atrial flutter is less common than other types of paroxysmal supraventricular tachycardia in
fetuses and neonates. It may be idiopathic, or associated with the same congenital heart lesions
as those producing other supraventricular tachycardias. The atrial rate may be 200 to 500 beats/
min. The AV node is generally not part of the tachycardia circuit, so that there need not be a 1:1
atrial:ventricular rate relationship. There is often some degree of AV node block resulting in
variable conduction, frequently with a 2:1 or 3:1 atrial:ventricular relationship. With a 2:1 block,
the ventricular rate at the highest atrial rate would be 250 beats/min, a rate sufficient to produce
congestive failure in infancy. The RR interval is constant except when the atrioventricular block
changes. The rare infant without AV nodal conduction block may have a very rapid rate and
shock. With higher degrees of AV nodal conduction block, a sawtoothed atrial pattern is
characteristically seen, often best in leads II or V1. In some patients the diagnosis can not be
ascertained from the surface electrocardiogram, especially if there is a 1:1 atrial:ventricular rate
relationship. Esophageal recordings may demonstrate the atrial
P.704
activity more clearly, or adenosine might be used to cause transient AV block and demonstrate
the flutter waves. Adenosine, although diagnostically helpful, does not convert atrial flutter (Fig.
33-47). Digoxin can be used to slow the ventricular rate, but is unlikely to convert the atrial
flutter. Overdrive atrial pacing from the esophagus and, if necessary, cardioversion by DC
countershock (initial dose 0.5 to 1 Joule/Kg) can be used to convert the rhythm to sinus. In the
absence of structural heart disease, there is generally a benign course once the arrhythmia is
converted to sinus rhythm, and subsequent chronic antiarrhythmic therapy is often not necessary
(106).
Figure 33-47 Electrocardiogram showing the effect of adenosine on atrial flutter. Before
adenosine, there is atrial flutter with 2:1 A-V conduction. The flutter waves are difficult to
discern. After adenosine, there is transient slowing of AV node conduction without termination of
the atrial flutter. This allows the flutter waves to be readily identified, confirming the diagnosis.
The atrial rate during atrial flutter is 500 beats/minute.
Atrial fibrillation is recognized by an irregularly irregular ventricular rhythm. It is rare in newborns

and usually seen in patients with structural heart disease. Digoxin can be used to slow the
ventricular rate. Cardioversion by DC countershock (initial dose 0.5 to 1 J/Kg) is usually necessary
to convert a sustained episode.
Treatments for Supraventricular Tachycardias

Treatment depends in part on the mechanism of tachycardia (see Tables 33-18 and 33-19) (100).
In arrhythmias involving the AV node, such as atrioventricular reciprocating tachycardias without
or with the WPW syndrome, transiently slowing or blocking AV node conduction can terminate the
tachycardia. Vagal maneuvers are often effective in terminating these types of supraventricular
tachycardia. These include the application of ice or very cold damp cloth to the face and rectal
stimulation. Ocular compression should not be used as a result of risks to the child's vision.
Rapid bolus of intravenous adenosine (0.1 mg/kg, increased to 0.2 mg/kg if needed) is the
treatment of choice for most infants with supraventricular arrhythmias involving the AV node that
are refractory to vagal maneuvers. Adenosine is an endogenous nucleoside with a very short halflife that can transiently block AV node conduction, interrupting these arrhythmias, resulting in
abrupt conversion to sinus rhythm. It is helpful to record the patient's electrocardiogram during
attempts at conversion so that one can see if there was an effect of the intervention on the
arrhythmia. There also might be transient evidence of WPW syndrome as the AV node conduction
is briefly blocked, in the form of a delta wave from antegrade conduction down an accessory
conduction pathway. If there was transient termination with rapid resumption of the arrhythmia, a
longer-acting agent might be necessary to control the arrhythmia. Significant side effects appear
to be rare, but include initiation of atrial fibrillation. Rarely, in patients with WPW syndrome, this
can cause a ventricular tachycardia or fibrillation. It is therefore recommended that a defibrillator
be readily available when administering adenosine. Methylxanthines (e.g., theophylline and
caffeine) are competitive antagonists to adenosine, and might make dosing more difficult.
Other antiarrhythmic agents can be used to terminate and control supraventricular tachycardias in
infants, in which vagal maneuvers or adenosine have resulted in only a transient termination of
the arrhythmia, or for supraventricular tachycardias that do not involve the AV node (i.e., ectopic
atrial tachycardia). These might include blockers (i.e., esmolol or propranolol), class I
antiarrhythmic agents (i.e., procainamide or flecainide), digoxin, or class III antiarrhythmic

agents (i.e., sotalol or amiodarone). For ectopic atrial tachycardias, a -blocker is often effective,
and intravenous infusion of esmolol is a good first-line therapy until the arrhythmia is controlled.
However, class I or class III antiarrhythmic medications may be necessary to adequately suppress
ectopic atrial or junctional tachycardias. Intravenous verapamil has been associated with
cardiovascular collapse and death in neonates and infants and should not be used in patients
under 1 year of age (107).
If available, atrial pacing using an esophageal pacing catheter can be effective in terminating AV
reciprocating tachycardias and atrial flutter. Atrial pacing will not terminate ectopic atrial
tachycardias or junctional ectopic tachycardias, but the atrial catheter can be helpful in confirming
the mechanism of the tachycardia.
If the patient is hemodynamically unstable, DC countershock (starting with 0.5-1.0 Joules/kg)
should be attempted for known or suspected reentry-type supraventricular tachycardias, including
AV reciprocating tachycardia, atrial flutter, and atrial fibrillation.
Prophylactic antiarrhythmic therapy is prescribed for most infants with supraventricular
tachycardia, as there is approximately a 20% recurrence risk of atrioventricular reciprocating
tachycardias. Often the recurrence risk diminishes after the first 6 to 12 months, and the
medication may be discontinued (108).
P.705
Digoxin may be used in patients with atrioventricular reciprocating tachycardia without evident
WPW syndrome. The use of digoxin in patients with manifest WPW syndrome is controversial.
There is evidence that digoxin can sometimes enhance antegrade accessory pathway conduction,
potentially allowing very rapid conduction over the accessory pathway if the patient develops
atrial fibrillation, possibly causing a hemodynamically unstable ventricular tachycardia or
ventricular fibrillation.
adrenergic blockers such as propranolol or atenolol are often used in patients with
atrioventricular reciprocating tachycardias, especially with WPW syndrome and for ectopic atrial
tachycardias. Care may be necessary in using adrenergic blockers for infants with significant
lung disease. For more refractory patients, combination therapy with propranolol and digoxin may
be effective (101).
If breakthrough of supraventricular tachycardia occurs, switching to other agents with a greater
potency (and toxicity), under the direction of a pediatric cardiologist, may suppress recurrences.
These other agents may include drugs with a combination of actions such as sotalol, or others
used alone or in combination such as type IA agents (e.g., quinidine, procainamide,
disopyramide), type IC agents (e.g., flecainide or propafenone), or amiodarone, and possibly oral
verapamil in older infants. The recommended management schemes vary between institutions
and will change as more data about existing medications and new medications become available.
Esophageal or intracardiac electrophysiology studies with programmed atrial stimuli can be used
to determine probability of recurrence on medications or after medications have been
discontinued, for those with reentry supraventricular tachycardia. Rarely, radiofrequency ablation
is used in infants with particularly refractory supraventricular tachycardia, often with associated
ventricular dysfunction (109). Radiofrequency ablation is more safely and routinely applied in later
childhood in those with persistent problematic arrhythmias.
Fetuses with supraventricular tachycardia may be recognized in utero by an unusually rapid heart
rate. The mechanism of the arrhythmia can usually be elucidated with echocardiography, and
usually involves a reentry mechanism (Fig. 33-6). If the arrhythmia has been present for some
time, these fetuses may develop hydrops. Treatment in utero by the administration of digoxin,
flecainide, quinidine, procainamide, or sotalol to the mother, or procainamide or amiodarone
directly into the umbilical vein, may be effective in terminating or suppressing the arrhythmia.
Figure 33-48 Electrocardiogram showing ventricular tachycardia in an infant at a rate of 158

beats/min. Because of retrograde conduction block, occasional sinus beats can be conducted to
the ventricles through the AV node and His-Purkinje system, resulting in a normal QRS complex.
Note that the QRS complex duration of the ventricular tachycardia measures 0.08 seconds,
compared a QRS duration of 0.06 seconds for the conducted sinus beats.
Ventricular Tachycardias
Wide complex tachyarrhythmias may represent ventricular tachycardia, SVT with aberrant
conduction, atrioventricular reciprocating tachycardia in WPW syndrome with antegrade
conduction over the accessory pathway and retrograde conduction over the His-Purkinje system
and AV node, or any tachycardia in the presence of bundle branch or intraventricular conduction
block (e.g., after cardiac surgery, hyperkalemia). In infants, the QRS complex in ventricular
tachycardia may be as narrow as 0.06 to 0.11 seconds, but it is always different than in sinus
rhythm (Fig. 33-48).
Although dissociation of the ventricles from the atria, with more ventricular beats, is a hallmark of
ventricular tachycardia (ventricular:atrial ratio > 1), there may be a 1:1 ventricular:atrial
relationship if there is retrograde conduction over the A-V node. For the purposes of initial
management, a wide QRS complex tachycardia should be considered ventricular tachycardia
unless one has an ECG in sinus rhythm showing the same, wide QRS complex (generally as a
result of a fixed bundle branch block). Chronic recurrent ventricular tachycardias are rare in
neonates, and are usually associated with structural or functional heart disease. When there is a
polymorphic ventricular arrhythmia in the absence of apparent structural or functional heart
disease, one should consider torsades de pointes associated with the long-QT syndrome. Very
rarely an incessant, monomorphic ventricular tachycardia is found in infants with cardiac tumors
(110).
Long-QT syndrome is a generally congenital abnormality of ventricular repolarization that can
cause ventricular tachycardia. The congenital long QT syndrome has recently been subdivided into
7 genotypes, caused by mutations in 1 of 6 proteins involved with transmembrane ion currents or
in an associated structural anchoring protein (111). It can be inherited in an autosomal dominant
fashion (Romano-Ward syndrome, without sensorineural hearing loss), or less often as an
autosomal recessive condition (Jervell and Lange-Nielsen syndrome, with sensorineural hearing
loss). Patients with this syndrome, including infants, generally have a prolonged
P.706
corrected QT interval (QT (RR) > 0.46 msec). Rarely, infants with the long QT syndrome have
2:1 A-V conduction block because the ventricles are refractory when the next atrial impulse
conducts through the A-V node and His-Purkinje system (96). Patients with the long QT syndrome
are at risk for torsades de pointes (a polymorphic ventricular tachycardia characterized by
changing QRS morphology/axis that appears to turn around the baseline point) and sudden death.
To help aid in the diagnosis of the long QT syndrome, a scoring system has been developed based
on ECG findings, the patient's arrhytmia (ventricular tachycardia) and symptoms (syncope,
aborted sudden death) and family history (112). Care must be made in diagnosing the long QT
syndrome in the first few days of life, especially without a family history of long QT syndrome or
documented torsades de pointes, because many neonates have a QTc greater than 440 msec in
the first few days of life.
Patients identified with the long-QT syndrome should be started on -blockers, although other
antiarrhythmic medications might eventually be preferable for some genotypes. Patients with
malignant ventricular tachycardia despite -blockers occasionally require cervicothoracic
sympathectomy, pacemakers, or automatic implantable cardiac defibrillators (113).
Acquired long QT syndrome with risk of torsades de pointes can occur secondary to medications,
electrolyte abnormalities (e.g., hypokalemia, hypocalcemia, or hypomagnesemia), endocrine
abnormalities (e.g., hyperparathyroidism, hypothyroidism, or pheochromocytoma), or CNS
disorders (e.g., encephalitis, head trauma, subarachnoid hemorrhage).
Treatments for ventricular Tachycardias

Because ventricular tachycardia can be life threatening, with the potential for degeneration into
ventricular fibrillation, it is safest to treat all wide complex tachycardias as ventricular tachycardia.
If the patient is hemodynamically stable with a monomorphic ventricular tachycardia,
antiarrhythmic therapy using lidocaine, procainamide, or amiodarone may suppress the
arrhythmia (see Table 33-19). Phenytoin can be especially useful if the arrhythmia is related to
digoxin toxicity. For torsades de pointes, treatment should include magnesium sulfate, lidocaine,
and possibly isoproterenol or cardiac pacing.
If the patient is, or becomes, hemodynamically unstable, DC cardioversion should be initiated with
1 to 2 Joules/kg initially.
Chronic prophylactic antiarrhythmic therapy should be guided by the type of ventricular
tachycardia, any underlying heart disease, and hemodynamic alterations associated with the
arrhythmia (see Table 33-19). Radiofrequency ablation is rarely necessary for infants with
monomorphic ventricular tachycardia not adequately suppressed by antiarrhythmic medications.
Rarely, automatic implantable cardiac defibrillators are necessary for infants with recurrent lifethreatening ventricular arrhythmias despite antiarrhythmic medications.
Atrioventricular Block
First Degree Atrioventricular Block
First degree A-V block is characterized by an abnormally prolonged P-R
interval for age and heart rate. Most commonly, this is as a result of
enhanced vagal tone and seen during sleep. Rarely this can be secondary
to antiarrhythmic medications (e.g., digoxin), hypothermia, increased
parasympathetic tone, hypothyroidism, or electrolyte disorders (hypo- or
hyperkalemia, hypo- or hypercalcemia, hypoglycemia, and
hypomagnesemia). First-degree AV block is generally well tolerated and
requires no specific therapy. However, an infant with a muscular
dystrophy, Kearns-Sayre syndrome, neonatal lupus, a family history of
complete heart block or maternal connective tissue disease should be
followed for possible progression of the conduction abnormality.
Second-Degree Atrioventricular Block

Second-degree AV block is defined as intermittent failure of conduction of some
atrial depolarizations to the ventricles. Second-degree AV block is subdivided into
Mobitz type I (Wenckebach) and Mobitz type II AV block. In Mobitz type I
(Wenckebach) AV block there is progressive delay in A-V conduction before a
single atrial beat fails to conduct. Mobitz type I AV block is generally caused by
factors similar to those causing first degree A-V block; it is usually well tolerated,
and rarely causes symptomatic bradycardia. Mobitz Type II AV block is recognized
by the intermittent failure of atrioventricular conduction, without an associated
prolongation of the P-R interval. There is often a fixed ratio of atrial to ventricular
depolarizations, although this ratio may vary over time. Asymptomatic patients
with isolated Mobitz Type II AV block do not require a permanent pacemaker, but
should receive close follow-up.
Third-degree (Complete) Atrioventricular Block

Third-degree (complete) atrioventricular (AV) block is defined by the absence of
conduction of atrial impulses to the ventricles. The ventricular rhythm may arise
from the AV node or His bundle (junctional escape rhythm), in which case there is
usually a narrow QRS complex. Alternatively, the ventricular rhythm may arise
from the ventricles, resulting in a wide QRS complex. The ventricular rate is
related to the origin of the escape rhythm, with AV junctional escape rhythms
generally being faster than idioventricular rhythms.
Congenital complete heart block is frequently recognized in utero. It is estimated
to occur in 1 of 20,000 live
P.707
births. In approximately 50% of infants with congenital heart block, there is an
associated cardiovascular malformation (e.g., l-transposition of the great arteries,
heterotaxy syndrome, endocardial cushion defect). In the absence of structural
heart disease, the heart block is usually related to maternal autoantibodies (antiRo and/or anti-La) that cross the placenta and interact with the developing
conduction system (114). These antibodies are associated with maternal
connective tissue disease, particularly lupus erythematosis and Sjgren syndrome.
In newborns with heart block and no structural cardiac abnormalities, testing of
their mothers for these antibodies is indicated because they may have no signs or
symptoms of connective tissue disease.
Occasionally newborns present with second-degree AV block and later progress to
complete heart block. Therefore, those infants with persistent second-degree
heart block (especially Mobitz type II) should undergo echocardiography to look
for structural heart disease, and their mothers should be tested for these
antibodies. Rarely myocarditis and fibrosis of the atrioventricular node or His
bundle has been implicated as the cause of neonatal heart block. Complete heart
block may occur as a complication of cardiac surgery, particularly in the correction
of l-transposition, endocardial cushion defects, tetralogy of Fallot, and ventricular
septal defects.
Infants with congenital complete heart block generally have a narrow, junctional,
escape rhythm between 30 and 110 beats/min. The ventricular rate tends to
decrease with increasing age. The stroke volume increases to compensate for the
low ventricular rate, to maintain an adequate cardiac output. Examination usually
reveals cardiac enlargement from an increased left ventricular end-diastolic
volume, and a systolic ejection murmur and apical mid-diastolic rumble from
increased stroke flow.
Most infants with isolated congenital complete heart bock without
hemodynamically significant cardiac malformation often tolerate the bradycardia
well, without symptoms, with normal growth and development, and do not require
immediate intervention. Symptoms are usually related to the severity of the
associated cardiovascular malformation and the degree of bradycardia. Less often,
an infant or fetus with isolated complete heart block develops congestive heart
failure or hydrops and some do not survive (115).
Patients with syncope or near syncope (Stokes-Adams attacks), congestive heart
failure, or postsurgical block require early initiation of permanent ventricular
pacing. Medical therapy with isoproterenol or with transcutaneous or transvenous
cardiac pacing is helpful in the acute situation, prior to surgery. The timing of
pacemaker implantation in other patients is more controversial. Criteria used to
select patients for permanent pacing in infants have included: awake, resting
ventricular rates less than 50 beats/min; awake, resting atrial rates greater than
140 beats/min; wide QRS escape rhythms; prolonged QT intervals; frequent,
complex ventricular ectopy; or ventricular tachycardia (116).
REFERENCES
1. Mitchell SC, Korones SB, Berendes HW. Congenital heart disease in 56,109
births. Circulation 1971;43:323.
2. Hoffman JIE, Kaplan S. The incidence of congenital heart disease. J Am Coll

Cardiol 2002;39:1890.
3. Fyler DC. Report of the New England Regional Infant Cardiac Program.
Pediatrics 1980;65(Suppl):375.
4. Hoyert DL, Freedman MA, Strobino DM, Guyer B. Annual summary of vital
statistics: 2000. Pediatrics 2001;108:1241.
5. Tweddell JS, Hoffman GM, Mussatto KA, at al. Improved survival of patients
undergoing palliation of hypoplastic left heart syndrome: lessons learned from
115 consecutive patients. Circulation 2002;106:I-82.
6. Newburger JW, Silbert AR, Buckley LP, Fyler DC. Cognitive function and age
at repair of transposition of the great arteries in children. N Engl J Med
1984;310:1495.
7. Bellinger DC, Jonas RA, Rappaport LA, et al. Developmental and neurologic
status of children after heart surgery with hypothermic circulatory arrest or lowflow cardiopulmonary bypass. N Engl J Med 1995;332(9):549.
8. Wernovsky G, Newburger J. Neurologic and developmental morbidity in

children with complex congenital heart disease. J Pediatr 2003;142:6.
9. Pyeritz, RE: Genetics and Cardiovascular Disease. In Braunwald E. Heart

disease: a textbook of cardiovascular medicine, 6th ed., W. B. Saunders
Company, 2001:1977.
10. Clark EB: Mechanisms in the pathogenesis of congenital cardiac

malformations. In: Pierpont MEM, Moller JH, eds. Genetics of cardiovascular
disease. Boston, Martinus Nihjoff, 1986:3.
11. Clark EB. Pathogenetic mechanisms of congenital cardiovascular

malformations revisited [review]. Semin Perinatol 1996;20(6):465.
12. Lacro RV. Dysmorphology. Nadas' pediatric cardiology. Philadelphia: Hanley

& Belfus, 1992:37.
13. Ferencz C, Correa-Villasenor A. Epidemiology of cardiovascular

malformations: the state-of-the-art. Cardiol Young 1991;1:264.
14. Olson EN, Srivastava D. Molecular pathways controlling heart development.

Science 1996;272:671.
15. Srivastava D, Olson EN. A genetic blueprint for cardiac development. Nature
2000;407:221.
16. Online Mendelian inheritance in man, OMIM (TM). Baltimore: Center for
Medical Genetics, Johns Hopkins University; Bethesda: National Center for
Biotechnology Information, National Library of Medicine. World Wide Web URL:
http://www.ncbi.nlm.nih.gov/omim/.
17. Gorlin RJ, Cohen MM, Levin LS. Syndromes of the head and neck, 3rd ed.
Oxford: Oxford University Press, 1990.
18. Jones KL. Smith's recognizable patterns of human malformation. 5th ed.
19. Greenwood RD. Cardiovascular malformations associated with extracardiac

anomalies and malformation syndromes: patterns for diagnosis. Clin Pediatr
1984;23(3):145.
20. Milerad J, Larson O, Hagberg C, Ideberg M. Associated malformations in

infants with cleft lip and palate: a prospective, population-based study.
21. Moore KL, ed. The developing human: clinically oriented embryology, 2nd
ed. Philadelphia: WB Saunders, 1997.
22. Kurnit DM, Layton WM, Matthysse S. Genetics, chance and morphogenesis.
Am J Human Genet 1987;41:979.
23. Shaw GM, O'Malley CD, Wasserman CR, Tolarova MM, Lammer EJ. Maternal
periconceptional use of multivitamins and reduced risk for conotruncal heart
defects and limb deficiencies among offspring. Am J Med Genet 1995;59:536.
24. Botto LD, Khoury MJ, Mulinare J, Erickson JD. Periconceptional multivitamin
use and the occurrence of conotruncal heart defects: results from a populationbased, case-control study. Pediatrics 1996;98:911.
25. Hernandez-Diaz S, Werler MM, Walker AM, Mitchell AA. Folic acid
antagonists during pregnancy and the risk of birth defects. N Eng J Med
2000;343:1608.
26. Junker R, Kotthoff S, Vielhaber H, et al. Infant methylenetetrahydrofolate

reductase 677TT genotype is a risk factor for congenital heart disease.
Cardiovasc Res 2001;51:251.
27. Rosenquist TH, Ratashak SA, Selhub J. Homocysteine induces congenital

defects of the heart and neural tube: effect of folic acid. Proc Natl Acad Sci USA
1996;93:15227.
P.708
28. Kapusta L, Haagmans MLM, Steegers EAP, et al. Congenital heart defects
and maternal derangement of homocysteine metabolism. J Pediatr
1999;135:773.
29. Dawson BV, Johnson PD, Goldberg SJ, Ulreich JB. Cardiac teratogenesis of
trichloroethylene and dichloroethylene in a mammalian model. J Am Coll Cardiol
1990;16:1304.
30. Goldberg SJ, Lebwitz MD, Graver EJ, Hicks S. An association of human
congenital cardiac malformations and drinking water contaminants. J Am Coll
Cardiol 1990;16:155.
31. Croen LA, Shaw GM, Sanbonmatsu L, Selvin S, Buffler PA. Maternal
residential proximity to hazardous waste sites and risk for selected congenital
malformations. Epidemiology 1997; 8:347.
32. Richard P, Charron P, Carrier L, et al. EUROGENE Heart Failure Project.

Hypertrophic cardiomyopathy: distribution of disease genes, spectrum of
mutations, and implications for a molecular diagnosis strategy. Circulation
2003;107:2227.
33. Seidman JG, Seidman C. The Genetic Basis for Cardiomyopathy: from
Mutation Identification to Mechanistic Paradigms. Cell 2001; 104,557.
34. Schwartz ML, Cox GF, Lin AE, et al. Clinical approach to genetic
cardiomyopathy in children. Circulation 1996;94:2021.
35. Splawski I, Timothy KW, Vincent GM, Atkinson DL, Keating MT. Molecular
basis of the long-QT syndrome associated with deafness. N Engl J Med
1997;336:1562.
37. Barcroft J. Researchers of pre-natal life. Oxford: Blackwell & Mott, 1944.
38. Dawes GS. Foetal and neonatal physiology: a comparative study of the
changes at birth. Chicago: Year Book, 1969.
39. Lind J, Stern L, Wegelius C. Human and foetal neonatal circulation.

Springfield, IL: Charles C Thomas, 1964.
40. Rudolph AM. Congenital diseases of the heart. Chicago: Year Book, 1974.
41. Friedman D, Buyon J, Kim M, Glickstein JS. Fetal cardiac function assessed
by Doppler myocardial performance index (Tei index) Ultrasound in Obstetrics
and Gynecology 2003;21:33.
42. Budorick NE, Millman SL New modalities for imaging the fetal heart. Semin
43. Davignon A, Rautaharju P, Boisselle E, et al. Normal ECG standards for

infants and children. Pediatr Cardiol 1979;1:123.
44. Fowler RS, Finlay CD. The electrocardiogram of the neonate. In Freedom
RM, Benson LN, Smallhorn JF, eds. Neonatal heart disease. London: SpringerVerlag, 1992.
45. Schwartz PJ, Garson A Jr, Paul T, et al. Guidelines for the interpetation of
the newborn electrocardiogram. A task force of the European society of
cardiology. Eur Heart J 2002;23:1329.
46. Quinones MA, Douglas PS, Foster E, et al. American College of Cardiology.
American Heart Association. American College of Physicians-American Society of
Internal Medicine. American Society of Echocardiography. Society of
Cardiovascular Anesthesiologists. Society of Pediatric Echocardiography.ACC/
AHA clinical competence statement on echocardiography: a report of the
American College of Cardiology/American Heart Association/American College of
Physicians-American Society of Internal Medicine Task Force on Clinical
Competence. J Am College Cardiol 2003;41:687.
47. Lock JE, Keane JF, Mandell VS, Perry SB. Cardiac catheterization. In Fyler
DC, ed. Nadas' pediatric cardiology. Philadelphia: Hanley & Belfus, 1992.
48. LaFarge CG, Miettinen OS. The estimation of oxygen consumption.

Cardiovasc Res 1970;4:23.
49. Kovalchin JP, Forbes TJ, Nihill MR, Geva T. Echocardiographic determinants
of clinical course in infants with critical and severe pulmonary valve stenosis. J
Am Coll Cardiol 1997;29(5): 1095.
50. Kaine SF, Smith EO, Mott AR, Mullins, Geva T. Quantitative
echocardiographic analysis of the aortic arch predicts outcome of balloon
angioplasty of native coarctation of the aorta. Circulation 1996;94:1056.
51. Rhodes LA, Colan SD, Perry SB, et al. Predictors of survival in neonates with
critical aortic stenosis. Circulation 1991;84:2325.
52. Egito ES, Moore P, O'Sullivan J, et al. Transvascular balloon dilation for
neonatal critical aortic stenosis: early and midterm results. J Am Coll Cardiol
1997;29(2):442.
53. Olley PM, Coceani F, Bodach E. E-type prostaglandins: a new emergency
therapy for certain cyanotic congenital heart malformations. Circulation

1976;53:728.
54. Lang P, Freed MD, Rosenthal A, et al. The use of prostaglandin E1 in an

infant with interruption of the aortic arch. J Pediatr 1977; 91:805.
55. Borow K, Green LH, Castaneda AR, et al. Left ventricular function after
repair of tetralogy of Fallot and its relationship to age at repair. Circulation
1980;61:1150.
56. Jatene AD, Fontes VF, Paulista PP, et al. Anatomic correction of
transposition of the great vessels. J Thorac Cardiovasc Surg 1976; 72:364.
57. Castaneda AR, Norwood WI, Jonas RA, et al. Transposition of the great
arteries and intact ventricular septum: anatomical repair in the neonate. Ann
Thorac Surg 1984;38:438.
58. Wernovsky G, Mayer JE Jr, Jonas RA, et al. Factors influencing early and
late outcome of the arterial switch operation for transposition of the great
arteries. J Thorac Cardiovasc Surg 1995; 109(2):289.
59. Geva T, Ayres NA, Pac FA, Pignatelli R. Quantitative morphometric analysis
of progressive infundibular obstruction in tetralogy of Fallot. A prospective
longitudinal echocardiographic study. Circulation 1995;92(4):886.
60. Walsh EP, Rockenmacher S, Keane JF, et al. Late results in patients with
tetralogy of Fallot repaired during infancy. Circulation 1988; 77:1062.
61. Lillehei CW, Varco RL, Cohen M, et al. The first open heart corrections of
tetralogy of Fallot. A 26-31 year follow-up of 106 patients. Ann Surg 1986;204
(4):490.
62. Giglia TM, Mandell VS, Connor AR, Mayer JE Jr, Lock JE. Diagnosis and
management of right ventricle-dependent coronary circulation in pulmonary
atresia with intact ventricular septum. Circulation 1992;86:1516.
63. Franklin RC, Spiegelhalter DJ, Sullivan ID, et al. Tricuspid atresia presenting
in infancy. Survival and suitability for the Fontan operation. Circulation
1993;87:427.
64. Gentles TL, Mayer JE Jr, Gauvreau K, et al. Fontan operation in five hundred
consecutive patients: factors influencing early and late outcome. J Thorac
Cardiovasc Surg 1997;114:376.
65. Norwood WI, Lang P, Hansen DD. Physiologic repair of aortic atresia
hypoplastic left heart syndrome. N Engl J Med 1984; 308:23.
66. Forbess JM, Cook N, Roth SJ, et al. Ten-year institutional experience with
palliative surgery for hypoplastic left heart syndrome. Risk factors related to
stage I mortality. Circulation 1995;92(9 Suppl):II-262.
67. Iannettoni MD, Bove EL, Mosca RS, et al. Improving results with first-stage
palliation for hypoplastic left heart syndrome. J Thorac Cardiovasc Surg
1994;107:934.
68. Razzouk AJ, Chinnock RE, Gundry SR, et al. Transplantation as a primary
treatment for hypoplastic left heart syndrome: intermediate-term results. Ann
69. Gutgesell HP, Massaro TA. Management of hypoplastic left heart syndrome
in a consortium of university hospitals. Am J Cardiol 1995;76:809.
70. Canter C, Naftel D, Caldwell R, et al. Survival and risk factors for death after
cardiac transplantation in infants. A multiinstitutional study. The Pediatric Heart
Transplant Study. Circulation 1997;96:227.
71. Hanley FL, Heinemann MK, Jonas RA, et al. Repair of truncus arteriosus in
the neonate. J Thorac Cardiovasc Surg 1993;105: 1047.
72. Rajasinghe HA, McElhinney DB, Reddy VM, Mora BN, Hanley FL. Long-term
follow-up of truncus arteriosus repaired in infancy: a twenty-year experience. J
Thorac Cardiovasc Surg 1997; 113(5): 869-878.
73. Piccoli G, Pacifico AD, Kirklin JW, Blackstone EH, Kirklin JK, Bargeron LM.
Changing results and concepts in the surgical treatment of double-outlet right
ventricle: analysis of 137 operations in 126 patients. Am J Cardiol 1983;52:549.
74. Radzik D, Davignon A, van Doesburg N, et al. Predictive factors for

spontaneous closure of atrial septal defects diagnosed in the first 3 months of
life [Comment]. J Am Coll Cardiol 1994;23:828.
75. Bharati S, Lev M. The spectrum of common atrioventricular orifice (canal).

Am Heart J 1973;86:553.
76. Hanley FL, Fenton KN, Jonas RA, et al. Surgical repair of complete
atrioventricular canal defects in infancy. Twenty-year trends. J Thorac
P.709
77. Slomp J, Gittenberger-de Groot AC, Glukhova MA, et al. Differentiation,
dedifferentiation, and apoptosis of smooth muscle cells during the development
of the human ductus arteriosus. Arterioscler Thromb Vasc Biol 1997;17:1003.
78. Friedman WF, Hirschklau MJ, Previtz MP, et al. Pharmacologic closure of
patent ductus arteriosus in the premature infant. N Engl J Med 1976;295:526.
79. Heymann MA, Rudolph AM, Silverman NH. Closure of the ductus arteriosus
in premature infants by inhibition of prostaglandin synthesis. N Engl J Med
1976;295:530.
80. Nehgme RA, O'Connor TZ, Lister G, Bracken MB. Patent ductus arteriosus.
In Sinclair JC, Bracken MB, eds. Effective care of the newborn infant. New York:
Oxford University Press, 1992;281.
81. Ment LR, Oh W, Ehrenkranz RA, et al. Low-dose indomethacin and

prevention of intraventricular hemorrhage: a multicenter randomized trial.
Pediatrics 1994;93(4):543.
82. Fowlie PW. Prophylactic indomethacin: systematic review and metaanalysis. Arch Dis Child 1996;74:F81.
83. Volpe JJ. Brain injury caused by intraventricular hemorrhage: is

indomethacin the silver bullet for prevention? Pediatrics 1994; 11:673.
84. Reynolds EOR. Prevention of periventricular hemorrhage. Pediatrics

1994;11:677.
85. Varvarigou A, Bardin CL, Beharry K, Chemtob S, Papageorgiou A, Aranda

JV. Early ibuprofen administration to prevent patent ductus arteriosus in
premature newborn infants. JAMA 1996; 275:539.
86. Wagner HR, Ellison RC, Zierler S, et al. Surgical closure of patent ductus
ateriosus in 268 preterm infants. J Thorac Cardiovasc Surg 1984;87:870.
87. Gersony WM, Peckham GJ, Ellison RC, et al. Effects of indomethacin in
preterm infants with patent ductus arteriosus. Results of a national collaborative
study. J Pediatr 1983;102:895.
88. Doty DB, Richardson JV, Falkovsky GE, Gordonova MI, Burakovsky VI.
Aortopulmonary septal defect: hemodynamics, angiography, and operation. Ann
89. Edwards JE. Malformations of the aortic arch system manifested as

vascular rings. Lab Invest 1953;2:56.
90. Yeager SB, Chin AJ, Sanders SP. Two-dimensional echocardiographic

diagnosis of pulmonary artery sling in infancy. J Am Coll Cardiol 1986;7:625.
91. Tonkin IL, Elliot LP, Bargeron LM. Concomitant axial cineangiography and
barium esophagography in the evaluation of vascular rings. Radiology
1980;135:69.
92. Backer CL, Ilbawi MN, Idriss FS, DeLeon SY. Vascular anomalies causing
tracheoesophageal compression. J Thorac Cardiovasc Surg 1989;97:725.
93. Swischuk LE. Anterior tracheal indentation in infancy and early childhood:
normal or abnormal? Am J Roentgenol Rad Ther Nucl Med 1971;112:12.
94. Nir A, Tajik AJ, Freeman WK, et al. Tuberous sclerosis and cardiac
rhabdomyoma. Am J Cardiol 1995;76:419.
95. De Giovanni JV, Dindar A, Griffith MJ, et al. Recovery pattern of left
ventricular dysfunction following radiofrequency ablation of incessant
supraventricular tachycardia in infants and children. Heart 1998;79:588.
96. Gorgels AP, Al Fadley F, Zaman L, et al. The long QT syndrome with
impaired atrioventricular conduction: a malignant variant in infants. J Cardio
Electrophys 1998;9:1225.
97. Nagashima M, Matsushima M, Ogawa A, et al. Cardiac arrhythmias in

healthy children revealed by 24-hour ambulatory ECG monitoring. Pediatr
Cardiol 1987;8:103.
98. Southall DP, Richards J, Mitchell P, et al. Study of cardiac rhythm in healthy
newborn infants. Brit Heart J 1980;43:14.
99. Van Hare GF, Stanger P. Ventricular tachycardia and accelerated ventricular
rhythm presenting in the first month of life. Am J Cardiol 1991;67:42.
100. Snyder CS, Fenrich AL, Friedman RA, et al. Usefulness of echocardiography
in infants with supraventricular tachycardia. Am J Cardiol 2003;91:1277.
101. Weindling SN, Walsh EP, Saul JP. Efficacy and risks of medical therapy for
supraventricular tachycardia in neonates and infants. Am Heart J 1996;131:66.
102. Bromberg BI, Lindsay BD, Cain ME, et al. Impact of clinical history and
electrophysiologic characterization of accessory pathways on management
strategies to reduce sudden death among children with Wolff-Parkinson-White
syndrome. J Am Coll Cardiol 1996;27:690.
103. Gollob MH, Green MS, Tang AS, et al. Identification of a gene responsible
for familial Wolff-Parkinson-White syndrome. N Eng J Med 2001;344:1823.
104. Bauersfeld U, Gow RM, Hamilton RM, et al. Treatment of atrial ectopic
tachycardia in infants <6 months old. Am Heart J 1995; 129:1145.
105. Villain E, Vetter V, Garcia JM, et al. Evolving concepts in the management
of congenital junctional ectopic tachycardia. Circulation 1990;81:1544.
106. Casey FA, McCrindle BW, Hamilton RM, et al. Neonatal atrial flutter:
significant early morbidity and excellent long-term prognosis. Am Heart J
1997;133:302.
107. Epstein ML, Kiel EA, Victorica BE. Cardiac decompensation following
verapamil therapy in infants with supraventricular tachycardia. Pediatrics
1985;75:737.
108. Perry JC, Garson A, Jr. Supraventricular tachycardia due to WolffParkinson-White syndrome in children: early disappearance and late recurrence.
J Am Coll Cardiol 1990;16:1215.
109. Friedman RA, Walsh EP, Silka MJ, et al. NASPE Expert Consensus
Conference: Radiofrequency catheter ablation in children with and without
congenital heart disease. Pacing Clin Electrophysiol 2002;25:1000.
110. Perry JC. Ventricular tachycardia in neonates. Pacing Clin Electrophysiol

1997;20:2061.
111. Antzelevitch A. Molecular Genetics of Arrhythmias and Cardiovascular

Conditions Associated with Arrhythmias. Pacing Clin Electrophysiol
2003;26:2194.
112. Schwartz PJ, Moss AJ, Vincent GM, et al. Diagnostic criteria for the long QT
syndrome. An update. Circulation 1993;88:782.
113. Hoorntje T, Sreeram N, de Vroet R. Device therapy for malignant neonatal

long QT syndrome. Internat J Cardiol 1999;71:289.
114. Reed BR, Lee LA, Harmon C, et al. Autoantibodies to SS-A/Ro in infants
with congenital heart block. J Pediatr 1983;103:889.
115. Buyon JP, Hiebert R, Copel J, et al. Autoimmune-associated congenital

heart block: demographics, mortality, morbidity and recurrence rates obtained
from a national neonatal lupus registry. J Am Coll Cardiol 1998;31:1658.
116. Gregoratos G, Abrams J, Epstein AE, et al. ACC/AHA/NASPE 2002

Guideline Update for Implantation of Cardiac Pacemakers and Antiarrhythmia
Devicessummary article: a report of the American College of Cardiology/

American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol
2002.
117. Mullins CE, Mayer DC. Congenital heart disease: a diagramatic atlas. New
York: Alan R Liss, 1988.
Chapter 34
Preoperative and Postoperative Care of the Infant with
John M. Costello
Wayne H. Franklin
INTRODUCTION
The contemporary strategy of early, complete repair for congenital heart defects originated in the 1970s, as
evidenced by the first report in 1976 of complete anatomic correction of d-transposition of the great arteries (dTGA) with ventricular septal defect (VSD) in early infancy (1). Aided by advances in surgical technique and
equipment, myocardial protection, cardiac anesthesia and perioperative care, the strategy of early primary
repair is now applied to many other complex congenital heart lesions, such as tetralogy of Fallot and truncus
arteriosus (2,3).Furthermore, infants with certain lesions that were considered before the 1980s to be
inoperable, such as hypoplastic left heart syndrome (HLHS), currently undergo staged surgical palliation with
constantly improving outcomes (4). The first stage of surgical palliation for infants with single ventricle
physiology usually must be performed in the neonatal period to minimize early morbidity and mortality, and to
optimize the cardiovascular physiology for the subsequent Fontan operation. Early repair or palliation (for single
ventricle lesions) benefits infants because prolonged preoperative periods of cyanosis (See Color Plate) and
heart failure, and the many associated complications are minimized (5,6,7). These potential associated
complications include impairment of cognitive function as a result of chronic hypoxemia, central nervous system
embolic events, and the development of pulmonary vascular obstructive disease. Postoperative complications
including pulmonary hypertensive crises are also reduced using a strategy of early surgical intervention (8). The
incidence of vascular complications and infection related to central lines in infants requiring prolonged
hospitalization prior to cardiac repair or palliation will also be minimized.
As a result of this early intervention strategy, approximately one-third of all patients with congenital heart
defects undergo primary repair or palliation in early infancy. Due to the immaturity of many organ systems,
these neonates possess limited physiologic reserve. They have limited ability to increase stroke volume, limited
functional residual capacity in the lung, limited fat and carbohydrate reserves, and a limited ability to regulate
temperature. Drug metabolism is altered by hepatic and renal immaturity, and total body water content. All of
these factors complicate the care of infants in the perioperative period, and some have prolonged intensive care
unit (ICU) stays, presenting a challenge for physicians involved with their management.
A variety of pediatric personnel with expertise in pediatric cardiology are involved in the day-to-day
management of these patients. These personnel include neonatologists, intensivists, cardiologists, cardiac
surgeons, anesthesiologists, nurses, and respiratory therapists. Because of the complexity inherent to this field,
and to ensure an optimal outcome for each patient, seamless communication of pertinent anatomic,
pathophysiologic, and operational/interventional details is required across all specialties.
This chapter will complement Chapter 33 by providing an overview of the key issues, concepts, and strategies
pertaining to the care of infants with critical congenital heart
P.711
disease. We have allowed some overlap of information between chapters in the interest of clarity. The expected
presentation and clinical course of specific congenital heart lesions will be discussed, and the general principles
that are widely applied to patient management will be reviewed. This chapter will not provide a comprehensive
description of every congenital heart defect and associated variants that may present in infancy. The reader
must appropriately apply the concepts and principles discussed in the following pages to each unique patient.
The first section will cover general preoperative issues, and the presentation, physiology and management of
common types of critical congenital heart disease. The second section will provide an overview of
cardiopulmonary bypass (CPB) and its sequelae, followed by a lesion specific discussion of operative
intervention, postoperative physiology and common complications.
PREOPERATIVE CARE
Fetal Diagnosis
Fetal echocardiography plays an important role in the management of critical congenital heart disease for a
variety of reasons (9). The identification of severe congenital heart disease during the second trimester allows
time for family counseling. Fetuses that are likely to require surgical or transcatheter intervention in the
neonatal period should ideally be delivered at an institution with a level II or III nursery that is located at or in
close proximity to a tertiary care pediatric cardiology center. In the absence of nonimmune hydrops fetalis,
fetuses with known critical congenital heart disease should be delivered at term to minimize the potential
complications of prematurity, such as respiratory distress syndrome, intraventricular hemorrhage and
necrotizing enterocolitis (NEC). As discussed below, infants with unrecognized critical congenital heart disease
may develop severe cyanosis (See Color Plate), shock, myocardial dysfunction and end organ damage.
Information from the fetal echocardiogram thus guides planning for initial stabilization upon delivery, such as
the need for a prostaglandin E1 (PGE1) infusion. Prenatal diagnosis is associated with reduced preoperative
morbidity and mortality in large series of neonates with HLHS and TGA (10,11,12,13). Management strategies
may be developed in advance for the uncommon patient with congenital heart disease who may become
critically ill in the first few minutes of life, such as some neonates with absent pulmonary valve syndrome, or
HLHS with an intact atrial septum. The impact of in utero diagnosis of critical congenital heart disease on
postoperative mortality is controversial. Some centers have reported lower postoperative mortality in patients
with a fetal diagnosis of HLHS or d-TGA, whereas others found no such impact (12,13,14,15,16).
TABLE 34-1 RISK FACTORS FOR CONGENITAL HEART DISEASE THAT WARRANT FETAL
ECHOCARDIOGRAPHY (SEE ALSO TABLE 33-6)
Maternal Risk Factors
1. Family history of congenital heart disease
2. Maternal disease (lupus, diabetes mellitus)
3. Environmental (alcohol, certain viruses,
medications)
Fetal Risk Factors

1.
2.
3.
4.
Extracardiac anomalies
Chromosomal anomalies
Arrhythmia
Abnormal fetal growth
5. Fetal distress
6. Obstetrician's suspicion of CHD from screening
US
CHD, congenital heart disease; US, ultrasound.
Known risk factors for congenital heart disease that warrant referral to a pediatric cardiologist for fetal
echocardiography are listed in Table 34-1 (17). Even if these screening guidelines are followed, the majority of
patients with congenital heart disease will not be detected until after birth. Fetal echocardiography has several
known limitations, most notably sub-optimal windows in mothers with large body habitus, and difficulty with
precise imaging of extracardiac vessels. Two serious congenital heart lesions that may not be always appreciated
with fetal echocardiography deserve comment. Isolated coarctation of the aorta may be difficult to definitively
diagnose or exclude in utero as a result of the obligatory presence of the ductus arteriosus (18). Total
anomalous pulmonary venous return may also be unrecognized. Thus, any diagnosis of fetal congenital heart
disease must be confirmed, and the anatomic details clarified, as soon as possible after birth with a
transthoracic echocardiogram.
Routine screening fetal ultrasound studies, which are interpreted by obstetricians, are usually limited to the
demonstration of a heart with four chambers. Doppler assessment of blood flow patterns and valvar function is
not usually performed, and data regarding longitudinal growth of the ventricular chamber sizes are not available
when a single study is performed in the second trimester. Thus, a history of a normal fetal ultrasound does not
exclude the possibility of critical congenital heart disease.
P.712
Transitional Circulation
An appreciation of the fetal-placental circulation, and the normal transition from fetal to newborn circulation, is
required to understand the timing and presentation of symptoms in the neonate with critical heart disease (19).
In the placenta, gas, nutrient, and waste exchange occur between the fetal and maternal circulations.
Oxygenated blood then returns from the placenta to the fetus through the umbilical vein and partially bypasses
the liver, via the ductus venosus, into the inferior vena cava. The oxygenated blood is preferentially shunted
through the foramen ovale to the left atrium (20). The oxygen-rich blood thenenters the left ventricle, from
where it is pumped out the aorta tosupply the coronary circulation and the brain. Deoxygenated blood from the
superior vena cava and distal inferior vena cava preferentially enters the right ventricle, and is then pumped to
the pulmonary artery. This preferential blood flow phenomenon is known as streaming. Due to the high
pulmonary vascular resistance in utero, the majority of the deoxygenated pulmonary artery blood bypasses the
lungs, and flows through the ductus arteriosus into the descending aorta, supplying the lower body and placenta.
An elegant series of events occur when the umbilical cord is clamped immediately after birth. The ductus
venosus functionally closes in large part as a result of decreased flow from the placenta (21). The low resistance
placental circulation is gone and thus systemic vascular resistance rises, although pulmonary vascular resistance
falls as a result of mechanical expansion of the lungs with respiration and higher oxygen tension (22,23). Blood
ejected from the right ventricle now perfuses the lungs rather than entering the ductus arteriosus, and the
increased blood returning to the left atrium leads to functional closure of the foramen ovale. The increased
oxygen tension also contributes to functional closure of the ductus arteriosus.
When the aforementioned events, in particular the fall in pulmonary vascular resistance and closure of the
ductus arteriosus, occur in the setting of critical congenital heart disease, a physiologic state of either
inadequate systemic or pulmonary blood flow exists which, if unrecognized, may progress to multi-organ system
failure and death.
Presentation of Critical Congenital Heart Disease

Despite the broad spectrum of congenital heart lesions, neonates with critical congenital heart disease
undetected in utero present in a limited number of ways: with cyanosis (inadequate pulmonary blood flow),
shock (inadequate systemic blood flow) or some combination of these physiologic states. Older neonates and
infants with significant left to right shunting at the ventricular or great artery level may present with congestive
heart failure. Critical congenital heart disease may be detected prior to the onset of symptoms as a result of
abnormal findings on cardiovascular examination or during echocardiographic screening performed in neonates
with known chromosomal abnormalities or noncardiac congenital malformations.
Critical coarctation of the aorta, critical aortic valve stenosis, and HLHS are frequently encountered lesions that
present in the neonatal period with symptomatic obstruction of the left ventricular outflow tract. In each case,
constriction or closure of the ductus arteriosus leads to inadequate systemic blood flow. These patients will
typically present within the first week or two of life with feeding difficulties, tachypnea, poor perfusion and
metabolic acidosis, a constellation of findings that may be mistaken for septic shock. Due to low systemic
cardiac output, findings such as pathologic heart murmurs may not be present and accurate four-extremity
blood pressure measurements may be difficult to obtain. Diminished or absent pulses in the lower extremities
and an increased right ventricular impulse are more reliable findings on physical examination, and a high index
of suspicion is required to make a timely diagnosis.
Critical pulmonary valve stenosis, severe tetralogy of Fallot, and pulmonary atresia are examples of lesions with
right ventricular outflow tract obstruction that manifest with progressive cyanosis in the newborn period as
constriction of the ductus arteriosus limits pulmonary blood flow.
Lesions with complete intracardiac mixing of the systemic venous and pulmonary venous circulations, such as
tricuspid atresia, truncus arteriosus, or complex single ventricles usually have variable degrees of obstruction to
either the systemic or pulmonary circulations, depending upon the anatomic details of the individual lesion. Less
commonly, they have no obstruction to either systemic or pulmonary blood flow, leading to mild hypoxemia
(less commonly clinical cyanosis) but substantial heart failure that presents later in the neonatal period or during
early infancy.
d-TGA is a common cyanotic lesion with the unique physiology of systemic and pulmonary circulations that run
in parallel. Infants with d-TGA, an intact ventricular septum, and no significant outflow tract obstruction usually
present with cyanosis soon after birth. Those with a significant VSD have less desaturation and can present later
with congestive heart failure.
The neonate with nonimmune hydrops fetalis may have an underlying cardiac problem. The differential diagnosis
includes serious congenital heart disease (particularly with significant atrioventricular valve regurgitation),
obstructive cardiac tumors, congenital myocarditis, tachyarrhythmias and bradycardia.
Initial Evaluation and Stabilization of the Neonate with Suspected

A careful history and cardiovascular examination will often reveal abnormal findings in the infant with critical
congenital heart disease, even before the onset of symptoms. Review of maternal records may identify a known
risk factor for congenital heart disease (Table 34-1) (17). The course of labor
P.713
should be reviewed, and factors that suggest acute infection, respiratory distress syndrome, or meconium
aspiration should be noted, as they may indicate a primary respiratory cause for hypoxemia. The physical
examination focuses on the detection of dysmorphic features, cyanosis, an abnormal precordium, abnormal
femoral and peripheral pulses, and/or a pathologic heart murmur. Measurement of blood pressure in all four
extremities is useful if an arm-leg gradient of greater than 10 mm Hg is detected, but in the setting of a large
patent ductus arteriosus, the pressures may be equal despite the presence of a critical aortic coarctation or
interrupted aortic arch. Although respiratory distress may be present in neonates with critical congenital heart
disease or pulmonary problems, those with heart disease typically have a shallow rapid respiratory pattern
(quiet tachypnea), as opposed to the more labored breathing pattern of neonates with a primary pulmonary
process.
The term differential cyanosis refers to a condition in which the lower body is visibly more cyanotic than the
upper body as a result of right to left shunting at the ductal level. The presence of differential cyanosis,
occasionally seen with severe pulmonary hypertension, may also be indicative of aortic arch obstruction or
coarctation of the aorta. In neonates with d-TGA, reverse differential cyanosis may be present if aortic arch
obstruction or severe pulmonary hypertension also exists. In such patients, oxygenated blood is ejected from
the left ventricle to the pulmonary artery, and then passes through the patent ductus arteriosus to the
descending aorta.
In newborns that present with cyanosis, a right radial arterial blood gas analysis on room air and 100% inspired
oxygen (the hyperoxia test) is useful for discriminating those with cyanotic heart disease from those with
pulmonary disease. The partial pressure of carbon dioxide (pCO2) is typically mildly decreased in newborns with
cardiac disease and mildly elevated in those with pulmonary disease. The arterial oxygen pressure (PaO2) is
often between 25 to 40 mm Hg on room air in both patient groups. In 100% FiO2, the PaO2 will usually rise to
more than 100 mm Hg in patients with pulmonary disease, provided that significant pulmonary artery
hypertension is not present. The PaO2 will remain unchanged or only increase slightly in most neonates with
cyanotic heart disease. There are limitations to the hyperoxia test. Patients with left sided obstructive lesions (e.
g., HLHS, critical coarctation of the aorta, critical aortic valve stenosis, interrupted aortic arch) may have a high
PaO2 (>60 mm Hg) in any extremity or a very high PaO2 (>150 mm Hg) in the right arm with certain anatomic
variants. Pulse oximetry should not be used as a substitute for arterial blood gas analysis when performing the
hyperoxia test. For example, a neonate receiving 100% fraction of inspired oxygen (FiO2) may have a pulse
oximetry reading of 100%, with a PaO2 of 80 mm Hg (suggestive of cyanotic heart disease), or a PaO2 more
than 100 mm Hg (suggestive of pulmonary disease). Potential complications exist when performing the
hyperoxia test. For example, if the test is continued for a prolonged period of time in neonates with left sided
obstructive lesions, pulmonary vascular resistance may drop further, and the ductus arteriosus may begin to
close, leading to clinical deterioration.
Interpretation of a chest radiograph (CXR) and 12-lead electrocardiogram (ECG) may provide insight to the
underlying cardiac diagnosis in cyanotic newborns. For example, cyanotic newborns with normal heart size,
normal to increased pulmonary arterial markings on CXR, and a normal ECG may have TGA. Cyanotic newborns
with normal heart size and decreased pulmonary blood flow on CXR are stratified by the ECG findings: those
with right axis deviation (90-180) on ECG may have tetralogy of Fallot; those with a normal adult axis (1090) on ECG may have pulmonary atresia with an intact ventricular septum; and those with an ECG with left
axis deviation (0 to -90) and left ventricular hypertrophy may have tricuspid atresia. Infants with a normal to
small heart size and increased pulmonary vascular markings on CXR may have total anomalous pulmonary
venous return, and those with a large heart with increased pulmonary vascular markings may have truncus
arteriosus, or a complex single ventricle lesion with unobstructed pulmonary blood flow. Neonates with a very
large heart and decreased pulmonary arterial markings may have Ebstein's anomaly of the tricuspid valve.
If the aforementioned evaluation is consistent with the presence of critical congenital heart disease, it is
reasonable to stabilize the infant, including the initiation of a PGE1 infusion in some cases, and then arrange
transport to a pediatric cardiology center in which the initial echocardiogram can be performed. A PGE1 infusion
should be initiated in neonates who present with shock, cyanosis with a minimal increase in PaO2 during the
hyperoxia test, or a fetal diagnosis of ductal dependent congenital heart disease. An echocardiogram may be
performed at the presenting institution if the presence of critical congenital heart disease is in doubt, or to guide
initial stabilization if the transport time will be prolonged. The echocardiographer and the sonographer must
have experience with congenital heart lesions, otherwise false negative and false positive evaluations may occur
(24).
The introduction of prostaglandin infusions to maintain ductal patency in the late 1970s represented a major
advance for infants with critical congenital heart disease (25,26). While receiving PGE1, infants can be safely
transported over long distances to congenital heart centers. Cardiac surgery can be postponed while infections
are treated and other major congenital anomalies identified. Patients who present in shock can be medically
managed for several days, allowing time for recovery of end-organ function prior to surgery. Detailed cardiac
diagnostic information can be obtained using echocardiography, cardiac catheterization, spiral computed
tomography (CT), magnetic resonance imaging (MRI), or electron beam tomography (EBT) to allow optimal
planning for intervention.
PGE1 can be administered through a peripheral intravenous line, although many practitioners prefer to use
central venous access to ensure uninterrupted medication delivery. A wide range of dosing of PGE1 is used
depending on the clinical state of the patient. Generally a PGE1 infusion
P.714
of 0.1 mcg/kg/min is used when the ductus arteriosus is severely constricted or functionally closed and a state
of shock or severe cyanosis exists. Lower doses (0.01- 0.05 mcg/kg/min) will maintain ductal patency in
neonates with presumed ductal dependent systemic or pulmonary circulation who are otherwise stable (27).
The most troublesome side effect of PGE1 infusion is apnea, which occurs in approximately one-third of neonates
(27,28). Tracheal intubation and mechanical ventilation are usually indicated when PGE1 is started for neonates
who present with severe cyanosis or shock and/or require transfer to another institution. Lower doses of PGE1
are associated with fewer side effects, particularly apnea (27,29). Patients who are stable upon presentation and
are started on a low dose of PGE1 may be less likely to develop apnea, and their respiratory status may be
observed without mechanical ventilation. This strategy may be used in patients who require transfer to another
institution, provided that members of the transport team are experienced and prepared to intubate the patient's
trachea should the need arise. Preliminary data suggest that aminophylline may be prescribed to minimize the
occurrence of apnea in neonates requiring PGE1 infusion (30). PGE1 is a potent vasodilator, and hypotension
may occur following the initiation of the drug, particularly if narcotics are also administered to facilitate
procedures. This problem should be anticipated, and the hypotension usually resolves with a dose reduction of
PGE1 and volume administration. Other common side effects are listed in Table 34-2 (27,28). Although unusual,
clinical deterioration may occur following the initiation of a PGE1 infusion, and in these instances, obstruction to
pulmonary venous return (e.g., obstructed total anomalous pulmonary venous return) or left atrial egress (e.g.,
TGA with intact ventricular septum and a restrictive atrial communication) may be present. These infants require
emergent surgical or catheter intervention. Some neonates have congenital absence of the ductus arteriosus
(such as those with tetralogy of Fallot with absent pulmonary valve syndrome) or some infants with pulmonary
atresia, ventricular septal defect, and major aortopulmonary collateral arteries. A PGE1 infusion in these patients
may worsen cyanosis by lowering systemic vascular resistance and thereby decreasing pulmonary blood flow.
When endotracheal intubation is indicated for a neonate with critical congenital heart disease, the use of
medications should be considered to blunt the stress and vagal responses to laryngoscopy, decrease oxygen
consumption and thus improve oxygen delivery, and provide pharmacologic paralysis to facilitate the procedure.
The choice of specific medications depends on the clinical scenario and the airway skills of the clinician. As
mechanical ventilation is initiated, care must be taken in many instances to avoid excessive use of supplemental
oxygen and hyperventilation. These maneuvers may lower pulmonary vascular resistance, thereby increasing
the ratio of pulmonary to systemic blood flow (Qp/Qs), and stealing blood flow from the systemic circulation.
This is discussed in greater detail in the HLHS section below.
TABLE 34-2 SIDE EFFECTS OF PROSTAGLANDIN E1 INFUSION
Respiratory
Cardiovascular
Endocrine/metabolic
Gastrointestinal
Hematologic
respiratory depression, apnea

hypotension, tachycardia
fever, seizures
hypocalcemia, hypoglycemia, cortical hyperostosis*
diarrhea, gastric outlet obstruction*
inhibition of platelet aggregation
Dermatologic
flushing, harlequin rash
* Seen with long-term use.
Stable intravenous access is required for all infants with critical congenital heart disease. The need for arterial
and central venous access should be determined on a case-by case basis, based upon the heart defect, clinical
presentation and anticipated preoperative course. Central access may be needed to allow for medication
delivery, frequent blood sampling, and the administration of parenteral nutrition. The most accessible vessels in
the newborn are the umbilical arteries and veins. The use of umbilical vessels for preoperative stabilization
leaves other vessels available for future cardiac catheterizations and surgical procedures. If umbilical lines
cannot be placed, the radial or femoral arteries and the femoral veins may be accessed. Some clinicians prefer
to avoid placing central venous catheters in the internal jugular or subclavian veins in patients with single
ventricle physiology to prevent scarring and clot formation, which may impede venous return following the bidirectional Glenn/hemi-Fontan and Fontan operations.
Adequate cardiac output and oxygen delivery usually exist in infants with critical congenital heart disease
following initial stabilization and infusion of PGE1. Infants who demonstrate signs of inadequate cardiac output
may benefit from inotropic infusions, but should be critically reassessed to ensure that the PGE1 is being
delivered correctly, that appropriate oxygenation and ventilation strategies are being employed to prevent
pulmonary over-circulation, and that noncardiac etiologies for shock are not present, such as pneumothorax or
sepsis. Fetal-maternal hemorrhage, twin-twin transfusion, and blood loss at delivery from delayed cord clamping
are examples of perinatal causes of anemia that may necessitate transfusion of packed red blood cells soon after
delivery. Most patients with ductal dependent congenital heart disease require a hemoglobin level of 13 to
P.715
15 gm/dL. Attention also needs to be given to acid-base status, and blood glucose and calcium levels.
Bacterial pneumonia and sepsis are often considered when an infant presents with cyanosis and shock, and
empiric antibiotics may be prescribed before a diagnosis of critical congenital heart disease is confirmed. If no
source of bacterial infection is identified, the antibiotics may be discontinued after a 48-hour period. Although it
has been suggested that the use of PGE1 increases the risk of bacterial infection, there is no published evidence
to support this concept.
Neonates with critical congenital heart disease require transport to congenital heart centers for further
evaluation and intervention. Experienced pediatric transport teams are used when available, and close
communication is required with the intensivist and pediatric cardiologist at the congenital heart center
throughout the transport process. During transport, care must be taken to maintain a normal patient
temperature as hypothermia will increase systemic vascular resistance and may be poorly tolerated. The
transport team members must understand that normal ventilation and the avoidance of excessive supplemental
oxygen are important to maintain a balanced circulation in many infants with ductal dependent systemic or
pulmonary blood flow.
Evaluation at the Congenital Heart Center

Upon arrival to the tertiary care center, the intensivist and nursing staff should obtain a detailed report from the
transport team. Historical features that are useful to direct the diagnostic evaluation and initial management
include the results of fetal echocardiogram(s), complications of the pregnancy, family history, estimated
gestational age of the fetus, severity of cyanosis or shock prior to initial resuscitation, presence of risk factors
for infection, and suspected noncardiac congenital anomalies.
A detailed physical examination includes a review of recent and current vital signs. Heart rates are usually within
the normal range for age. Four-extremity blood measurements should be obtained. Pulse oximetry results are
interpreted in the context of the patient's physiology. The presence of dysmorphic facial features may suggest a
specific genetic syndrome. The cardiac examination begins with inspection and palpation of the precordium. An
increased right ventricular impulse is often present with significant left ventricular outflow tract obstruction and/
or pulmonary hypertension. Auscultation is performed with attention to splitting and quality of the second heart
sound, systolic ejection clicks, and the presence of pathologic systolic or diastolic murmurs. The span of the liver
is determined, and the quality of the peripheral pulses and perfusion are noted.
The ECG and CXR are reviewed with attention to features that may suggest an underlying cardiac diagnosis as
described above. The CXR should also be reviewed to ensure that any tubes or lines placed prior to transport
have not migrated. Cardiopulmonary instability in the infant with critical congenital heart disease is often as a
result of noncardiac issues, such as respiratory distress syndrome, tension pneumothorax, or endotracheal tube
plugging or malposition.
Even if an echocardiogram has been performed at the referring institution, a complete echocardiogram is
obtained upon arrival to the congenital heart center to confirm the diagnosis and clarify anatomic and
physiologic details. Complete diagnostic information is obtained in the vast majority of patients with critical
congenital heart disease using transthoracic echocardiogram, eliminating the need for diagnostic cardiac
catheterization (31,32). Occasionally a critically ill infant with congenital heart disease will not tolerate a
prolonged echocardiogram, as a result of temperature instability related to exposure, or hypotension related to
abdominal compression during subcostal imaging.
Limited indications exist for diagnostic cardiac catheterization in the current era. For example, infants with
pulmonary atresia, VSD and major aortopulmonary collateral arteries (MAPCAs) usually require catheterization
to define all sources of pulmonary arterial blood flow. Coronary anatomy can usually be clarified by
echocardiography, but occasionally angiography is required, as is the case in some infants with tetralogy of
Fallot, pulmonary atresia with intact ventricular septum, or TGA (33). Hemodynamic data, such as pulmonary
vascular resistance, may need to be obtained by cardiac catheterization in selected older infants prior to a
definitive procedure. Cardiac catheterization is more commonly indicated to perform interventional procedures.
For example, neonates with TGA and intact ventricular septum may require enlargement of the atrial
communication, and those with HLHS and intact atrial septum require emergent creation of an adequate atrial
septal defect (ASD) to allow egress of pulmonary venous blood from the left atrium into the right atrium
(34,35). Balloon valvuloplasty of critical aortic or pulmonary valve stenosis is commonly performed in many
centers. Cardiac MRI, spiral CT, and electron beam tomography are evolving tools that are currently available in
selected centers and are most useful for clarifying anatomy of extracardiac vessels (36,37).
Evaluation of other organ systems is indicated in infants with critical congenital heart disease prior to
intervention. Basic laboratory studies are obtained to evaluate acid-base status, oxygenation and ventilation,
and the hepatic, renal and hematologic systems. Recovery of end-organ function in infants who present in shock
is to be expected in nearly all cases following appropriate resuscitation and stabilization. Infants who are
premature or who present with severe cyanosis or shock should have a head ultrasound performed to evaluate
for intracranial hemorrhage. If not clear by echocardiography, an abdominal ultrasound is indicated in patients
with heart disease associated with heterotaxy syndrome to determine the sidedness of the liver and spleen. An
upper gastrointestinal series may be needed in such infants to evaluate for malrotation. Any suspected
hematologic issues, such as thrombocytopenia or a bleeding diathesis, should be thoroughly evaluated prior to
intervention. Other major birth defects may be seen in up
P.716
to 25% of patients with significant congenital heart disease. For example, infants with conotruncal heart defects,
such as tetralogy of Fallot, pulmonary atresia with VSD, interrupted aortic arch, and truncus arteriosus have an
increased risk compared with the general population of having an associated oral cleft, omphalocele,
trachoesophageal fistula, or imperforate anus (38). There is an increased incidence of renal anomalies in infants
with congenital heart disease who have other major congenital anomalies, and a screening renal ultrasound may
be indicated in this subset of patients (39).
A chromosomal abnormality or easily definable genetic syndrome may be identified in approximately 20% of
infants with congenital heart disease. Table 34-3 lists selected common chromosomal abnormalities and genetic
syndromes, and the congenital heart lesions with which they are associated (40,41,42).
Preoperative Supportive Care: General Principles

Cardiopulmonary Management
Infants with ductal dependent congenital heart disease require a balanced circulation to achieve adequate
systemic and pulmonary blood flow. Those who present with severe cyanosis or shock, or require prolonged
interhospital transport are tracheally intubated and mechanical ventilation is provided. Great care must be taken
throughout the preoperative period to ensure that appropriate ventilation and oxygenation strategies are used to
prevent an iatrogenic imbalance between systemic and pulmonary blood flow. In particular, in patients with
ductal dependent systemic blood flow (e.g., HLHS or interrupted aortic arch), hyperventilation and provision of
supplemental oxygen will decrease pulmonary vascular resistance, thereby promoting increased pulmonary to
systemic flow ratio (Qp/Qs), and steal from the systemic and coronary circulations. Such patients will have
inappropriately high systemic oxygen saturations (greater than 90%), poor systemic perfusion and are at risk of
developing myocardial dysfunction as a result of volume overload and coronary ischemia, renal insufficiency,
and/or NEC. Interventions that may be employed to increase pulmonary vascular resistance and improve the
balance between systemic and pulmonary circulations, including subambiant FiO2 and hypoventilation, are
discussed below in the section on preoperative management of HLHS. Neonates with left ventricular outflow
tract obstruction who are stable at presentation may be allowed to spontaneously breathe while awaiting
surgical intervention, with close monitoring of their cardiopulmonary status. These neonates will typically have
relatively high systemic oxygen saturations (greater than 90%) and quiet tachypnea, but will maintain
adequate peripheral pulses, urine output, and low
P.717
lactate levels. Diuretics and low dose inotropic support are occasionally needed to alleviate pulmonary edema
and to support the volume-loaded ventricular myocardium. In neonates with right ventricular outflow tract
obstruction, the ductus arteriosus is often somewhat stenotic and takes a tortuous course from the aorta to the
pulmonary artery, related to the in utero blood flow pattern, which typically will prevent torrential pulmonary
blood flow. Thus marked pulmonary overcirculation is less of a clinical concern in ductal-dependent neonates
with right ventricular outflow tract obstruction.
TABLE 34-3 SELECTED COMMON CHROMOSOMAL ABNORMALITIES AND GENETIC SYNDROMES
ASSOCIATED WITH CONGENITAL HEART DISEASE (SEE ALSO TABLE 33-5)
Patient Group
Incidence of CHD
Most Common Lesions
Trisomy 21 (Down S.)
40-50%
1
AVSD
2
VSD
3
ASD
Trisomy 18 (Edwards S.)

Trisomy 13
4p- (Wolf-Hirschhorn S.)
5p- (Cri-du-Chat S.)
Monosomy X (Turner S.)
Noonan S.
Holt-Oram S.
90%
80%
50%
30%
20%
66%
90%
VSD
VSD
VSD
variable
CoA
PS
ASD
PDA
PDA
PDA
ASD
ASD
ASD
BAV
HCM
VSD
AS
Williams S.
22q11 deletion (DiGeorge)
75%
80%
supravalvar AS
IAA
PPS
truncus
PS
TOF
Goldenhar S.
VATER/VACTERL
CHARGE association
Beckwith-Wiedmann S.
Marfan's S.
25%
variable
75%
common
100% (neonatal)
VSD
PDA
TOF
conotruncal
HCM
dilated AAo
MVP/MR
AR
MVP
AR, aortic regurgitation; AS, aortic valve stenosis; AAo, ascending aorta; ASD, atrial septal defect;
AVSD,
atrioventricular septal defect; BAV, bicuspid aortic valve; CHARGE, acronym for coloboma, heart
defects,
atresia of the posterior choanae, retarded growth/development, genital hypoplasia, ear anomalies/
deafness;
CoA, coarctation of aorta; HCM, hypertrophic cardiomyopathy; IAA, interrupted aortic arch; MVP,
mitral valve prolapse; MR, mitral regurgitation; PDA, patent ductus arteriosus; PPS, peripheral
pulmonary
stenosis; PS, pulmonary valve stenosis; S, syndrome; TOF, tetralogy of Fallot; VATER/VACTERL,
acronyms
for vertebral anomalies, anal atresia, cardiac defects, tracheoesophageal fistula, esophageal atresia,
renal
defects, limb defects; VSD, ventricular septal defect.
Inotropic support is occasionally necessary for neonates with congenital heart disease, particularly those with
volume loaded single ventricles, but escalating inotropic or vasopressor requirements often signify a noncardiac
problem or inadequate systemic cardiac output as a result of Qp/Qs imbalance. Inotropic agents are discussed in
the postoperative section on low cardiac output.
Prematurity
The premature infant with critical congenital heart disease presents a significant challenge. Problems inherent to
prematurity, including temperature instability, decreased nutritional reserve, intraventricular hemorrhage,
respiratory distress syndrome, infection, and NEC, may develop during the pre- and postoperative course. In
preterm infants with critical congenital heart disease, a management strategy of prolonged medical therapy to
achieve weight gain prior to surgery is fraught with complications related to infection, heart failure and feeding
intolerance (43,44). Advances in cardiopulmonary bypass techniques and miniaturization of surgical equipment
allow for the safe conduct of open-heart surgery in these patients. Several reports depict the recent experience
with palliative or definitive surgical intervention in premature infants with critical congenital heart disease. Reddy
and associates reported the University of California at San Francisco experience with complete (two-ventricle)
repair in 102 infants with significant congenital heart disease weighing less than 2,500 grams. These
investigators found higher preoperative morbidity in the group of patients having delayed repair. The early and
late mortalities were 10 and 8%, respectively, and no patient had postbypass intracranial hemorrhage (44).
Investigators from the same center reported a series of 20 very low birth weight (LBW)(< 1,500 gm) infants
with symptomatic congenital heart disease who underwent complete repair with only two deaths (10%) (45).
Rossi and associates reported a series of infants weighing less than 2 kg, 73% of which were born prematurely.
Hospital survival was 83%, and 24% of the patients had neurologic complications (46). Bacha and associates, in
a study of 18 infants weighing less than 2 kg with coarctation of the aorta, found that repair can be
accomplished with low operative mortality, but the residual or recurrent coarctation rate was 44% (47). Results
for single ventricle palliation are less encouraging. The reported mortality for infants weighing less than 2.5 kg
at the time of Norwood operation for HLHS or other single ventricle variants is 45% to 51% (48,49). Thus,
although prematurity and LBW complicate the care of infants with significant heart disease, early intervention,
including corrective surgery, can be performed with acceptable morbidity and mortality.
Nutrition
Although the enteral route is the preferred method for providing nutrition to infants in the intensive care unit,
those with critical congenital heart disease awaiting intervention often have several pathophysiologic features
that must be considered prior to initiation of enteral feedings. The combination of myocardial dysfunction,
cyanosis and falling pulmonary vascular resistance, the later allowing increased diastolic runoff through the
ductus arteriosus, may result in inadequate mesenteric oxygen delivery. In one large, retrospective study,
approximately 3% of neonates admitted to a busy cardiac intensive care unit developed NEC (50). Independent
risk factors for developing NEC were prematurity, a history of resuscitation from severe cyanosis or shock,
HLHS, or the presence of diastolic runoff through the ductus arteriosus or other aorto-pulmonary connections
(50). In part as a result of the absence of randomized studies, the use of enteral nutrition in infants with ductaldependent congenital heart disease remains controversial and clinical practice varies widely. The risk-benefit
ratio for providing enteral nutrition must be individually determined for each patient. One approach is to avoid
enteral feeding in those infants who have one or more of the aforementioned risk factors. It is also reasonable to
consider enteral feedings in those infants with an acceptable diastolic blood pressure despite the presence of a
patent ductus arteriosus, provided that additional risk factors are not present.
Congenital Complete Heart Block

Congenital complete heart block may be associated with structural heart disease including atrioventricular canal
defects and complex lesions with ventricular inversion. In neonates with normal intracardiac anatomy, there is a
high incidence of maternal autoimmune disease, particularly systemic lupus erythematosus (51). Congenital
complete heart block is diagnosed electrocardiographically when atrioventricular dissociation is present with an
atrial rate that is appropriate for age and a slow ventricular rate. The diagnosis may be made in utero by fetal
echocardiography and may be associated with hydrops fetalis. Dual chamber epicardial pacemaker placement is
indicated when third degree congenital atrioventricular heart block is associated with a wide QRS escape rhythm,
complex ventricular ectopy, or ventricular dysfunction. A pacemaker is also indicated in an infant with a
ventricular rate less than 50 to 55 beats per minute, or with congenital heart disease and a ventricular rate less
than 70 beats per minute (52).
P.718
PREOPERATIVE PHYSIOLOGY AND MANAGEMENT: SPECIFIC

LESIONS
Right Ventricular Outflow Tract Obstruction
Critical Pulmonary Valve Stenosis
In neonates with critical pulmonary valve stenosis, the pulmonary valve leaflets are thickened and fused to a
variable extent, and their mobility is decreased (Fig. 34-1, see also color plate). The pulmonary valve annulus is
hypoplastic in severe cases. The pulmonary arteries are usually of normal size, but may be stenotic or dilated.
The right ventricle is often hypertrophied and mildly hypoplastic. Significant tricuspid regurgitation, and a right
to left atrial shunt are present (53). Infants with isolated critical pulmonary valve stenosis will develop
significant cyanosis (See Color Plate) upon closure of the ductus arteriosus. Such infants are dependent on a
PGE1 infusion to maintain ductal patency for adequate pulmonary blood flow. Although open surgical valvotomy
was routinely performed in the past to relieve right ventricular outflow tract obstruction, most of these infants
are candidates for a balloon valvuloplasty in the cardiac catheterization laboratory. Balloon valvuloplasty is
successful in approximately 60% to 80% of cases (53,54,55,56,57,58). Neonates who fail an attempted balloon
valvuloplasty may have smaller right heart structures, and are referred for a surgical valvotomy +/- a systemic
to pulmonary artery shunt (59,60).
Tetralogy of Fallot
Tetralogy of Fallot, the most common cyanotic congenital heart lesion, is comprised of an anterior malalignment
VSD in a heart with two ventricles and aortic to mitral fibrous continuity. Anatomic details that need to be
clarified at the time of diagnosis include the severity and location of right ventricular outflow tract obstruction,
the pulmonary artery anatomy, the presence of additional VSDs, the sidedness of the aortic arch (rightward in
25%), and the coronary artery anatomy. The left coronary artery may arise from the right coronary artery and
cross the right ventricular outflow tract in 5% of cases. Complete anatomic information is usually obtainable by
transthoracic echocardiogram (33). If uncertainty exists about the coronary artery anatomy, some surgeons
may request a cardiac catheterization, as great care must be taken not to injure the coronary vessels during
right ventricular outflow tract reconstruction. Chromosome 22q11 deletion is present in 15% of patients with
tetralogy of Fallot, and those with a right aortic arch are at higher risk (41).
Figure 34-1 Critical pulmonary valve stenosis. A. Two-dimensional echocardiogram from the parasternal long
axis view demonstrating thickened and doming pulmonary valve leaflets. Note the poststenotic dilation of the
main pulmonary artery. B. Application of color Doppler during ventricular systole demonstrates a turbulent flow
jet originating at the pulmonary valve leaflets and extending into the main pulmonary artery. MPA, main
pulmonary artery; PV, pulmonary valve; RVOT, right ventricular outflow tract. (See color plate)
In neonates with tetralogy of Fallot, a wide spectrum of presentation exists, depending in large part upon the
degree of malalignment of the conal septum into the right ventricular outflow tract. Infants with a minimal
degree of obstruction to pulmonary blood flow, commonly referred to as pink tets, are usually asymptomatic
and fairly well oxygenated (systemic saturations > 85%) soon after birth. They are commonly observed in the
hospital until the ductus arteriosus closes and then sent home to await surgical repair within the first 6 months
of life. Occasionally a pink tet will mimic the pathophysiology of an infant with a large VSD and develop
congestive heart failure within the first few months of life as pulmonary vascular resistance falls. More
commonly these patients develop progressive right ventricular outflow tract obstruction and cyanosis. The exact
timing of repair varies depending
P.719
upon several variables including the development of increasing cyanosis, hypercyanotic spells (TET spells), and
institutional preference (2,61). Infants with tetralogy of Fallot and more severe right ventricular outflow tract
obstruction or pulmonary atresia will develop excessive cyanosis (oxygen saturation < 75%-80%) upon closure
of the ductus arteriosus. Such patients will require surgical intervention during the initial hospitalization. Either a
complete repair or a systemic to pulmonary shunt is performed, depending on the details of the anatomy and
surgeon preference (62).
Hypercyanotic episodes, or TET spells, are potentially life-threatening events notable for significant
desaturation and irritability as a result of an acute decrease in pulmonary blood flow. The precise etiology of
these spells is unclear but has been attributed to spasm of the infundibular conus, an imbalance between the
systemic and pulmonary vascular resistances, increased systemic venous return, and/or tachycardia, all of which
may lead to a progressive cycle of decreased pulmonary blood flow, increased cyanosis, and eventually
metabolic acidosis. Although TOF spells can occur at any age, the incidence seems to increase after 4 to 6
months of age, which is one of the reasons that early complete repair before this age is commonly
recommended. TET spells can be triggered by medical procedures including placement of intravenous catheters
and cardiac catheterization. Infants with tetralogy of Fallot have a fixed component of obstruction to pulmonary
blood flow, and thus will normally develop transient desaturation with agitation and feeding as a result of
increased oxygen consumption and cardiac output. Clinical judgment is required to differentiate a true TET spell
from these common and benign desaturation episodes. The absence of a murmur during a true TET spell,
signifying a substantial reduction of blood flow across the right ventricular outflow tract, may be useful in this
regard. Treatment strategies are implemented with the goal of decreasing patient agitation and heart rate,
increasing systemic vascular resistance and pulmonary blood flow, and correcting metabolic acidosis (Table 344). Prompt surgical intervention is indicated once an infant has had a TET spell.
Absent pulmonary valve syndrome, identified by the presence of rudimentary pulmonary valve leaflets and free
pulmonary insufficiency, is a relatively rare lesion, present in approximately 3% of all patients with tetralogy of
Fallot. This lesion is most notable for in utero development of aneurysmal dilation of the pulmonary arteries,
which may occur as a result of the lack of a ductus arteriosus or as a result of the pulsatile blood flow pattern
across the right ventricular outflow tract. When compared with infants with typical tetralogy of Fallot, there may
be a higher incidence of chromosome 22q11 deletion in those with absent pulmonary valve (63). A nearly
pathognomonic to-and-fro (sawing wood) murmur is heard as a result of pulmonary annulus stenosis and free
pulmonary insufficiency (Fig. 34-2, see also color plate). Such patients may have minimal if any respiratory
symptoms and behave much like a pink tet, or present with severe problems with oxygenation and ventilation
soon after birth as a result of compression of the bronchi by the dilated pulmonary arteries (64). Intubation,
mechanical ventilation (with judicious use of positive end-expiratory pressure), and deep sedation may be life
saving in neonates with symptomatic airway obstruction. Prone positioning may be advantageous as gravity will
allow the pulmonary arteries to fall off the mainstem bronchi, thus permitting adequate gas exchange (65).
Prompt surgical repair is indicated in symptomatic neonates.
TABLE 34-4 TREATMENT FOR HYPERCYANOTIC EPISODES (TOF SPELLS)
Intervention
Effect on Pathophysiology
Knee-chest position
systemic vascular resistance
Oxygen
pulmonary vascular resistance
Morphine
agitation
Ketamine
agitation, systemic vascular resistance
Propranolol
infundibular spasm, heart rate
Sodium bicarbonate
metabolic acidosis
Phenylephrine
systemic vascular resistance

rescue therapy when above measures fail
Cardiopulmonary bypass
Pulmonary Atresia, Ventricular Septal Defect, Major

Aortopulmonary Collateral Arteries
Infants with pulmonary atresia, VSD and MAPCAs represent a challenging subset of infants with pulmonary
atresia. This lesion is also referred to as tetralogy of Fallot, pulmonary atresia and MAPCAs. In addition to the
anatomic features of tetralogy of Fallot as described above, the true pulmonary arteries are often diminutive or
absent, and a variable number of aortopulmonary collaterals arise from the aorta or bracheocephalic vessels to
supply pulmonary blood flow. Infants with this lesion will present with either cyanosis or congestive heart failure,
depending upon the cardiac anatomy (66). The physical examination may be notable for the presence of a
widely radiating continuous murmur produced by the aortopulmonary collateral blood flow. These infants are
generally not dependent upon prostaglandin infusion unless a major aortopulmonary
P.720
collateral vessel arises from the ductus arteriosus. Cardiac catheterization is required to identify all sources of
pulmonary blood flow. Surgical strategy must then be individualized depending upon the pulmonary artery and
collateral vessel anatomy. True pulmonary arteries, MAPCAs, or both may supply blood to individual segments of
the lungs. The ultimate goal is to encourage true pulmonary artery growth and incorporate as many arterial
blood vessels and lung segments into the repair as possible to maximize the effective pulmonary arterial
vascular bed, and thus minimize right ventricular hypertension following VSD closure. In situations in which dual
blood supply to a lung segment exists, the MAPCAs are coil occluded in the cardiac catheterization laboratory or
ligated at the time of surgery to eliminate left to right shunting. As with other conotruncal defects, these infants
should undergo genetic testing for chromosome 22q11 deletion.
Figure 34-2 Tetralogy of Fallot with absent pulmonary valve. A. Two-dimensional echocardiogram from the
parasternal long axis view demonstrating rudimentary pulmonary valve leaflets and aneurysmal dilation of the
main pulmonary artery. B. Application of color Doppler during ventricular diastole revealing free regurgitation
from the main pulmonary artery into the right ventricular outflow tract. MPA, main pulmonary artery; PV,
pulmonary valve; RVOT, right ventricular outflow tract. (See color plate)

Pulmonary atresia is characterized by a membranous or plate-like obstruction of the right ventricular outflow
tract at the level of the pulmonary valve, with variable degrees of hypoplasia of the tricuspid valve and right
ventricle, and occasional coronary artery fistulous connections to the right ventricle. Systemic venous return to
the right atrium flows through an obligatory atrial communication to the left atrium. Infants with pulmonary
atresia and intact ventricular septum will manifest with significant cyanosis upon closure of the ductus
arteriosus, and are thus dependent on prostaglandin infusion. Intervention in these patients depends on two key
issues: the adequacy of the right ventricular size to support a two-ventricular circulation, and on the presence of
right ventricle dependent coronary circulation. The right ventricle must have adequate size to support the
entire right heart circulation if these patients are to be considered for a two-ventricle repair. The size of the
tricuspid valve annulus plays an important role in the decision making process (67). Infants with pulmonary
atresia and intact ventricular septum, particularly those with small tricuspid valve annulus size, may have
significant fistula communications between the right ventricle and the coronary circulation, known as coronarycameral fistulae, which may be identified by echocardiogram and further characterized by cardiac catheterization
(68,69). If coronary stenoses exist such that a significant amount of coronary blood flow to the myocardium is
derived from and dependent upon the coronary cameral fistulae, then right ventricular dependent coronary
circulation is present. If the right ventricular outflow tract obstruction is relieved in an infant with right
ventricular dependent coronary circulation, coronary perfusion pressure falls and the patient may develop a
myocardial infarction. Thus, if the right ventricle is not of adequate size, or if right ventricle dependent coronary
circulation exists, these patients are referred for a systemic to pulmonary artery shunt in anticipation of an
eventual Fontan operation. If a complete two-ventricle repair is deemed appropriate, this may be performed as a
primary operation in the neonatal period, although more commonly a systemic to pulmonary artery shunt and
right ventricular outflow tract decompression is initially performed to encourage right ventricular growth, with
complete repair in later infancy (70). Some infants with membranous pulmonary atresia and intact ventricular
septum who meet criteria for a two ventricular repair may undergo perforation of the pulmonary valve using a
stiff wire, a radiofrequency ablation catheter, or a laser catheter, followed by balloon valvuloplasty (54,58,71).
In well-selected patients, the use of transcatheter intervention
P.721
may eliminate the need for surgical intervention. The decision to attempt transcatheter intervention is made
considering patient size, co-morbidities, anatomic details, and institutional experience.

Ebstein's anomaly of the tricuspid valve is a malformation in which the septal and posterior leaflets of the
tricuspid valve are displaced to a variable extent into the anatomic right ventricle, and the anterior leaflet,
although not inferiorly displaced, is redundant. Tricuspid regurgitation of variable severity occurs (Fig. 34-3, see
also Color Plate). One or more accessory connections may be present at the tricuspid valve annulus, creating the
necessary substrate for atrio-ventricular reentrant tachycardia. The clinical presentation is usually one of rightsided congestive heart failure and arrhythmias in adolescence or early adulthood. A small subset of infants with
Ebstein's anomaly, those with severe tricuspid regurgitation, a large atrialized portion of the right ventricle,
severe cardiomegaly, and pulmonary hypoplasia are critically ill soon after birth and have a high mortality (72).
These neonates will present with severe cyanosis as a result of functional right ventricular outflow tract
obstruction, and right to left shunting at the atrial level. Infants with significant cyanosis should be started on
PGE1, and judicious use of positive end-expiratory pressure is required to re-expand the lungs, which are
compressed by the generous heart size (73). Additionally, standard measures including inhaled nitric oxide
should be employed to decrease pulmonary vascular resistance and cyanosis. In many of these patients, as
pulmonary vascular resistance falls over time, forward flow increases across the right ventricular outflow tract,
cyanosis decreases, and surgery can be deferred. For those neonates with refractory cyanosis, referral is made
for either a primary complete repair, consisting of a reduction atrioplasty, fenestrated closure of the atrial
septum and tricuspid valvuloplasty, or palliation toward the Fontan pathway starting with a systemic to
pulmonary artery shunt (73,74,75).
Figure 34-3 Ebstein's anomaly of the tricuspid valve. A. Two-dimensional echocardiogram from the apical fourchamber view demonstrating significant displacement of the tricuspid valve leaflets into the right ventricle. B.
During ventricular diastole, color Doppler demonstrating free retrograde flow across the tricuspid valve annulus.
ARV, atrialized right ventricle; LA, left atrium; LV, left ventricle; RA, right atrium; TV, tricuspid valve. (See color
plate)
Left Ventricular Outflow Tract Obstruction

Critical Aortic Valve Stenosis
Isolated congenital aortic valve disease may present in a variety of ways, depending primarily upon the severity
of stenosis. A bicuspid aortic valve without stenosis or insufficiency may be an incidental finding in an otherwise
healthy patient. At the other extreme, newborns with critical aortic valve stenosis develop shock upon closure of
the ductus arteriosus. The murmur is typically softer than expected as a result of the low cardiac output state,
and the gradient estimated by echocardiography may not correlate with the severity of the stenosis. Initiation of
PGE1 infusion allows blood from the right ventricle to provide systemic perfusion until an intervention can be
performed (76). Once the infant is stabilized, echocardiographic information must be carefully reviewed to
predict whether the left ventricle is of adequate size to sustain the systemic circulation following a two-ventricle
repair (77). Infants thought to be suitable for a two-ventricle repair usually will be referred for a balloon aortic
valvuloplasty; although a surgical aortic valvotomy is an acceptable option (78,79,80). If hypoplasia of multiple
left sided structures suggests that the left ventricle will not be able to support the circulation, then a Norwood
operation or heart transplantation may be considered.
Critical Coarctation of the Aorta

A coarctation of the aorta is present when a significant narrowing exists in the thoracic aorta just distal to the
left
P.722
subclavian artery. A posterior shelf is present opposite the insertion site of the patent ductus arteriosus.
Coarctation of the aorta may exist in isolation or can be associated with aortic arch hypoplasia, a VSD, or other
complex intracardiac lesions. When present in isolation, coarctation of the aorta is critical when ductal
dependent systemic blood flow exists, which represents only about 10% of cases. Neonates with isolated critical
coarctation of the aorta also will present with shock if the diagnosis is not made before significant constriction of
the ductus arteriosus occurs (Fig. 34-4). Prior to the development of shock, the diagnosis should be suspected
by the detection of a brachial-femoral pulse discrepancy, which can be confirmed by measuring an arm-leg blood
systolic pressure gradient. A systolic blood pressure gradient between the upper and lower extremities of greater
than 10 mm Hg is clinically significant. In approximately 5% of infants with isolated coarctation of the aorta, the
right subclavian artery will arise aberrantly from the descending thoracic aorta, distal to the coarctation. Thus it
is important to palpate pulses and measure blood pressure in both upper extremities in all infants with
suspected coarctation. Occasionally differential cyanosis will be present as the lower body is being perfused by
deoxygenated blood from the ductus arteriosus. PGE1 infusion will almost always reopen the ductus arteriosus
and allow recovery of end-organ function prior to surgical repair (76). In the absence of data from randomized
clinical trails, the method of nutritional support for these patients should be individualized and deserves
comment. Those neonates with critical coarctation who present in shock with metabolic acidosis and laboratory
evidence of end-organ dysfunction should be maintained on bowel rest prior to surgery, and nutritional support
is provided with total parenteral nutrition. However, enteral feeds may be considered in the subset of infants
who are diagnosed prior to the development of shock, started on PGE1, and subsequently have reasonable
femoral pulses and diastolic blood pressure distal to the coarctation.
Figure 34-4 Schematic diagram of hemodynamic data and oxygen saturation levels in a neonate with critical
coarctation of the aorta. A: Hemodynamic data and oxygen saturation levels prior to ductal closure. B:
Hemodynamic data and oxygen saturation levels following ductal closure. Note the development of systemic
hypertension and left atrial hypertension, the persistence of pulmonary hypertension, and the decrease in
cardiac output as demonstrated by the lower oxygen saturation level in the inferior vena cava (a, atrial; v,
ventricle; m, mean). Reprinted from Rudolph AM. Aortic arch obstruction. In: Congenital diseases of the heart:
clinical-physiological considerations, 2nd ed. Armonk, NY: Futura, 2001:382-383, with permission.

HLHS exists when there is severe mitral and aortic valve stenosis or atresia, a diminutive left ventricle,
ascending aorta, and aortic arch, and often a juxtaductal coarctation of the aorta. Following birth there are two
key communications that need to remain open to permit cardiopulmonary stability. A patent ductus arteriosus
(PDA) is necessary to allow blood flow from the pulmonary artery to reach the systemic circulation, and shock
will develop once the ductus arteriosus constricts. An adequate ASD is also necessary to allow egress of the
pulmonary venous
P.723
blood flow from the left atrium to the right atrium. The 5% of infants with HLHS who have a severely restrictive
or intact atrial septum will develop profound cyanosis, pulmonary edema, and shock almost immediately
following delivery.
Initially, most patients with HLHS (assuming an unrestrictive or mildly restrictive atrial communication) will be
hemodynamically stable following the initiation of a PGE1 infusion. Over several days, however, as pulmonary
vascular resistance falls, the ratio of pulmonary to systemic blood flow (Qp/Qs) will increase, leading to an
increase in arterial oxygen saturation. Poor systemic circulation may manifest, and the neonate may develop
right ventricular volume overload, and coronary and end-organ ischemia. The Qp/Qs may be calculated by the
following equation which is derived from the Fick principle: Qp/Qs =(SaO2-SvO2)/(PvO2-PaO2), in which SaO2
=aortic saturation; SvO2 =mixed venous saturation (estimated by the superior vena cava oxygen saturation);
PvO2 =pulmonary venous oxygen saturation (assumed to be > 95% if not directly measured); and PaO2
=pulmonary artery oxygen saturation (assumed equal to the aortic saturation). Most preoperative neonates will
not have a catheter positioned in the superior vena cava to measure SvO2, and thus the Qp/Qs cannot be
precisely calculated. Determination of an absolute value for Qp/Qs is of less importance than having an
appreciation for the physiologic disturbance that develops as these neonates become progressively over
circulated. Theoretical computer models of newborns with HLHS demonstrate that, beyond a certain point, slight
increases in arterial oxygen saturation are associated with significant decreases in oxygen delivery, and that
maximal oxygen delivery occurs at a Qp/Qs less than 1 (Fig. 34-5 and 34.6) (81,82). This model also
demonstrates that a SvO2 less than 40%, or a SaO2-SvO2 difference of more than 40%, is likely associated with
a severe derangement in systemic oxygen delivery, related to either a high Qp/Qs or low cardiac output (82).
Figure 34-5 Systemic arterial oxygen saturation vs. systemic oxygen (O2) delivery in the theoretical newborn
with hypoplastic left heart syndrome. A computer model was used to generate the curves by setting the cardiac
output (CO) at 300 or 450 mL/kg/min and varying Qp/Qs from 0.2 to 10. The short line on each curve
represents the point at which Qp/Qs = 1. As SaO2 increases, oxygen delivery increases and reaches a peak,
and then decreases rapidly. Peak oxygen delivery occurs at a Qp/Qs 1. Reprinted from Barnea O, Santamore
WP, Rossi A, et al. Estimation of oxygen delivery in newborns with a univentricular circulation. Circulation
1998;98:1407-13, with permission.
In the typical neonate with HLHS, interventions that lower the pulmonary vascular resistance, such as
supplemental oxygen and hyperventilation, may increase Qp/Qs and should be avoided. Patients with pulmonary
over-circulation (high Qp/Qs) and evidence of poor end-organ perfusion (e.g., poor urine output, rising lactate
levels, diminished peripheral pulses) may be treated with sub-ambient FiO2, obtained using supplemental
inhaled nitrogen to increase pulmonary vascular resistance and thus restore the balance between systemic and
pulmonary blood flow (83,84). The supplemental nitrogen may be delivered through the endotracheal tube or
through a nasal cannula in neonates who are spontaneously breathing. Pulmonary vascular resistance may also
be increased in such patients using supplemental inspired carbon dioxide (CO2) (85). In a prospective,
randomized, crossover study, the impact of hypoxia (17% FiO2) vs. hypercarbia (2.7% FiCO2) on systemic
oxygen delivery in preoperative neonates with HLHS was evaluated (86). All patients were receiving mechanical
ventilation, deep sedation or anesthesia, and
P.724
neuromuscular blockade. The study found that both hypoxia and hypercarbia were effective in reducing Qp/Qs
ratio (Fig. 34-7). Hypercarbia increased cerebral oxygen saturation, mean arterial blood pressure, oxygen
delivery and, although not directly measured, cardiac output, whereas hypoxia did not change these variables,
suggesting that hypercarbia may be the preferred therapy (Fig. 34-8) (86,87). However, administration of
supplemental CO2 requires the concurrent use of sedation and paralysis, thus limiting its applicability for most
neonates with HLHS prior to surgery.
Figure 34-6 Systemic oxygen (O2) delivery vs. Qp/Qs in the theoretical newborn with hypoplastic left heart
syndrome. A computer model was used to generate the curves by setting the cardiac output (CO) at 300 or 450
mL/kg/min. Note that increasing cardiac output can increase oxygen delivery, and that oxygen delivery
decreases significantly once Qp/Qs exceeds 1. Reprinted from Barnea O, Santamore WP, Rossi A, et al.
Estimation of oxygen delivery in newborns with a univentricular circulation. Circulation 1998;98:1407-1413,
with permission.
Neonates with HLHS and myocardial dysfunction should receive inotropic support to improve cardiac output and
systemic oxygen delivery. However, care should be taken to avoid inotropic agents that increase systemic
vascular resistance because this will favor pulmonary blood flow over systemic blood flow. Low-dose dopamine
or milrinone are good inotropic agents for neonates with HLHS and myocardial dysfunction.
The neonate with HLHS and a severely restrictive or intact atrial septum will present with profound cyanosis and
shock immediately after birth as a result of obstruction of blood egress from the left atrium. The short- and longterm outlook for such infants has historically been poor, regardless of whether they undergo the Norwood
operation or are listed for heart transplantation, due in large part to the in utero development of pulmonary
venous hypertension and lymphatic abnormalities that persist following birth (88,89,90). The immediate
management of these critically ill neonates involves prompt intervention to decompress the left atrium.
Emergent surgical intervention, involving either open atrial septectomy or a Norwood procedure, has usually
resulted in early death. A Brockenbrough atrial septoplasty (transcatheter, trans-septal needle puncture followed
by serial balloon dilation of the new ASD and possibly stent placement) will serve to decompress the left atrium,
increase pulmonary venous return and pulmonary blood flow, and alleviate cyanosis (35). Following this
procedure, the patient may be medically managed for a few days with a PGE1 infusion and diuretics to allow
pulmonary vascular resistance to fall and pulmonary edema to improve prior to the Norwood operation. This
multidisciplinary approach to the neonate with HLHS and severely restrictive or intact atrial septum has resulted
in a contemporary 53% survival rate following the Norwood operation, which is better than results obtained with
other management strategies for this lesion, but remains poor when compared with the outcome of neonates
without atrial obstruction who undergo the Norwood operation (91). Theoretically, alleviation of the left atrial
hypertension in utero may improve outcome in patients with HLHS and a severely restrictive or intact atrial
septum. Fetal intervention to dilate the atrial septum using percutaneous, transcatheter needle puncture of the
atrial septum followed by balloon dilation appears to be feasible and is associated with minimal maternal risk,
but technical improvements and additional experience will be required before this procedure is widely adopted
(92).
Figure 34-7 Difference in Qp/Qs between condition (either hypoxia or hypercarbia) and baseline (mean
SEM) in preoperative neonates with hypoplastic left heart syndrome. Reprinted from Tabbutt S, Ramamoorthy
C, Montenegro LM, et al. Impact of inspired gas mixtures on preoperative infants with hypoplastic left heart
syndrome during controlled ventilation. Circulation 2001; 104:I159-I164, with permission.
Figure 34-8 Difference in oxygen delivery between condition (either hypoxia or hypercarbia) and baseline
(mean SEM) in preoperative neonates with hypoplastic left heart syndrome. Oxygen delivery was calculated
as SaO2 (SaO2-SvO2), in which SaO2 and SvO2 were directly measured. Reprinted from Tabbutt S,
Ramamoorthy C, Montenegro LM, et al. Impact of inspired gas mixtures on preoperative infants with
hypoplastic left heart syndrome during controlled ventilation. Circulation 2001;104:I159-I164, with permission.
Figure 34-9 Hospital survival following stage 1 palliation in patients with hypoplastic left heart syndrome by
year at a single center. Hospital survival improved coincident with the application of new treatment strategies
beginning in July of 1996 (arrow). S1P, stage 1 palliation (Norwood operation). Reprinted from Tweddell JS,
Hoffman GM, Mussatto KA, et al. Improved survival of patients undergoing palliation of hypoplastic left heart
syndrome: lessons learned from 115 consecutive patients. Circulation 2002;106: I82-I89, with permission.
P.725
P.726
Initial reports suggested that severe tricuspid regurgitation was a significant risk factor for infants with HLHS
(93,94). However, in the preoperative infant with HLHS, the severity of tricuspid regurgitation may be in part
related to a high Qp/Qs, with right ventricular volume overload and stretching of the tricuspid valve annulus.
Following the Norwood operation, the Qp/Qs may decrease as a result of the fixed resistance to pulmonary blood
flow provided by the shunt, and subsequently the degree of tricuspid regurgitation may diminish. Thus, clinical
judgment and an appreciation of tricuspid valve morphology are required in such cases.
Figure 34-10 The three types of interrupted aortic arch. In type A, the interruption is at the aortic isthmus
between the left subclavian artery and the ductus. In type B, the interruption is at the distal aortic arch
between the left carotid and left subclavian arteries. In type C, the interruption is at the proximal aortic arch
between the innominate and left carotid arteries. Type B is the most common form of this lesion. Type C is
rare. Reprinted from Chang AC, Starnes VA. Interrupted aortic arch. In: Chang AC, Hanley FL, Wernovsky G, et
al., eds. Pediatric cardiac intensive care. Baltimore: Williams and Wilkins, 1998:243-247, with permission.
Several centers have recently reported overall survival rates of more than or equal to 90% following initial
surgical palliation for HLHS (Fig. 34-9) (4,95). Contemporary risk factors contributing to higher mortality
following the Norwood operation include prematurity, the presence of multiple congenital anomalies, and an
intact atrial septum (48,88). Although palliative care is still offered at some centers for neonates with HLHS,
fewer clinicians and parents are opting for no intervention given the improving surgical results (96). Some
centers offer heart transplantation as a primary option for HLHS (97). Problems with this strategy include the
high morbidity and mortality incurred while waiting for a heart to become available from a very limited donor
pool, suboptimal neurologic outcome in many patients, and complications inherent to heart transplant
(98,99,100). For these reasons, many pediatric cardiology centers recommend the Norwood operation as the
initial surgical intervention in neonates with HLHS, and heart transplantation is typically reserved for the small
subset that develop subsequent irreversible myocardial failure.
Interrupted Aortic Arch

An interrupted aortic arch (IAA) is most commonly associated with a VSD, but also is seen with a variety of
other congenital cardiac malformations, including truncus arteriosus and TGA (101). If a VSD is not present in a
patient with an IAA, an aorto-pulmonary window is likely to be found. IAA is classified by the location of the arch
interruption in relation to the bracheocephalic vessels (Fig. 34-10) (102). IAA type B is the most common
subtype (>50% of cases), and type C is rare. Chromosome 22q11 deletion is present in approximately 60% of
neonates with IAA type B (41). The infant with an unrecognized IAA will present with shock, similar to a critical
coarctation, upon closure of the ductus arteriosus. Infants with IAA are dependent upon ductal patency to
maintain blood flow to the lower body, and
P.727
preoperative management strategies to balance the systemic and pulmonary circulations mimic those used in
infants with HLHS. Early surgical intervention is warranted.
Mixing Lesions
Transposition of the Great Arteries
In d-TGA, the aorta arises from the anatomic right ventricle and the pulmonary artery arises from the anatomic
left ventricle. Approximately 40% of neonates with TGA have an associated VSD, which occasionally is a
malalignment type defect. If anterior malalignment exists, there may be associated right ventricular outflow
tract obstruction, aortic valvar stenosis, coarctation of the aorta or rarely an IAA. A posterior malalignment VSD
is associated with left ventricular outflow tract obstruction, pulmonary stenosis or atresia. Coronary artery
branching abnormalities are present in approximately 30% of cases.
In this unique parallel circulation, deoxygenated systemic venous blood returns to the right heart and is pumped
back to the systemic arterial circulation, and oxygenated pulmonary venous blood passes through the left heart
and is pumped back to the lungs. Unless there is adequate mixing between these parallel circulations, severe
cyanosis, metabolic acidosis and death will occur. Such intercirculatory mixing represents effective pulmonary
and systemic blood flows, and may take place at the atrial, ventricular, or great artery level (Fig. 34-11). Infants
with TGA and an intact ventricular septum typically have a patent foramen ovale (PFO) or ASD that allows some
mixing at the atrial level. However, a PGE1 infusion is usually necessary to open or maintain patency of the
ductus arteriosus, which will increase effective pulmonary blood flow, provided that the pulmonary vascular
resistance is lower than the systemic vascular resistance and that there is an adequate atrial communication.
Occasionally infants with TGA may develop severe cyanosis and shock following the initiation of PGE1 infusion,
as a result of increasing blood volume returning to the left atrium and functional closure of the foramen ovale. If
this occurs, or if the PFO is restrictive, as assessed by echocardiography, and excessive cyanosis is present, an
emergent balloon atrial septostomy should be performed to enlarge the atrial communication (Fig. 34-12, see
also Color Plate) (34). The balloon atrial septostomy may be performed at the bedside using echocardiographic
guidance, or in the cardiac catheterization laboratory (103). Excessive cyanosis may persist despite a technically
successful balloon atrial septostomy. High pulmonary vascular resistance may limit effective pulmonary blood
flow, and measures should be taken to lower pulmonary vascular resistance. Because the majority of systemic
blood flow comes from the systemic venous circulation in TGA, neonates who remain excessively cyanotic
following a balloon atrial septostomy will also improve following maneuvers that will increase the mixed venous
oxygen saturation. These include interventions to decrease oxygen consumption, such as sedation and paralysis,
and improve oxygen delivery, such as correcting anemia and administering inotropic agents. To summarize,
infants with TGA and intact ventricular septum who remain excessively cyanotic despite the initiation of PGE1
require a balloon atrial septostomy, and if necessary, interventions to lower pulmonary vascular resistance and
increase systemic venous oxygen saturation.
Figure 34-11 The circulation in transposition of the great arteries. A: The systemic and pulmonary circulations
are in series in the normal circulation, whereas they are in parallel in TGA. Solid arrows, relatively
unoxygenated blood; stippled arrows, oxygenated blood; dashed arrows, intercirculatory shunts. B: Circulation
schema demonstrating flows and shunts in infants with TGA/intact ventricular septum. Note that the anatomic
left to right shunt constitutes the effective SBF, and the anatomic right-to-left shunt constitutes the effective
PBF. Ao, aorta; IVC, inferior vena cava; LA, left atrium; LV, left ventricle; L R, left to right; RA, right atrium;
RV, right ventricle; R L, right to left; PA, pulmonary artery; PBF, pulmonary blood flow; PV, pulmonary veins;
SBF, systemic blood flow; SVC, superior vena cava. Reprinted from Paul MH, Wernovsky G. Transposition of the
great arteries. In: Emmanouilides GC, Riemenschneider TA, Allen HD, et al, eds. Moss and Adams' heart
disease in infants, children and Adolescents, including the fetus and young adult. Baltimore: Williams & Wilkins,
1995:1154-1225, with permission.
Many clinicians recommend that a semi-elective balloon atrial septostomy be performed in nearly all patients
with TGA and intact ventricular septum, even those without
P.728
excessive cyanosis. Once the atrial septum is enlarged, the PGE1 can usually be safely discontinued, thus
avoiding complications related to that medication and a large ductus arteriosus, such as apnea and NEC.
Furthermore, the left atrium is decompressed following a balloon atrial septostomy, which may allow the
pulmonary vascular resistance to fall prior to surgery. Finally, the infant may be transferred from the ICU to the
general ward to initiate feedings and await surgery. The major risks associated with the performance of a
balloon atrial septostomy include myocardial perforation or avulsion of the inferior vena cava from the right
atrium during pullback of the balloon, both of which rarely occur. Following initial stabilization of the neonate
with TGA and intact ventricular septum, an early arterial switch operation is required before the left ventricle
becomes deconditioned.
Figure 34-12 Restrictive atrial septum in a newborn with transposition of the great arteries. A. Two-dimensional
echocardiogram with color Doppler from the subcostal window demonstrating a tiny patent foramen ovale with
left to right flow across the atrial septum. B. Following successful balloon atrial septostomy, a wide
communication now exists between the left and right atria. LA, left atrium; PFO, patent foramen ovale; RA,
right atrium. (See color plate)
Infants with TGA and moderate to large VSD generally are well oxygenated and do not require a PGE1 infusion
provided that the ventricular outflow tracts are unobstructed. Early congestive heart failure typically develops,
and surgical repair is indicated within the first few weeks of life. A malalignment-type VSD may be associated
with clinically significant left or right ventricular outflow tract obstruction. Patients with sub-aortic obstruction
may have an associated coarctation of the aorta or an IAA. Such an infant may present with the very unique
finding of reverse differential cyanosis, with low oxygen saturation in the right arm and high oxygen saturation
in the leg. A PGE1 infusion is needed, followed by early surgical intervention. Infants with d-TGA and VSD who
are referred for surgery after the first few months of life (typically from underdeveloped nations) are at risk for
having developed pulmonary vascular obstructive disease. These patients may require a cardiac catheterization
to evaluate their pulmonary vascular resistance prior to surgery.
Tricuspid Atresia
Tricuspid atresia is present when there is agenesis of the tricuspid valve and an absent communication between
the right atrium and right ventricle. The presentation of infants with this lesion is variable and depends primarily
upon the presence and size of a VSD, whether the great arteries are normally related or transposed, and the
degree of ventricular outflow tract obstruction (Table 34-5) (104,105). These factors determine which of these
neonates will be dependent on a PGE1 infusion. All of the systemic venous blood must pass through an obligate
atrial communication to the left atrium, in which it mixes with the pulmonary venous blood. In the presence of
normally looped ventricles and normally related great arteries, the left ventricle then ejects blood to the aorta,
and through the VSD, if present, to the pulmonary artery. If the great arteries are transposed, the left ventricle
pumps blood to the pulmonary artery (unless pulmonary atresia is present), and through the VSD to the aorta.
Although all infants with tricuspid atresia eventually require a Fontan palliation, the type of initial operation
depends upon the extent of cyanosis or systemic outflow obstruction. For example, infants with tricuspid atresia,
normally looped ventricles and normally related great vessels, a large VSD and no or minimal right ventricular
outflow tract obstruction (type I-C) will be well oxygenated and develop early heart failure. In the neonatal
period, they require placement of a pulmonary artery band or transection of the pulmonary artery and
placement of a systemic to pulmonary artery shunt. Those with a moderate size VSD and a moderate degree of
right ventricular outflow tract obstruction (type I-B) will have a balanced circulation with an acceptable amount
of cyanosis. Such infants may be discharged home with the expectation of a bi-directional
P.729
Glenn or hemi-Fontan operation within the first six months of life. Infants with tricuspid atresia and a small or
absent VSD (type I-A or I-B) will have significant cyanosis and thus be dependent on a PGE1 infusion until a
systemic to pulmonary artery shunt is performed. Infants with tricuspid atresia and transposed great arteries, a
small or absent VSD (type II-A or II-B) will develop shock unless ductal patency is maintained. These infants
require a Damus-Kaye-Stansel operation (aortopulmonary anastomosis) or a modified Norwood operation with a
systemic to pulmonary artery shunt.
TABLE 34-5 CLASSIFICATION SYSTEM FOR TRICUSPID ATRESIA
Type and Description
Type I: Normally related great arteries

A. Intact ventricular septum with pulmonary atresia
B. Small ventricular septal defect with pulmonary stenosis
C. Large ventricular septal defect without pulmonary stenosis
Type II: d-Transposition of the great arteries
A. Ventricular septal defect with pulmonary atresia
B. Ventricular septal defect with pulmonary stenosis
C. Ventricular septal defect without pulmonary stenosis
Type III: I-Transposition of the great arteries
Frequency
70%80%
12%25%
3%6%
Total Anomalous Pulmonary Venous Return

Total anomalous pulmonary venous return (TAPVR) is a congenital heart defect in which all four pulmonary veins
drain to a systemic vein or the right atrium. TAPVR results from a failure of the common pulmonary vein to fuse
with the posterior surface of the left atrium early in fetal life. The pulmonary veins then decompress via the
coronary sinus or primitive venous structures, which eventually lead to the right atrium, in which mixing occurs
with the systemic venous blood. An anatomic classification system divides TAPVR into supracardiac, cardiac,
infracardiac, and mixed, depending upon which primitive venous pathways are employed. Supracardiac TAPVR
exists when all of the pulmonary veins come to a confluence behind the left atrium and drain via a vertical vein
to the innominate vein or superior vena cava. Occasionally the vertical vein is obstructed as it passes between
the left mainstem bronchus and left pulmonary artery. Cardiac TAPVR exists when the pulmonary veins drain to
the coronary sinus or right atrium. Obstruction is uncommon in cardiac TAPVR. In infracardiac TAPVR, the
pulmonary venous confluence drains inferiorly through the diaphragm and into the portal or hepatic veins, in
which obstruction is common. Patients with mixed TAPVR have drainage to more than one site. For example, the
left sided pulmonary veins may drain through a vertical vein to the innominate vein, although the right
pulmonary veins drain to the right atrium. An obligatory intracardiac (usually atrial) communication exists in
neonates with TAPVR to allow some oxygenated blood to reach the left heart and systemic arterial circulation.
TAPVR may be associated with significant intracardiac disease as is seen in patients with heterotaxy syndrome,
or may exist as an isolated lesion.
The presentation of isolated TAPVR is primarily based upon the presence and degree of obstruction between the
pulmonary veins and the right heart. Most patients with infracardiac TAPVR and some with supracardiac TAPVR
will have obstructed pulmonary venous pathways that result in pulmonary edema, pulmonary hypertension,
cyanosis, and significant respiratory distress soon after birth. The clinical presentation and CXR may mimic those
seen with neonatal pneumonia or respiratory distress syndrome, leading to delays in diagnosis. If a venous
blood gas is drawn from an umbilical venous catheter positioned just above the liver in a cyanotic newborn, and
the oxygen saturation level is in the high 90s, the diagnosis of infracardiac TAPVR should be suspected, as the
sample is reflective of pulmonary venous return. The umbilical venous catheter should be removed so that it
does not further obstruct pulmonary venous flow and cause clinical decompensation. Stabilization for the
neonate with obstructed TAPVR involves mechanical ventilation, sedation and urgent surgical intervention (106).
Hypoxemia may be abated somewhat by interventions that increase mixed venous oxygen saturation, including
pharmacologic paralysis, correction of anemia, and inotropic support. Although neonates with obstructed TAPVR
may be quite cyanotic, measures to lower pulmonary vascular resistance, including the use of hyperventilation
and nitric oxide (NO), are usually not beneficial. The use of PGE1 infusion in such patients is controversial, but
generally is discouraged. Some physicians feel that PGE1 may maintain patency of the ductus venosus in infants
with infracardiac obstructed TAPVR, thus allowing for better drainage back to the right heart. Other clinicians
believe that maintaining patency of the ductus arteriosus may allow left to right shunting at the arterial level (if
the pulmonary vascular resistance is less than the systemic vascular resistance) and thus exacerbate pulmonary
edema.
P.730
Infants with unobstructed isolated TAPVR and an adequate atrial communication will be mildly hypoxemic
without overt clinical cyanosis at birth but develop congestive heart failure (CHF) within a few weeks to months.
Surgical repair is performed within the first few days to weeks of life.
Truncus Arteriosus
Truncus arteriosus is present when a single (common) arterial trunk gives rise to the aorta, at least one
coronary artery, and at least one pulmonary artery. A single semilunar (truncal) valve is always present, and
an unrestrictive VSD is almost always present. About one-third of patients with truncus arteriosus have a
22q11.2 microdeletion (DiGeorge syndrome) that can be diagnosed with florescent in-situ hybridization (FISH)
(107). Several classification systems exist that may be used to describe the origins of the pulmonary arteries
from the arterial trunk and the presence or absence of an IAA (108). Newborns with truncus arteriosus who
present with mild cyanosis do not require a PGE1 infusion unless an IAA (about 10% of cases) or ductal origin of
one of the pulmonary arteries is present.
Infants with truncus arteriosus usually develop too much pulmonary blood flow and congestive heart failure
within a few weeks after birth as pulmonary vascular resistance falls and systemic vascular resistance increases,
unless stenosis is present at the origin of the pulmonary arteries from the aorta. Supplemental oxygen should be
avoided to minimize the risk of decreasing pulmonary vascular resistance. Congestive heart failure is
exacerbated in the presence of significant truncal valve insufficiency, which occurs in 50% of cases, or
significant truncal valve stenosis (less common). For this reason, and to prevent the development of pulmonary
vascular obstructive disease before surgery and minimize the incidence of pulmonary hypertension crises
postoperatively, complete surgical repair is generally performed prior to discharge from the nursery (8,109,110).
Left to Right Shunts

Atrial Septal Defect
ASDs may be present in isolation or with nearly any other type of congenital heart defect. ASDs are categorized
by location and include secundum, ostium primum and sinus venosus types. The most common type is the
secundum ASD, which exists when there is a deficiency in the septum primum in the central portion of the atrial
septum. Although secundum ASDs may become smaller and close spontaneously, particularly those that are less
than 5 mm in diameter, the ostium primum and sinus venosus defects will always require surgical intervention.
The ostium primum defects are located in the inferior region of the atrial septum, and are almost always
associated with an additional endocardial cushion abnormality. The sinus venosus defects usually occur at the
junction between the superior vena cava and the right atrium, and are almost always associated with partial
anomalous pulmonary venous return.
Isolated ASDs only rarely cause symptoms in infants, as significant left to right shunting does not occur until
right ventricular compliance falls below left ventricular compliance, and this often takes years to occur. Likewise
it is uncommon for patients with large ASDs to require oral medications (digoxin or diuretics). However, clinical
judgment regarding the medical management and timing of surgical intervention is required in the neonate with
a large ASD and additional medical problems (e.g., chronic lung disease) that complicate the clinical picture. The
primary indication for surgery in infancy is the presence of symptoms secondary to left to right shunting.
Toddlers require intervention if they have an ostium primum or sinus venosus ASD, or a secundum ASD with
evidence for volume overload of the right ventricle. Intervention is indicated in early childhood in asymptomatic
patients with significant left to right atrial shunts to prevent late complications including pulmonary vascular
obstructive disease, thromboembolic events, arrhythmias, and right heart failure. A variety of devices exist that
are used to close secundum ASDs in the catheterization laboratory, but the majority of clinical experience thus
far exists with older children (111).

VSDs may be found in a variety of complex congenital heart lesions, or may exist in isolation. VSDs are
categorized by location and size. VSDs may be located in the perimembranous, muscular, inlet or conal (outlet,
subarterial, supracristal) regions of the ventricular septum, and there may be more than one VSD present.
Malalignment of the conal septum may exist, resulting in variable degrees of right or left ventricular outflow
tract obstruction. Conal VSDs often develop involvement of the aortic valve over many years (112). The size
may be determined by an estimation of the diameter of the VSD in relation to the aortic annulus, and by degree
of restriction to blood flow as estimated by Doppler echocardiogram. Perimembranous and muscular VSDs may
become smaller or close with time, particularly if the initial diameter of the VSD is less than 5 mm (113). In
patients with isolated moderate to large VSDs, congestive heart failure develops over time once an imbalance
exists between systemic and pulmonary vascular resistance and a significant left to right shunt develops.
Neonates with large VSDs will typically develop pulmonary overcirculation, left ventricular volume overload, and
symptoms of congestive heart failure within the first several weeks to months of life as pulmonary vascular
resistance falls. Any associated left ventricular outflow tract obstruction, such as coarctation of the aorta, will
exacerbate this problem. Premature infants with left to right shunting at the ventricular or great artery level may
develop congestive heart failure sooner as a result of decreased musculature in the pulmonary arterioles that
allows for a precipitous drop in pulmonary vascular resistance.
A trial of medical management is usually indicated in infants with moderate to large VSDs. Provision of adequate
nutrition and administration of medications including
P.731
loop diuretics and sometimes digoxin and/or angiotensin converting enzyme inhibitors (ACE inhibitors) can often
allow surgery to be deferred for a few months (114). Supplemental oxygen and hyperventilation should be
avoided, as they will cause pulmonary vascular resistance to fall, and Qp/Qs to rise. Infants who continue to
have failure to thrive despite these measures are referred for surgical repair, or in selected cases, transcatheter
intervention (115).
Atrioventricular Canal Defect

The presence or absence of a common atrioventricular valve oriface and the relative size of the physiologic VSD
characterizes the different types of atrioventricular septal defects, also known as endocardial cushion defects. In
patients with an incomplete atrioventricular canal, an ostium primum ASD exists along with a cleft in the mitral
valve. There is no physiologic VSD present. A transitional atrioventricular canal defect has an ostium primum
ASD, dense attachments of the atrioventricular valve leaflets to the crest of the ventricular septum such that
two functionally separate atrioventricular valves exist, and a small physiologic VSD. A complete atrioventricular
canal defect has a primum ASD, a common atrioventricular valve and a large inlet VSD. Associate anomalies
include a left superior vena cava to the coronary sinus, additional VSDs, and aortic arch obstruction.
Uncommonly, right ventricular outflow tract obstruction is seen, and such patients have features of tetralogy of
Fallot and an atrioventricular canal (116). In most cases the ventricles are balanced, such that both are of
adequate size to support the full cardiac output following surgery. Occasionally one of the ventricles is
hypoplastic (unbalanced), precluding a two ventricular repair. Such patients may be considered for a one and
a half ventricular repair, the single ventricle pathway, or heart transplantation.
Infants with complete atrioventricular canal defects typically develop congestive heart failure within the first few
weeks to months of life. The congestive heart failure will be exacerbated if significant atrioventricular valve
regurgitation or aortic arch hypoplasia is present. There is an increased incidence of Down syndrome (trisomy
21) in patients with endocardial cushion defects, and such patients are predisposed to the early development of
pulmonary vascular obstructive disease. A trial of medical management is indicated in symptomatic infants, as
discussed above for patients with large VSDs. Surgical repair is required within the first three to six months of
life to eliminate the symptoms of congestive heart failure, prevent the development of pulmonary vascular
obstructive disease and minimize the incidence of postoperative pulmonary hypertensive crises. Surgical repair
is not commonly performed in the newborn period as a result of the difficulty with suturing the paper-thin
atrioventricular valve leaflets.
Infants with incomplete or transitional atrioventricular canal defects often follow the clinical course of patients
with isolated ASDs, and these lesions are typically repaired sometime between 6 months and 4 years of age.

The ductus arteriosus normally is functionally closed within the first hours to days of life. PDAs may exist in
isolation or with many other types of congenital heart disease. An isolated, large PDA will cause symptoms of
congestive heart failure once the pulmonary vascular resistance falls, allowing a left to right shunt from the
aorta to the pulmonary artery. Left atrial and left ventricular volume overload are present and pulmonary edema
may occur. Diastolic runoff exists leading to a steal from systemic blood flow, which may contribute to
mesenteric ischemia and the development of NEC. PDAs are more common and cause more symptoms in
premature infants. Small PDAs typically do not cause symptoms but are a long-term risk for bacterial
endocarditis. Symptoms of congestive heart failure may be treated with diuretics, digoxin and the provision of
adequate nutrition. Approximately two-thirds of PDAs in premature infants will close following a course of
indomethacin or ibuprofen (117). Surgical ligation is indicated for PDAs that do not respond to medical therapy
in premature infants, and for PDAs in any symptomatic term infant. Older children with a continuous murmur
but no symptoms should have the PDA closed surgically or in the cardiac catheterization laboratory to minimize
the risk of endocarditis.
An aortopulmonary window is an uncommon, usually large communication between the ascending aorta and
pulmonary artery in the presence of two semilunar valves. This lesion usually exists in isolation, but may be
associated with IAA, VSD, or other congenital heart defects (118,119). Infants with large aortopulmonary
windows will usually develop congestive heart failure at several weeks of life. Prompt surgical repair is indicated
in most cases to eliminate the symptoms of congestive heart failure, prevent the development of pulmonary
vascular obstructive disease and minimize the incidence of postoperative pulmonary hypertension.
Cardiomyopathy and Heart Transplantation

Cardiomyopathy
A multidisciplinary team, lead by the intensivist, a cardiologist with expertise in heart failure and transplantation,
and a congenital heart surgeon, is required to answer several key questions for each infant who presents with a
severe cardiomyopathy: What is the primary cause for the cardiomyopathy? Is it treatable or reversible? What
level of medical or mechanical support is required? Is the infant a candidate for heart transplantation?
Infants may develop a severe cardiomyopathy from a wide variety of causes, including infection, structural heart
disease, chronic arrhythmia, metabolic disorders, and primary myopathies. The broad differential diagnosis for
infantile
P.732
cardiomyopathies is covered in detail in Chapter 33 and will not be repeated here. However, several conditions
that may present with severe cardiomyopathy, are difficult to diagnose and are sometimes erroneously referred
for heart transplantation, yet are readily treatable deserve comment. Infants born with anomalous left coronary
artery from the pulmonary artery (ALCAPA) develop coronary ischemia when the pulmonary artery pressures fall
after birth. These patients typically present in the first few weeks to months of life with severe left ventricular
dysfunction, mitral regurgitation (as a result of papillary muscle ischemia), and pulmonary edema. The diagnosis
may be suspected by the presence of characteristic findings on the electrocardiogram, and confirmed by
echocardiography with careful examination of the coronary anatomy and blood flow pattern (120,121). Following
reimplantation of the left coronary artery in the aorta, the vast majority of infants with ALCAPA recover left
ventricular and mitral valve function (122). Infants with incessant arrhythmias, such as automatic atrial
tachycardia, may present with severe cardiomyopathy. The initial rhythm may be thought to be sinus
tachycardia secondary to the cardiomyopathy, but the correct diagnosis may be made by careful examination of
the P wave morphology on the electrocardiogram. Myocardial function will recover once the arrhythmia is
controlled with medications or ablation. Severe viral myocarditis is another condition that may be difficult to
differentiate from a primary cardiomyopathy. Myocardial biopsy may provide clarity but is not without risk in sick
infants. Many infants with viral myocarditis will recover ventricular function over time with supportive care (123).
Infants with cardiomyopathy may present with congestive heart failure or shock. Many of the principles for
supporting cardiac output and monitoring of patients with severe cardiomyopathy are discussed below in the
postoperative care section, and thus will not be reiterated. Mechanical support, typically with venoarterial
extracorporeal membrane oxygenation (ECMO) in infants, may be used as a bridge to recovery or
transplantation and is discussed further below. Indications for mechanical support in an infant with severe
cardiomyopathy include worsening end-organ function and metabolic acidosis despite maximal medical support.
TABLE 34-6 CONGENITAL HEART DEFECTS THAT MAY WARRANT PRIMARY HEART TRANSPLANTATION

Severe Shone's syndrome
Severe Ebstein's anomaly of the tricuspid valve
Atrioventricular septal defects with unbalanced ventricles
Pulmonary atresia/intact ventricular septum with coronary-cameral fistula and myocardial
dysfunction
Multiple cardiac tumors with myocardial dysfunction
Adopted from Wong PC, Starnes VA. Pediatric heart and lung transplantation. In Chang AC, Hanley FL,
Wernovsky G et al, eds.Pediatric cardiac intensive care. Lippincott Williams and Wilkins, Baltimore,
1998, with
permission.
Heart Transplantation
Heart transplantation is a widely accepted option for many infants with severe congenital heart disease and
irreversible cardiomyopathies. Approximately 100 infant heart transplants are reported annually to the registry
of the International Society for Heart and Lung Transplantation. Seventy-five percent of these transplants are
for congenital heart disease, and the majority of the remainder are for cardiomyopathies (124). Heart
transplantation is considered a primary option in some institutions for infants with complex congenital heart
disease if surgical intervention is believed to result in an unacceptably high mortality rate (Table 34-6).
Transplantation is also an option if uncorrectable anatomical issues and/or irreversible myocardial dysfunction
are present following initial surgical intervention. The donor pool is very limited, and as a result as many as 25%
of infants will die while waiting for a heart to become available. If more institutions adopted a policy of primary
transplantation for the lesions listed in Table 34-6, particularly HLHS, the mortality while waiting for a donor
heart to become available would certainly increase. Recent experience with ABO incompatible transplantation
has shortened the wait times at a few centers (125). Relative contraindications to heart transplantation include
high, fixed pulmonary vascular resistance (>6-8 Woods units per m2), recent or recurrent malignancy, serious
infection, significant systemic disease, other organ system disease (e.g., hepatic, renal, or neurological), and
chromosomal, metabolic or genetic abnormalities with a poor long-term prognosis.
Although the 1-year survival following heart transplantation in infancy is lower than that reported for older
children, the long-term survival appears to be better, perhaps related to immune tolerance or better compliance
with medications (124). The 10-year survival following infant heart transplantation is approximately 60%.
Infants who have a successful heart transplant face potential long-term complications related to
immunosuppression (i.e., renal insufficiency, infection, diabetes mellitus, and malignancy), transplant coronary
artery disease, and rejection (124). In depth conversations about these issues
P.733
with parents are always required before an infant is listed for transplantation.
Comprehensive supportive care is required to maintain patients who are awaiting transplantation in the ICU. An
appropriate level of cardiopulmonary support is required to maintain end-organ function. This often entails the
use of inotropic agents, vasodilators, and diuretics. Mechanical support, usually ECMO in infants, is occasionally
required (126).
Mechanical ventilation may be beneficial to minimize oxygen consumption, and positive end-expiratory pressure
will reduce wall stress and thus provide afterload reduction for a failing systemic ventricle. Close attention must
also be given to the prevention of infection and the provision of adequate nutrition. Anticoagulation may be
indicated for infants with severe myocardial dysfunction to minimize the chance of luminal clot formation.
Management of the infant with HLHS awaiting transplantation deserves comment. Ductal patency is maintained
with a low dose PGE1 infusion, and occasionally these infants can be discharged to home while awaiting
transplantation (127). Infants with HLHS awaiting heart transplantation tend to develop pulmonary
overcirculation as pulmonary vascular resistance falls over time, which further volume overloads the right
ventricle and may exacerbate tricuspid regurgitation. The extended use of subambiant FiO2, achieved using
supplemental inspired nitrogen, to increase pulmonary vascular resistance and limit pulmonary blood flow has
been reported in this patient population (84,127). Limitation of pulmonary blood flow by banding of the
pulmonary artery, either surgically or using innovative transcatheter techniques, has also been reported (128).
POSTOPERATIVE CARE
Intraoperative and Immediate Postoperative Management
Cardiopulmonary Bypass (CPB)
Selected pediatric cardiac surgical procedures may be performed without the use of CPB such as modified
Blalock-Taussig shunt or repair of coarctation of the aorta. However, the majority of cardiac operations
performed require CPB. Thus, physicians caring for these patients must be familiar with the equipment and
techniques used to perform CPB, its sequelae on end-organ function, and several intraoperative strategies used
to minimize the associated morbidities. The primary function of CPB is to temporarily replace the major functions
of the heart and lungs while surgical interventions are performed on these organs. A typical CPB circuit used to
perform these functions includes venous cannula(s) that drain systemic venous blood from the vena cavae or
systemic venous atrium, a reservoir, a heat exchanger, a membrane oxygenator, a roller pump, a filter, and an
arterial cannula to return blood to the aorta. Before initiation of CPB, the circulated blood must be modified to
provide an appropriate composition of oxygen, carbon dioxide, pH, temperature, hematocrit, oncotic pressure,
electrolytes, and glucose (129). The circuit is primed with standardized quantities of crystalloid solution,
albumin, mannitol, sodium bicarbonate, heparin, calcium, and packed red blood cells. The patient is
anticoagulated with heparin for the duration of CPB time, and cooled to a variable extent to minimize metabolic
needs and oxygen consumption. Because hypothermia causes increased viscosity and red cell rigidity,
hemodilution is used during hypothermic cardiopulmonary bypass.
To obtain a motionless heart for intracardiac repairs, the aorta is cross-clamped and a potassium-rich
cardioplegia solution is injected into the proximal ascending aorta. Asystole develops once the cardioplegia
solution perfuses the coronary circulation. The combination of cardioplegia and hypothermia provides myocardial
protection for several hours. Following placement of the aortic cross clamp, blood from aortopulmonary
collaterals will continue to return to the left atrium. To eliminate the left atrial blood return and facilitate certain
complex left heart operations, deep hypothermic circulatory arrest (DHCA) may be used. Deep hypothermia
refers to cooling of the core temperature to 18C to 20C. During circulatory arrest the CPB pump is shut off. In
addition to the absence of blood returning to the left atrium, the perfusion cannula may be removed from the
surgical field, creating optimal conditions for an accurate repair. Due to concerns about neurologic outcome
following DHCA, regional perfusion techniques have recently been designed to minimize or avoid the use of
circulatory arrest (130). The neurologic sequelae of CPB and DHCA are discussed in the section on neurologic
complications later in this chapter. Following rewarming, weaning and separation from CPB, protamine is
administered to reverse the effect of heparin (129). Additional blood components may be required to control
bleeding immediately following CPB.
A number of adverse effects from CPB will impact on the postoperative course. During CPB, formed elements of
the blood are exposed to artificial surfaces and sheer stress. Ischemia reperfusion injury occurs, as does
microembolization of gas bubbles and particulate matter. Hemodilution may cause problems with oxygen
delivery and dilution of clotting factors. These events trigger a complex neurohumoral and inflammatory
response. CPB is associated with the release of endogenous catecholamines, vasopressin, endothelin, and
activation of the renin-angiotensin-aldosterone axis, all of which contribute to elevation of systemic and
pulmonary vascular resistances and fluid retention (131,132,133,134,135,136,137,138,139,140,141,142). A
generalized inflammatory response occurs following CPB, and the complement, coagulation and fibrinolytic
systems are activated (143,144,145,146). Capillary leak also occurs, related to fluid retention, the inflammatory
response, and dilution of plasma proteins. White blood cells and platelets are also activated, leading to additional
release of inflammatory mediators and proteolytic enzymes (147,148,149). Pulmonary leukosequestration
occurs, as does oxygen free radical generation (150,151). Platelet counts
P.734
fall following CPB, and clotting factors are diluted (140,152). Myocardial systolic dysfunction also occurs
following CPB in infants and children (153). The end result of these processes is a clinical picture of end-organ
dysfunction of variable severity. The cardiovascular, pulmonary, hematologic and renal systems tend to be the
most severely affected. Myocardial dysfunction, elevated systemic and pulmonary vascular resistance, abnormal
gas exchange and decreased pulmonary compliance, bleeding and fluid retention all may be present in the early
postoperative period.
Several pharmacologic agents and management strategies may be employed in the operating room to minimize
these adverse effects of CPB. Mannitol is administered to the priming solution to induce an osmotic diuresis and
act as an antioxidant (154). Dexamethasone, when administered prior to CPB, has been shown to reduce the
inflammatory response to CPB, and is associated with several favorable clinical effects including a reduced need
for supplemental fluids, a lower alveolar-arterial oxygen gradient, and decreased duration of mechanical
ventilation (155,156). Further studies are needed to determine the optimal timing and dosing of dexamethasone
prior to CPB. Aprotinin is a serine protease inhibitor that is used at many centers to attenuate the inflammatory
and hemostatic activation related to CPB (157). In a series of pediatric studies, this medication, when used in a
high dose regimen, has been demonstrated to safely decrease postoperative bleeding and operative closure
time, and improve respiratory function, primarily in infants and in those patients undergoing a repeat
sternotomy (158,159,160,161,162,163,164,165). Beneficial effects on the systemic inflammatory response were
seen when methylprednisolone was used in addition to aprotinin in adults, but this combination of drugs has not
been specifically studied in infants (166). New pharmacologic strategies to further blunt the inflammatory
response to CPB are being investigated (167,168).
To aid in the removal of edema and hemoconcentrate the infant's blood, ultrafiltration is typically used during
rewarming on CPB. Some surgeons use a technique called modified ultrafiltration (MUF) immediately following
CPB, which may be a more efficient method to filter and hemoconcentrate the patient's blood. Following
pediatric CPB, MUF has been shown to have favorable effects on hemodynamics, blood product requirements,
and total body water balance (169,170,171). In a study of pediatric patients at risk for postoperative pulmonary
hypertension, those receiving MUF had a lower pulmonary to systemic pressure ratio following CPB when
compared to the group that had conventional ultrafiltration (171). Modified ultrafiltration has also been
demonstrated to improve intrinsic left ventricular systolic function, improve diastolic compliance, increase blood
pressure, and decrease inotropic medication use in the early postoperative period in children (172,173). Some
investigators have found that the use of MUF is associated with a decrease in mechanical ventilation time
following pediatric CPB (174).
Stabilization in the Intensive Care Unit Following Surgery

Ideally, the physicians caring for infants following cardiac surgery will attend preoperative multidisciplinary
conferences in which data are presented and discussed to gain familiarity with each patient's anatomic and
pathophysiologic issues and the surgical plan. When an infant is returned from the operating room following
cardiac surgery, the intensivist must obtain a detailed report from the anesthesiologist and the surgeon. A
description must be obtained of the anesthetic regimen, the operative findings and surgical procedure
performed, and CPB, aortic cross-clamp and circulatory arrest times. Prolonged CPB times are a risk factor for
significant postoperative morbidity and mortality in infants and children (15,175). Circulatory arrest times more
than 45 to 50 minutes may be associated with increased postoperative neurologic complications (176). A
transesophageal echocardiogram may be performed before and following CPB for complex open heart cases in
infants weighing >3 kg, and postoperative findings pertaining to myocardial function and residual lesions should
be conveyed to the intensive care unit staff (177,178).
Invasive hemodynamic monitoring is required in nearly all infants following cardiac surgery. Some centers rely
heavily on data obtained from invasive hemodynamic monitoring to evaluate myocardial function and detect
residual lesions. Other institutions use a combination of minimal invasive monitoring and liberal use of
transesophageal echocardiogram to achieve similar information. In any case, it is essential that the clinicians
caring for infants following cardiac surgery are familiar with the interpretation and limitations of data obtained
from invasive monitoring, and the complications associated with these catheters and pacing wires. A variety of
factors may contribute to erroneous data obtained from invasive monitoring, including inappropriate transducer
height, and bubbles or clots in the catheters. Information obtained from invasive monitoring cannot be used in
isolation, but when placed in the context of the overall clinical picture, can be very useful to guide management
in the early postoperative period.
Central venous access is required following CPB in infants, and the anesthesiologist or surgeon will chose a site
depending upon the patient's anatomy, anticipated postoperative course and personal preference. For example,
physicians may chose to minimize the placement of central venous lines in the subclavian and jugular veins in
patients with single ventricle physiology as a result of concerns about thrombosis. Intracardiac lines may be
inserted by the surgeon prior to chest closure through the right atrial appendage to the right atrium (RA line),
or through the right upper pulmonary vein or left atrial appendage to the left atrium (LA line). An OA line
refers to an intracardiac line placed in the common atrium in patients with single ventricle physiology. A
pulmonary artery catheter may be placed through the right ventricular
P.735
outflow tract in patients at high risk for postoperative pulmonary hypertension, residual VSDs, or residual right
ventricular outflow tract obstruction. Continuous monitoring of pulmonary artery pressure provides precise
knowledge of the severity of pulmonary hypertension, and immediate feedback regarding the effectiveness of
interventions to lower pulmonary artery pressure. Significant lability in pulmonary artery pressures during
suctioning of the endotracheal tube or awakening from sedation may be a sign that the patient is not ready to
be weaned. Measurement of a step-up in oxygen saturation from a superior vena cava or right atrial catheter to
a pulmonary artery catheter may be helpful for the detection of significant residual left to right shunting (179). A
pullback pressure tracing from the pulmonary artery to the right ventricle may be obtained at the time of
removal of the pulmonary artery catheter, which quantifies any residual gradient across the right ventricular
outflow tract (179). Some pulmonary artery catheters also have a thermistor tip, thus allowing cardiac output to
be calculated by the thermodilution technique.
The report to the intensivist should include information about right and left heart filling pressures, and
pulmonary artery pressures (if available) measured before the patient left the operating room. Proper
interpretation of these loading conditions is beneficial for the detection of residual lesions, titration of volume
administration, and the implementation of interventions that modify vascular tone. Appropriate interpretation of
the atrial waveforms may provide insight into the presence of significant atrioventricular valve regurgitation or
rhythm disturbances.
Complications associated with intracardiac lines are uncommon but include air embolus, thrombus, infection,
bleeding, and catheter retention. Care must be taken when using LA lines in patients with two ventricular
repairs, and with any intravenous or intracardiac line in those patients with single ventricle physiology, not to
inject air into the systemic circulation. Thus LA lines are only used for monitoring whenever possible, and
medication infusions and parenteral nutrition are typically administered through lines that are in the right heart.
Although the reported risk of thrombosis or infection is very low with intracardiac catheters, LA and pulmonary
artery catheters are removed as soon as possible following surgery, often on the first postoperative day or at the
time of chest closure (180,181). Complications at the time of catheter removal include retention and bleeding;
the latter has been shown to occur more commonly with pulmonary artery catheters (180,181). Institutional
protocols designed to minimize the chance of bleeding at the time of removal of intracardiac lines include
guidelines for acceptable coagulation times and platelet counts. Packed red blood cells should be readily
available, particularly during the removal of pulmonary artery lines.
Arterial access, necessary for continuous blood pressure monitoring and frequent arterial blood gas sampling, is
obtained prior to CPB in the radial, umbilical, femoral, dorsalis pedis or posterior tibial arteries (182,183,184).
Care should be taken to ensure that blood pressure measurements are accurate. If a radial or pedal artery is
used for continuous blood pressure monitoring, one must take into account the effect of pulse amplification
(higher systolic and lower diastolic blood pressure measured peripherally when compared with central
recordings) when interpreting blood pressure values. Dampened waveforms or pressures measured distal to
stenotic arteries may give the false impression of hypotension. For example, arm blood pressure measurements
in a patient who has, or had in the past, an ipsilateral Blalock-Taussig shunt may be diminished as a result of
arterial stenosis. Proper interpretation of the waveform and pulse pressure is important during postoperative
management. For example, significant diastolic runoff may produce a wide pulse pressure if an oversized
systemic to pulmonary shunt is used, or if severe aortic regurgitation is present. A narrow pulse pressure, along
with tachycardia and hypotension, suggests cardiac tamponade. Patients who have had a repair of coarctation or
aortic arch reconstruction should have four extremity blood pressure measurements taken to document the
absence of a residual gradient.
Accurate assessment of the patient's heart rhythm is an important part of the initial evaluation following
surgery. The heart rate and rhythm should be continuously monitored. An electrocardiogram should be obtained
in the immediate postoperative period to serve as a new baseline should the infant subsequently develop a
tachyarrhythmia or myocardial ischemia. Atrioventricular synchrony is important for optimizing cardiac output.
Temporary pacing wires are commonly placed on the surface of the atrial and/or ventricular myocardium before
chest closure, and are tunneled from the mediastinum through the upper abdominal wall where they are secured
with tape. These pacing wires may be interrogated when the underlying rhythm is not apparent using the
bedside monitor or a 12-lead electrocardiogram (see section on arrhythmias). These wires are quite safe and are
usually removed prior to patient discharge (185).
Temperature should be monitored and regulated closely, as high temperature will place increased metabolic
demands on the circulatory system, and low temperature will increase systemic vascular resistance (186). Some
centers measure central and peripheral temperatures as an indirect estimate of systemic cardiac output.
However, a recent investigation found a poor correlation between the core-peripheral temperature gap and
lactate levels or hemodynamic parameters, such as cardiac index derived from a pulmonary artery catheter
(187).
A directed physical examination should be performed to assess the cardiopulmonary status and adequacy of the
surgical repair. Palpation of the precordium should be performed to assess the location and quality of the
ventricular impulses. Any murmurs, rubs or gallops should be noted, although dressings and chest tubes may
limit the auscultatory findings. It is common to hear a friction rub in the first few days following cardiac surgery,
usually as a result of a small accumulation of fluid in the pericardium. The liver span should be noted. Adequate
chest rise should be
P.736
found upon inspiration, and breath sounds should be noted bilaterally. The quality and symmetry of peripheral
pulses and warmth of the extremities are useful means of assessing the adequacy of the systemic circulation.
Caution must be used when attempting to estimate the adequacy of cardiac output by assessing capillary refill in
the postoperative period, as capillary refill time has a poor correlation with cardiac index, systemic vascular
resistance index and lactate levels in this setting (187).
Chest tubes should be assessed for location and proper function. Some surgeons will place only a mediastinal
chest tube, whereas others will also place pleural tubes. Generally the tubes in infants are removed when
drainage falls to less than 20 to 30 cc/day and when there is no evidence for chylothorax or air leak. The
stomach should be routinely decompressed postoperatively with a nasogastric tube in all tracheally intubated
patients. A CXR should be obtained upon admission to the ICU and daily thereafter, as this practice in infants
results in a high percentage of films with an abnormality requiring intervention (188). Particular attention should
be given to the location of all tubes and lines, and the heart size and lung fields.
The surgeon will occasionally leave the chest open for a few days, with the skin closed either primarily or using
a silastic patch, until hemodynamic stability can be achieved, bleeding controlled, and myocardial edema can
decrease (189). The sternum is routinely left open in many centers following the Norwood operation
(4,190,191). The risk for mediastinitis may be increased when the chest is left open (190). When the chest is
eventually closed respiratory compliance may decrease necessitating an increase in ventilatory support
(190,192).
Cardiopulmonary interactions play an important role in the physiology of an infant following cardiac surgery
(193). Ventilatory manipulations in the early postoperative period of PaCO2, PaO2, pH and mean airway pressure
may be used in the context of the patient's physiology to improve hemodynamics. Mechanical ventilation is also
useful to minimize oxygen consumption in infants with limited cardiac reserve. Arterial oxygen saturation is
monitored continuously by pulse oximetry. An arterial blood gas analysis is obtained frequently and attention
should be given to ensure adequate oxygenation and ventilation for the individual patient's physiology.
Ventilator settings used in the operating room are often quite different to those needed in the ICU, primarily as
a result of the different ventilators and respiratory tubing used in these settings. Respiratory acidosis has been
demonstrated to increase pulmonary vascular resistance in infants following CPB and in most cases should be
avoided (193,194). Pulmonary overdistension increases pulmonary vascular resistance and decreases cardiac
output (195). A low functional residual capacity following the discontinuation of mechanical ventilation has been
associated with increased pulmonary vascular resistance following congenital heart surgery (193). Thus
ventilator settings are ordered to maintain relatively normal lung volumes in the early postoperative period.
Although early extubation policies have been reported for older infants and children, most neonates and young
infants require at least 12 to 24 hours of mechanical ventilation following congenital heart surgery (196).
Criteria for extubation following cardiac surgery in infants include the presence of adequate cardiac output,
appropriate mental status to maintain the airway, muscular strength to support respiratory pump function,
acceptable gas exchange, and the absence of significant arrhythmias, bleeding or fever.
Standard laboratory values need to be assessed in the early postoperative period. Electrolytes, including
magnesium and ionized calcium levels, are monitored and normalized. The complete blood count is obtained
daily, and hemoglobin levels are monitored more frequently. In general, a minimum hemoglobin level of 10 to
12 gm/dL is appropriate for infants following a two ventricular repair, and a hemoglobin level of 13 to 15 gm/dL
is appropriate for infants following a palliative operation with ongoing cyanosis (See Color Plate). Anemia may
place unnecessary workload on the myocardium in the postoperative period. Transfusion of erythrocytes will
improve oxygen delivery following pediatric cardiac surgery (197). An excessively high hemoglobin level may
increase ventricular afterload or predispose to shunt thrombosis by increasing viscosity (198). Measurement of
prothrombin and partial thromboplastin times are obtained initially, and repeated as clinically indicated.
In addition to the physical examination, several clinical parameters may be used to assess the adequacy of
cardiac output and oxygen delivery in the immediate postoperative period. The presence of a metabolic acidosis,
as quantified by a base deficit or lactate level, suggests inadequate systemic cardiac output and requires
investigation. Lactic acidosis develops when inadequate tissue oxygen delivery leads to anaerobic metabolism.
Following congenital heart surgery, elevated lactate levels in infants and children upon admission to the ICU are
associated with increased morbidity and mortality (175,199-201). Some centers routinely obtain venous blood
gases from the superior vena cava to estimate cardiac output. An oxygen extraction ratio (SaO2-SvO2/SaO2) of
more than 0.5 at 6 hours after infant cardiac surgery suggests a severe derangement in oxygen transport that
may be predictive of mortality (Fig. 34-13) (202). Urine output provides a good estimate of the systemic cardiac
output and ideally should be at least 1 cc/kg/hour. Rising blood urea nitrogen (BUN) and creatinine levels (Cr)
may be secondary to low cardiac output. Infants often require some inotropic support following CPB, and low
dose dopamine is the initial drug of choice at many centers. Additional inotropic support is discussed further in
the section below on low cardiac output.
Infants may develop significant fluid retention following CPB, which may impair myocardial, respiratory, and
gastrointestinal function. Strategies used to minimize this problem in the operating room, including the use of
steroids and ultrafiltration, are discussed earlier. Despite the presence of total body fluid overload,
P.737
intravascular volume depletion is common in the first few hours following surgery and several fluid boluses may
be required. Furosemide is typically initiated 12 to 24 hours after surgery, either as bolus doses or continuous
infusion (203). Additional diuretics with alternative mechanisms of action are used as clinically indicated. Low
dose (renal dose) dopamine is commonly used to augment urine output despite a lack of convincing data
supporting its effectiveness in infants. Electrolyte disturbances, particularly hypokalemia, hyponatremia, and a
hypochloremic metabolic alkalosis, are commonly encountered as diuresis occurs in the first few days following
CPB.
Figure 34-13 Changes in mean oxygen extraction ratio (OER) for survivors and nonsurvivors after infant
congenital heart surgery over time (circles = survivors, squares = nonsurvivors). Reprinted from Rossi AF,
Seiden HS, Gross RP, et al. Oxygen transport in critically ill infants after congenital heart operations. Ann
Thorac Surg 1999;67:739-744, with permission from the Society of Thoracic Surgeons.
Analgesia is required for all infants following cardiopulmonary bypass. High dose fentanyl is well tolerated and
blunts the stress response in neonates following CPB (204,205,206). Morphine is another narcotic commonly
used in the early postoperative period. Benzodiazepines may be given for amnesia and sedation. Either
intermittent boluses or continuous infusions of these agents may be used in infants depending on the
hemodynamics and perceived risk for pulmonary hypertension. The indications for the use of neuromuscular
blocking agents vary at different centers, but these medications may be used to eliminate coughing and
minimize oxygen consumption in patients with labile hemodynamics.
Gastrointestinal tract motility is decreased following cardiac surgery in infants as a result of the inflammatory
effects of CPB, anesthesia, fluid retention, postoperative use of narcotics and possibly low cardiac output.
Furthermore, some neonates may have been exposed to a preoperative period of mesenteric hypoperfusion. If
these considerations preclude the initiation of enteral nutrition following surgery, the parenteral nutrition is
prescribed. Histamine-2 receptor antagonists are typically administered to minimize the risk of upper
gastrointestinal bleeding until enteral nutrition is established.
Common Postoperative Complications

Low Cardiac Output
The product of heart rate and stroke volume determines cardiac output. Stroke volume is determined by
preload, afterload and contractility. Low cardiac output may thus be caused by abnormalities in one or more of
these variables. An alteration in the heart rate or rhythm may impact cardiac output, and this is discussed
further in the subsequent section on arrhythmias. Decreased preload may occur from hemorrhage, excessive
diuresis, insufficient fluid replacement, or cardiac tamponade. Increased afterload may result from pulmonary
hypertension, peripheral vasoconstriction or ventricular outflow obstruction. A decrease in myocardial
contractility may result from a combination of factors including acidosis, electrolyte imbalance, hypothermia, or
myocardial injury secondary to inflammation, a ventriculotomy, inadequate myocardial protection or ischemia
reperfusion. Low cardiac output, when defined as a cardiac index less than 2.0 L/min/m2, occurs following
approximately 25% of complex two ventricle repairs in infancy (Fig. 34-14) (140). Other signs of low cardiac
output include tachycardia, hypotension, poor peripheral perfusion, poor urine output and metabolic acidosis. A
detailed assessment must be undertaken to identify the cause of low cardiac output so that proper treatment
may be initiated and end-organ failure avoided (207).
A suboptimal surgical repair with significant residual intracardiac shunts, ventricular outflow or aortic
obstruction,
P.738
systemic or pulmonary venous stenosis, and/or valvular lesions may be the primary etiology for low cardiac
output. The incidence of hemodynamically significant residual lesions following congenital heart surgery has
likely decreased in recent years as a result of improving surgical technique and the liberal use of
transesophageal echocardiogram immediately following CPB (178). That being said, when an infant develops low
cardiac output or follows an atypical clinical course following CPB, a primary responsibility of the intensivist and/
or cardiologist is the early detection of residual lesions that require reintervention. Failure to do so will expose
the patient to prolonged ICU stays with the associated morbidity and mortality. A combination of factors will lead
to the early identification of significant residual lesions, including an appreciation of the anticipated normal
postoperative course, close attention to data obtained by intracardiac monitoring and physical examination,
open communication with the cardiovascular surgeon, and a high index of suspicion (208). Early, focused
echocardiography or cardiac catheterization is then required to define the extent of the problem (209,210).
Figure 34-14 Scatterplots showing serial measurements of cardiac index as determined by thermodilution
techniques in 122 patients after the arterial switch operation for TGA. Cardiac index fell during the first
postoperative night, returning to baseline values by 24 hours after surgery. There was no significant difference
between patients randomized to deep hypothermic circulatory arrest (DHCA) and those randomized to low-flow
cardiopulmonary bypass (LF). Values are depicted as the mean and one side of each 95% CI. Reprinted from
Wernovsky G, Wypij D, Jonas RA, et al. Postoperative course and hemodynamic profile after the arterial switch
operation in neonates and infants. A comparison of low-flow cardiopulmonary bypass and circulatory arrest.
Circulation 1995;92:2226-2235, with permission.
Treatment of low cardiac output in neonates differs from that in older children and adults as a result of
differences in cardiovascular physiology. Because neonates have a greater ratio of noncontractile to contractile
myocardial mass, ventricular diastolic compliance is diminished, and they have a limited ability to increase their
stroke volume, which is fixed at approximately 1.5 cc/kg (211,212,213). Thus the cardiac output is ratedependent in this patient population. Heart rate may be optimized by pacing, or by using intravenous infusions
of chronotropic agents including dopamine, dobutamine or rarely isoproterenol. For infants with complete heart
block, atrioventricular sequential pacing at an appropriate rate is indicated to increase the cardiac output. Other
postoperative arrhythmias are discussed in detail below.
Optimization of cardiac loading conditions is a key component to the management of the infant with low cardiac
output (214). As described by the Frank-Starling mechanism, augmentation of the end-diastolic volume
increases the number of interactions between actin and myosin molecules, resulting in a larger stroke volume
and thus higher cardiac output. Conversely, hypovolemia may result in decreased ventricular filling and low
cardiac output. Certain cardiac operations in infants, such as repair of tetralogy of Fallot or truncus arteriosus,
result in particularly poor right ventricular compliance, which will require additional preload in the early
postoperative period to maintain cardiac output. Volume replacement therapy should also be guided by close
attention to filling pressures, arterial pressures, and physical examination signs including gallops, liver distention
and peripheral pulses. The type and amount of fluid replacement is based upon the hematocrit, albumin level,
and percentage of volume loss. Boluses of fluid are given in increments of 5 to 10 ml/kg over several minutes. A
left atrial pressure above 14 to 16 mm Hg rarely produces an additional increase in cardiac output, and a left
atrial pressure above 20 mm Hg can cause pulmonary edema (215). Furthermore, as a result of the large
venous capacitance of infants, the right atrial pressures may not necessarily reflect the volume administered and
should not be used in isolation to estimate preload.
Afterload, defined as the sum of forces that oppose systolic performance, is best quantified by systolic wall
stress and vascular impedance, both of which are difficult to measure at the bedside. Afterload may be clinically
estimated by using a pulmonary artery catheter to aid in the calculation of the resistance of the systemic and
pulmonary vascular beds. An increase in vascular resistance can significantly reduce both stroke volume and the
extent and velocity of wall shortening, resulting in decreased cardiac output and ventricular function. Increased
vascular resistance is commonly seen following cardiopulmonary bypass in neonates (140). Physiologic factors
such as hypoxia, acidosis, hypothermia, and pain may further increase systemic and pulmonary vascular
resistance. Increased afterload may also be secondary to residual right or left ventricular outflow tract
obstruction. In the setting of decreased cardiac contractility, increased afterload may be a compensatory
response necessary to maintain systemic blood pressure.
Vascular resistance, and thus afterload, can be pharmacologically decreased by vasodilatation of the vascular bed
(s). Systemic afterload reduction is beneficial for infants with significant aortic or mitral valve regurgitation, left
ventricular dysfunction or hypertension. Afterload reduction in the pulmonary circulation is beneficial for infants
with tricuspid valve regurgitation, right ventricular dysfunction and pulmonary hypertension. In addition to its
inotropic properties, milrinone is a direct vasodilator of the systemic and pulmonary vascular beds (216). The
systemic and pulmonary vascular resistances are also effectively lowered by intravenous infusion of sodium
nitroprusside or nitroglycerine, which are both NO donors that cause smooth muscle relaxation. With the use of
any of these vasodilators, volume augmentation may be necessary to fill the expanded vascular space and
maintain adequate preload. Inhaled NO selectively vasodilates the pulmonary vascular bed, thereby decreasing
right ventricular afterload and allowing for recovery of right ventricular systolic function (217). Positive pressure
ventilation may reduce wall stress and left ventricular afterload in two ventricle hearts provided that preload is
maintained (218,219). Occasionally an infant may have vasodilatory shock with inappropriately low systemic
vascular resistance following CPB that is refractory to catecholamines. Favorable hemodynamic responses to
vasopressin infusions have been described in this setting (220).
Cardiac contractility is the load-independent ability of the myocardium to generate force. Contractility may be
chronically impaired preoperatively by pressure and/or volume overload related to the specific cardiac lesion.
Contractility may be depressed intraoperatively by medications, anesthesia, myocardial ischemia, an extensive
ventriculotomy, or by myocardial resection. Postoperatively,
P.739
hypoxia, acidosis, and pharmacologic agents may affect contractility. In infants and children the cardiac index
tends to reach its nadir approximately 4 hours after CPB, and begins to rise to normal values after 9 to 12 hours
(221). The fall in cardiac index and rise in systemic and pulmonary vascular resistance occurring between 6 to
18 hours following the arterial switch operation in neonates with TGA has been well described (140). If the
infant still shows evidence of low cardiac output after optimizing the heart rate, the preload, and afterload,
myocardial contractility should be enhanced with pharmacologic agents. Several inotropic drugs are currently
available, and each has its own characteristic effects that may be more suitable for use in various clinical
situations.
Dopamine activates dopaminergic, , and -receptors, depending on the dosage used. When prescribed at 1 to 5
mcg/kg/min, dopamine preferentially dilates mesenteric and renal vessels, and increases renal blood flow.
Dosing at 5 to 10 mcg/kg/min tends to increase cardiac output with a mild increase in heart rate (222). Higher
doses of dopamine (10-20 mcg/kg/min) are usually avoided because of an increase in pulmonary vascular
resistance related to alpha-receptor stimulation. Low-dose dopamine is the initial drug of choice to increase
inotrope at many centers.
Dobutamine, a synthetic catecholamine, acts primarily on myocardial beta-receptors. Contractility increases with
infusion of dobutamine, but there is less effect on heart rate or vascular tone than with dopamine (223,224,225).
Milrinone is a phosphodiesterase inhibitor that exerts a positive inotropic effect by increasing intracellular levels
of cyclic adenosine monophosphate (cAMP), leading to improving cardiac contractility. It also has lusitropic
properties, and acts as a direct vasodilator of the systemic and pulmonary vascular beds (216). Milrinone has
been shown to decrease the incidence of low cardiac output syndrome in a prospective, randomized clinical trial
of pediatric patients less than 6 years of age undergoing biventricular repair, and to improve hemodynamics in
neonates with existing low cardiac output (226,227).
Patients who have marked myocardial dysfunction that does not improve with one or a combination of the first
line agents listed above may respond to intravenous epinephrine at 0.01 to 0.05 mcg/kg/min. When
administered in this dosage range, epinephrine primarily activates -1 cardiac receptors causing increased
inotrope, and -2 peripheral receptors causing reduced afterload. High-dose epinephrine (>0.1 mcg/kg/min) is
not frequently used because of marked -adrenergic action and adverse effect on renal perfusion. High-dose
epinephrine has also been shown to cause myocardial necrosis in neonatal pigs (228).
Infants with refractory low cardiac output, myocardial dysfunction and impending cardiovascular collapse may
benefit from having their sternotomy incision reopened in the ICU. The combination of edema of the
myocardium and other mediastinal structures, and any fluid or blood collections around the heart may contribute
to poor ventricular filling and compliance. Reopening the chest will expand the mediastinal space until edema
improves, and any fluid collections or blood clots can be easily removed. Should low cardiac output persist, an
open chest allows easy access for cannulation for mechanical support.
In patients with excessive or prolonged inotropic or vasopressor needs following CPB, a state of relative adrenocortical insufficiency may exist. Myocardial -receptor down regulation has been demonstrated to occur in
infants and small children with congenital heart disease in the perioperative period (229). Corticosteroids,
probably acting by improving vascular tone and up regulating adrenergic receptors, may allow weaning of high
dose catecholamine infusions in infants (230).
Thyroid hormone plays a critical role in the regulation of the cardiovascular system, and decreased levels of triiodothyronine (T3) and thyroxine levels occur in some infants following CPB (231,232,233). Preliminary data
suggest that exogenous supplementation with T3 is well tolerated and may improve mixed venous oxygen
saturation and decrease inotropic requirements following neonatal CPB (234). Favorable changes in
hemodynamics have also been reported in older infants and children who were given T3 following CPB
(235,236). Patient enrollment has been completed for a prospective, randomized, double-blind, placebocontrolled clinical trail designed to evaluate the effects of a 72-hour infusion of T3 on early postoperative
outcomes in neonates undergoing the Norwood procedure or repair of IAA at Children's Hospital Boston, and the
results will be available in early 2005.
Arrhythmias
The development of a postoperative arrhythmia may severely compromise cardiac output in the infant following
CPB. Prolonged CPB and aortic cross clamp times, and higher postoperative serum troponin levels, are
associated with the development of arrhythmias in infants and children (237). An accurate diagnosis of the
arrhythmia may be obtained with a 12-lead electrocardiogram. However, it is often the case in infants with fast
heart rates that the relationship of the P waves to the QRS complex is uncertain on the standard 12-lead ECG,
and in these instances, interrogation of temporary atrial pacing wires is very helpful (185,238). Several
techniques may be used to interrogate the temporary atrial pacing wires. One such method is to connect the two
leg leads of the ECG machine to the temporary atrial pacing wires, and the two arm leads are placed on the
patient's arms in the usual fashion. A rhythm strip of leads I, II and III is then recorded from the ECG machine.
Lead I will be a surface electrogram, and a sharp atrial electrogram (indicating atrial depolarization) is produced
in leads II and III, which may be compared to the surface QRS complex in lead I to determine the relationship of
P waves to the QRS complexes. The presence of cannon a-waves may give insight into the presence of an
arrhythmia and its mechanism. Cannon a-waves, depicted on the bedside monitor as an increase in amplitude of
the atrial pressure waveform obtained by intraatrial lines,
P.740
occur when the atria contract against a closed atrioventricular valve during various arrhythmias. Nearly any type
of arrhythmia may occur following infant CPB. Commonly encountered arrhythmias are discussed in the
following paragraphs, and further information may be found in an excellent book on pediatric arrhythmias (239).
Premature atrial complexes (PACs) are occasionally seen and may be related to central lines, electrolyte
disturbances or surgical incisions; they are usually benign. Reentrant forms of supraventricular tachycardia
(SVT), such as atrial flutter and atrioventricular reentrant tachycardia, are relatively uncommon following infant
cardiac surgery. Although rate-related aberrancy and antegrade conduction over accessory connections may
occur, in most cases of SVT the QRS complex in tachycardia is similar in morphology and axis to that seen on
the baseline postoperative ECG. The baseline QRS complex may be wide if a bundle branch block developed
during surgery, as is common following repair of tetralogy of Fallot, VSDs, and complete atrioventricular canal
defects. Atrial flutter is more likely to be seen following complex atrial baffling procedures. Atrial flutter is
characterized by a rapid, regular atrial rate with variable atrioventricular conduction. Adenosine is helpful
diagnostically as the flutter waves will persist in the presence of atrioventricular block. Atrial flutter may be
terminated using rapid atrial pacing, via the temporary pacing wires or a transesophageal electrode, or
synchronized cardioversion starting at 0.5 to 1 J/kg. Atrioventricular reentrant tachycardia may be seen in
infants with accessory connections, such as some patients with Ebstein's anomaly, L-looped ventricular
inversion. Pre-excitation (Wolfe-Parkinson-White Syndrome) may be present on the electrocardiogram in sinus
rhythm. Atrioventricular reciprocating tachycardia is characterized by retrograde P waves following the QRS
complex in a one-to-one relationship, and is terminated with adenosine, rapid atrial pacing, or synchronized
cardioversion. Atrioventricular node reentrant tachycardia and atrial fibrillation are rare in infants and will not be
discussed further.
Ectopic atrial tachycardia is an uncommon form of SVT following congenital heart surgery. It is characterized by
an automatic atrial rhythm with warm up and cool down behavior at its onset and termination. Infants
having disruption of the atrial septum (e.g., tricuspid atresia or TGA following balloon atrial septostomy), longer
CPB times, higher inotropic needs, and potassium depletion may be at increased risk for developing this
arrhythmia (240). Treatment strategies include normalization of electrolytes and temperature, minimizing
inotropic infusions, and administration of a variety of antiarrhythmic agents including beta-blockers and
amiodarone.
Junctional ectopic tachycardia (JET) is a common type of SVT seen in the postoperative period in infants,
particularly following repair of tetralogy of Fallot or VSDs (241,242). JET is an automatic rhythm originating from
the bundle of His, and although thought to be caused by some form of trauma to the atrioventricular (AV) node
during surgery, may occasionally be seen following cardiac surgical interventions distant from the AV node. The
occurrence of postoperative JET has been reported to prolong postoperative mechanical ventilation and ICU
times (243). Electrophysiological characteristics of JET are as follows: a QRS morphology similar to that seen in
sinus rhythm; atrio-ventricular dissociation with the ventricular rate faster than the atrial rate, or 1:1 retrograde
conduction; warm up behavior as seen with automatic arrhythmias; and failure to respond to adenosine,
overdrive pacing or cardioversion (244). Cannon a-waves will have variable amplitude in patients with JET if
ventricular-atrial (V-A) dissociation is present, but constant increased amplitude will be seen with JET if the
junctional rhythm is conduced to the atria in a 1:1 retrograde pattern. A 12-lead ECG or interrogation of
temporary atrial pacing wires will confirm the diagnosis. Although JET usually resolves spontaneously in the first
few days following surgery, this arrhythmia will often cause hemodynamic compromise when the ventricular rate
exceeds 170 to 180 beats per minute. Treatment strategies must be individualized to the individual patient's
heart rate and hemodynamic status and include fever control, provision of adequate analgesia to minimize
endogenous catecholamine release, minimizing the use of exogenous catecholamines, normalization of
electrolytes and acid-base status, atrial pacing at a rate faster than the junctional rate, and mild hypothermia
(245,246). Medications used to treat JET when the above measures fail include amiodarone, esmolol and
procainamide (244, 246,247). As these drugs have negative inotropic and chronotropic properties, close
monitoring of hemodynamics and back-up pacing capabilities are necessary.
Premature ventricular contractions (PVCs) may reflect myocardial irritability, electrolyte disturbances or hypoxia.
Lidocaine is often administered for frequent PVCs or nonsustained ventricular tachycardia (VT). VT is
characterized by a fast, wide QRS complex that differs in morphology and axis compared to the postoperative
baseline, and has either 1:1 retrograde V-A conduction, or V-A dissociation. The presence of complex ventricular
ectopy or VT is suggestive of myocardial ischemia, and those infants who had coronary manipulation as a
component of their operation should have a prompt evaluation of coronary blood flow. Sustained VT with
hemodynamic compromise may be terminated by synchronized cardioversion starting at 1 J/kg. Pharmacologic
therapy, starting with either lidocaine or procainamide, may be considered for patients with hemodynamically
stable VT. Ventricular fibrillation (VF) is characterized by a wide complex, irregular rhythm that requires
immediate defibrillation, starting at 2 J/kg. Cardiopulmonary resuscitation is required during VF until a perfusing
rhythm is reestablished.
Complete heart block typically is apparent when the patient is rewarmed following CPB, but much less commonly
may develop in the first few days following surgery. Atrioventricular sequential temporary pacing is used for
treatment (248). The capture threshold of the temporary ventricular pacing wire should be determined
frequently in patients who have developed or are at high risk for complete heart block. An alternative method of
pacing
P.741
may be required for patients with very high or rapidly rising capture thresholds. If atrioventricular conduction
does not return within 7 to 14 days, a permanent pacemaker is placed to prevent low cardiac output and sudden
death (52).
Pulmonary Hypertension
Although the movement toward early complete neonatal repair has lead to a decreased incidence of pulmonary
hypertensive crises for patients with many congenital heart defects, this complication continues to impede the
recovery of infants following complete surgical repair or palliation in the contemporary era (3,8,142,249,250).
Severe pulmonary hypertension, defined as a ratio of pulmonary to systemic arterial pressure greater than or
equal to 1.0, was present postoperatively in 2% of infants and children having congenital heart surgery at a
single center (median age 4.2 months) (251). In this series, severe pulmonary hypertension was associated with
an increase in mortality.
Pulmonary hypertension following infant CPB may be caused by a combination of preoperative, intraoperative,
and postoperative factors, the sum effect of which may be additive (Table 34-7). The presence of several
preoperative clinical features allows for the identification of infants at increased postoperative risk for pulmonary
hypertension. Specific lesions, all characterized by having preoperative large left to right shunts or obstructed
pulmonary venous return, are associated with postoperative pulmonary hypertension (8,252). Infants with large
communications at the ventricular or arterial level by definition have pulmonary hypertension prior to going to
the operating room. Infants with obstructed pulmonary venous return, mitral stenosis or left ventricular outflow
obstruction (critical aortic valve stenosis or coarctation of the aorta) may have severe preoperative pulmonary
hypertension and increased pulmonary vascular resistance. Down syndrome is also a risk factor for severe
postoperative pulmonary hypertension (251).
TABLE 34-7 FACTORS CONTRIBUTING TO PERIOPERATIVE PULMONARY HYPERTENSION IN INFANTS
Preoperative
left to right shunts
obstructed pulmonary venous return
Intraoperative
microemboli
pulmonary leukosequestration
excess thromboxane production
duration of CPB
endothelial injury
Postoperative
mechanical obstruction to pulmonary blood flow
residual left to right shunt
atelectasis
hypoxic pulmonary vasoconstriction
catecholamines (endogenous and exogenous)
CPB is associated with increased pulmonary vascular resistance in infants and children (140,141). CPB causes
partial ischemia of the pulmonary vasculature, leading to endothelial cell dysfunction and decreased endogenous
production of NO (141,253-256). CPB leads to increased plasma levels of catecholamines, endothelin-1 and
other pulmonary vasoconstrictors in infants and children (252,257-259). In infants, prolonged CPB time has
been associated with increased pulmonary vascular resistance in the postoperative period (142). The presence
of significant pulmonary hypertension soon after weaning from CPB is predictive of subsequent pulmonary
hypertension in the ICU and the need for prolonged ventilatory support (142,260).
Postoperatively, large residual left to right shunts or obstruction to pulmonary venous or distal pulmonary
arterial blood flow, or left ventricular outflow tract obstruction all may cause pulmonary hypertension. Noxious
stimuli, particularly suctioning of the endotracheal tube, may trigger a pulmonary hypertensive crisis (261).
Pulmonary hypertension may manifest as low cardiac output following a two-ventricular repair when both septa
are completely intact. In infants with a significant risk for pulmonary hypertensive spells, such as those
undergoing complete repair of truncus arteriosus, an atrial communication is often intentionally left open so that
right to left shunting may occur to preserve cardiac output in the event that pulmonary artery pressures become
significantly elevated. In such infants, and those with palliated single ventricle physiology, excessive cyanosis
will be present if postoperative pulmonary hypertension occurs. Pulmonary artery catheters, although not
commonly used in infants at many congenital heart centers, provide direct and continuous measurement of
pulmonary artery pressure following two
P.742
ventricular repairs. The severity of pulmonary hypertension may also be estimated echocardiographically by
Doppler interrogation of a tricuspid regurgitation jet in the absence of right ventricular outflow tract obstruction.
TABLE 34-8 CRITICAL CARE STRATEGIES FOR TREATMENT OF PULMONARY HYPERTENSION
Encourage
1. Anatomic investigation
Avoid
2. Right to left atrial pop off
1. Residual anatomic disease

2. Intact atrial septum
3. Sedation/analgesia
4. Moderate hyperventilation
3. Agitation/pain
4. Respiratory acidosis
5. Moderate alkalosis
5. Metabolic acidosis
6. Adequate inspired oxygen

7. Normal lung volumes
6. Alveolar hypoxia
7. Atelectasis or overdistension
8. Optimal hematocrit
9. Inotropic support
8. Excessive hematocrit
9. Low output and coronary perfusion
10. Pulmonary vasodilators
10. Pulmonary vasoconstrictors
Modified from Wessel DL. Managing low cardiac output syndrome after congenital heart surgery. Crit
Care Med 2001;29:S220-30, with permission.
A combination of relatively simple postoperative strategies should be sufficient to prevent the development of
pulmonary hypertensive crises in many at-risk patients (Table 34-8) (262). Adequate oxygenation and
ventilation are important in the early postoperative period to maintain low pulmonary vascular resistance, as is
the maintenance of relatively normal lung volumes (193-195,263). Deep sedation and analgesia in the
immediate postoperative period are also used to minimize the incidence of pulmonary hypertension crises. High
dose fentanyl infusions have been shown to reduce the stress response and postoperative morbidity in infants
following CPB (257). In particular, fentanyl has been shown to blunt the elevation in pulmonary artery pressure
and pulmonary vascular resistance related to endotracheal suctioning following the repair of congenital heart
defects in infants (205). Benzodiazepines are commonly used for additional sedation. Prophylactic alphareceptor blockade and early definitive repair can reduce the incidence of postoperative pulmonary hypertension
(8). Although nitric oxide is well tolerated and decreases the incidence of pulmonary hypertensive crises in highrisk infants, this drug is not commonly used for primary prevention of pulmonary hypertension crises in part
related to its high cost (250).
Before assuming that increased pulmonary vascular resistance causes pulmonary hypertension, the intensivist
must ensure that an anatomic etiology for elevated pulmonary artery pressures is not present. For example,
significant residual pulmonary arterial obstruction distal to the point of pulmonary artery pressure measurement
will cause pulmonary hypertension, but the pulmonary artery pressure will not fall with medical management
including inhaled NO (208).
All of the management strategies mentioned above regarding the prevention of pulmonary hypertension are
used in the event of a pulmonary hypertensive crisis. Hyperventilation has been shown to reduce pulmonary
artery pressure and pulmonary vascular resistance in infants and children with pulmonary hypertension in the
early postoperative period. Although hyperventilation is a first-line intervention for pulmonary hypertensive
crises, the following associated effects must be considered. Systemic vascular resistance increases with
hyperventilation, as a result of alkalosis, leading to decreased cardiac output (264,265). Depending on the
ventilator strategy used to achieve hyperventilation, the mean airway pressure may increase, which may
increase pulmonary vascular resistance. Other concerning side effects of hyperventilation include reduced
cerebral and coronary blood flow, and a leftward shift of the oxy-hemoglobin dissociation curve.
In an infant with a two-ventricular repair and intact atrial septum who experiences a pulmonary hypertension
crisis and develops refractory hypotension, careful infusion of volume through the left atrial line will bypass the
pulmonary vascular bed and may temporarily stabilize the blood pressure until the pulmonary vascular
resistance is reduced.
Nonselective vasodilators were used in the past to treat severe pulmonary hypertension, however, current
standard therapy centers on the use of NO. NO, when produced by the vascular endothelium, diffuses to
adjacent smooth muscle cells, in which relaxation occurs by activation of guanylate cyclase, which increases
intracellular guanosine 3, 5-monophosphate (cyclic GMP) (266). Because NO is rapidly inactivated by
hemoglobin, it acts as a selective pulmonary vasodilator when delivered by inhalation (267,268). Side effects
associated with NO include a rebound effect upon withdrawal of the drug, and methemoglobinemia
(269,270,271). Preliminary observations suggest that sildenafil may be useful to facilitate the weaning of NO in
infants who previously demonstrated significant rebound pulmonary hypertension (272).
One randomized study of high-risk infants and young children found that NO did not lower pulmonary artery
pressure or prevent pulmonary hypertensive crises. Most
P.743
patients in the study were receiving nitroprusside, a NO donor, and about one-half were receiving milrinone.
Furthermore, this study was not controlled for the use of neuromuscular blockade, and several patients in the
study had underlying lung disease (273). However, the vast majority of studies found a beneficial effect of NO in
alleviating pulmonary hypertension following CPB in infants and children. In a randomized, double-blind, placebocontrolled clinical trial of infants and children (mostly infants with VSDs or atrioventricular canal defects) who
emerged from CPB with a mean pulmonary artery pressure more than 50% of the mean systemic artery
pressure, NO was shown to reduce the mean pulmonary artery pressure in the immediate postoperative period
(260). Several other nonrandomized reports demonstrate the efficacy and safety of NO for selectively lowering
pulmonary artery pressure in infants and young children following CPB (141,274-278). Transient right
ventricular dysfunction following repair of congenital heart defects may be exacerbated by pulmonary
hypertension. NO improves right ventricular ejection fraction and cardiac output while decreasing pulmonary
artery pressure and vascular resistance in infants and young children following a biventricular repair (217). NO
also decreases pulmonary artery pressure in infants with single ventricle physiology following the bi-directional
Glenn operation (278,279). NO may reduce the need for extracorporeal life support in infants with critical
pulmonary hypertension after CPB (280).
The benefits of intravenous prostacycline on pulmonary vascular resistance in children with congenital heart
disease are offset by tachycardia and systemic hypotension (281,282,283). However, more recent studies have
found that inhaled prostacycline, which causes vasodilation by increasing intracellular concentrations of cAMP, is
as effective as NO for lowering pulmonary vascular resistance following repair of congenital heart disease
(284,285). Further clinical investigation and clinical experience are required before inhaled prostacycline can be
recommended for widespread clinical use following CPB in infants.
Cyanosis
Excessive cyanosis following surgical repair or palliation may be attributed to a variety of anatomic problems
depending upon the individual patient's physiology. In the infant with palliated single ventricle physiology,
excessive cyanosis may be attributable to inadequate pulmonary blood low, pulmonary venous desaturation and/
or systemic venous desaturation. Decreased pulmonary blood flow may be as a result of stenosis or thrombosis
in the systemic to pulmonary artery shunt, or stenosis in the superior vena cava or pulmonary arteries following
the bi-directional Glenn or hemi-Fontan operation. If an anatomic problem is suspected with the systemic to
pulmonary artery shunt or a bi-directional Glenn, a cardiac catheterization is usually required. Pulmonary
hypertension may also cause decreased pulmonary blood flow, which can be attributed to CPB, or impairment of
blood egress from the pulmonary veins (e.g., pulmonary venous obstruction or left atrial outlet obstruction). In
palliated single ventricle patients with excessive cyanosis, treatment of pulmonary hypertension includes the use
of NO if standard measures to lower pulmonary vascular resistance are ineffective (278,279). Pulmonary venous
desaturation may be caused by parenchymal lung disease, pleural effusions, or systemic to pulmonary venous
collateral vessels. Systemic venous desaturation, secondary to anemia, low cardiac output or increased oxygen
consumption, may also contribute to cyanosis following single ventricle palliation. Similar causes for excessive
cyanosis may exist following a two-ventricle repair, particularly if an atrial communication exists allowing right to
left shunting. For example, following complete repair of truncus arteriosus, excessive cyanosis related to
decreased pulmonary blood flow may be present in the postoperative period as a result of poor right ventricular
compliance, right ventricular outflow tract obstruction, or pulmonary hypertension with subsequent right to left
shunting at the atrial level. Note that excessive systemic venous desaturation will not contribute to cyanosis
following a two-ventricle repair unless a residual intracardiac communication or veno-venous collaterals exist
that allow right to left shunting.
Ventilatory management of the infant with excessive cyanosis following the bi-directional Glenn/hemi-Fontan
operation deserves special consideration. As pulmonary blood flow is passive and heavily influenced by
intrathoracic pressure, ventilator settings that minimize mean airway pressure and maximize exhalation time will
enhance pulmonary blood flow. As these operations place the cerebral and pulmonary vascular beds in series,
pulmonary blood flow is derived in large part from venous return from the brain to the superior vena cava.
Working on the principle that increased partial pressure of CO2 (PaCO2) results in cerebral arterial vasodilation
and increased cerebral blood flow, investigators have demonstrated that hyperventilation decreases SaO2
following the bi-directional Glenn/hemi-Fontan operation, and the hypoventilation improves SaO2 (286,287). In
addition to CO2 induced increased cerebral blood flow, however, other uncontrolled factors that may have
resulted in improved SaO2 in these patients included a decrease in mean airway pressure during
hypoventilation, and CO2 stimulated increase in cardiac output leading to increased oxygen saturation in the
inferior vena cava (287).
Bleeding
Excessive bleeding may occur from suture lines and/or abnormalities in the coagulation system following CPB
(152). Generally accepted risk factors for postoperative bleeding include repeat sternotomy, cyanosis, and
operations involving extensive suture lines in the aorta. Excessive bleeding may be defined as greater than 5cc/
kg of blood from the chest tube in any given hour or greater than 3 cc/kg/hour 3 hours. Intraoperative
administration of aprotinin may minimize postoperative bleeding, as
P.744
discussed earlier (164). Aminocaproic acid and tranexamic acid, both antifibrinolytic agents, may be
administered intraoperatively to infants at risk for bleeding (288,289,290,291). The platelet count is maintained
more than 50,000 to 100,000/mL in the early postoperative period. Fresh frozen plasma is administered for a
prothrombin time (PT) more than 25 seconds or for excessive bleeding. Hypertension might exacerbate bleeding
and should be controlled. Inadequate neutralization of heparin will be manifested by a prolonged partial
thromboplastin time (PTT), and additional protamine is indicated if this diagnosis is suspected.
Bleeding may occur when intracardiac lines are removed. Bleeding is more common following the removal of
pulmonary artery catheters and left atrial lines, when compared to right atrial lines (180,181). The pulmonary
artery and left atrial lines are generally removed on the first day following surgery if no longer needed clinically,
or at the time of chest closure. Coagulation times and platelet counts should be at acceptable levels, and crossmatched blood should be available when these catheters are removed.
Cardiac tamponade may occur when significant bleeding is not evacuated by the chest tube(s). External
compression of the heart by blood or blood clots leads to impaired ventricular filling, increased central venous
pressure, tachycardia, a narrow pulse pressure and eventually systemic hypotension. The diagnosis can be
confirmed by echocardiography, and surgical exploration of the chest is required, either emergently at the
bedside, or urgently in the operating room.
Cardiac Arrest
Infants having a cardiac arrest following congenital heart surgery have better survival rate (41% in one series)
when compared with other general pediatric intensive care unit patient populations (292). The better survival
rate may be attributed to a variety of issues unique to the cardiac population, including the increased incidence
of an acute ventricular arrhythmia, the absence of multi-organ system failure in the majority of the cardiac
patients, and the common presence of central venous access, arterial access and epicardial pacing wires.
Survival to hospital discharge was achieved in 5/22 infants despite resuscitation times >30 minutes in one series
of patients, making the use of predetermined resuscitation end points somewhat irrelevant (292). However, the
long-term neurologic outcome of these 5 infants was not reported in this study. Algorithms exist for the
management of infants experiencing a cardiopulmonary arrest and will not be repeated here (293). If
spontaneous circulation has not returned despite a few minutes of resuscitation, the chest may be reopened in
infants who have had a recent sternotomy, and open cardiac massage performed. Rapid-deployment
extracorporeal membranous oxygenation (ECMO; also see Chapter 32) has been used to salvage some infants
receiving ongoing cardiopulmonary resuscitation and is discussed further below (294).
Mechanical Cardiac Support

A recently published textbook includes a detailed discussion of pediatric mechanical cardiac support (295). At
most congenital heart centers, postoperative ECMO is used in approximately 3% of infants and children following
CPB (296). ECMO is used to provide mechanical cardiopulmonary support in most infants, as aortic balloon
pumps and ventricular assist devices currently are not widely available for this patient population as a result of
size limitations and other technical issues. Some experience does exist, however, with a centrifugal pump
ventricular assist device in infants, including those with isolated left ventricular failure following repair of
anomalous origin of the left coronary artery from the pulmonary artery, a lesion that is characterized by having
severe but reversible left ventricular systolic dysfunction (297,298).
Following surgical repair or palliation, infants who cannot be weaned from CPB may be converted to an ECMO
circuit before leaving the operating room. Once in the ICU, those infants who develop refractory low cardiac
output, severe cyanosis, arrhythmias, or pulmonary hypertension despite maximal medical therapy, or have an
unexpected cardiac arrest may be candidates for mechanical cardiopulmonary support. Mechanical support may
be used as a bridge to myocardial recovery or cardiac transplantation (126,299,300). Venoarterial ECMO is
required in nearly all of infants requiring mechanical support for primary cardiac failure, in contrast to neonates
with primary respiratory failure who may be supported with venous-venous ECMO. Relative contraindications to
ECMO include multi-organ system failure, an irreversible or inoperable disease process, significant neurologic
impairment, uncontrolled bleeding, extremes of size and weight, and inaccessible vessels during
cardiopulmonary resuscitation (262). The decision to employ ECMO following infant CPB should be made prior to
the development of irreversible end-organ failure, as patients who have been placed on ECMO after extended
efforts at medical management have been shown to do poorly (301,302).
Although some investigators have found single ventricle physiology to be a risk factor for mortality with cardiac
ECMO, relatively good outcomes (50% survival to hospital discharge) were reported in a recent series of patients
requiring ECMO after a Norwood palliation (301,302). Management of the systemic to pulmonary artery shunt in
infants with single ventricle physiology who require ECMO deserves special consideration. As is the case prior to
initiation of mechanical support, the physiology if these patients is such that high flow through the shunt may
lead to coronary ischemia, volume overload of the single ventricle and systemic hypoperfusion, all of which will
impede efforts to wean from ECMO. If the shunt is left patent when ECMO is initiated, the flow on the ECMO
circuit must be increased to compensate. Pulmonary overcirculation may also occur. Partial or complete ligation
of the shunt will eliminate the large left to right shunt, but may lead to pulmonary infarction and shunt
thrombosis (301,303).
P.745
The intensive care management of infants on ECMO has been reviewed elsewhere in great detail and is beyond
the scope if this chapter (304). However, two critical issues deserve comment. In infants with biventricular
physiology on ECMO, it essential that the left atrium and left ventricle are decompressed to reduce myocardial
wall stress and pulmonary edema, thus optimizing the chances for myocardial recovery. Left ventricular
distension may be assessed by echocardiogram, and decompression may be accomplished by increasing the
ECMO flow rate, using low-dose inotropic agents and afterload reduction to improve myocardial ejection, and/or
by decompressing the left atrium by surgically placing a vent. The ventilator strategy used in infants on ECMO
for cardiac failure is also important to optimize the probability of myocardial recovery. Even on full venoarterial
ECMO flow, the coronary circulation is typically perfused by blood ejected from the left ventricle. Thus a primary
goal of ventilator management is to ensure that minimal ventilation-perfusion mismatch occurs in the lungs, so
that pulmonary venous blood, which eventually perfuses the myocardium, is adequately oxygenated (304). This
is in direct contrast with the typical ventilator strategy used in neonates with primary respiratory failure on
ECMO, which focuses on lung rest and avoidance of further lung injury.
Inability to wean from ECMO within 3 to 5 days and the development of renal failure are ominous signs for
pediatric cardiac patients (296,302,303,305-307). A high index of suspicion for residual lesions is necessary for
infants who fail to wean from ECMO in a timely fashion (306). If transthoracic echocardiographic windows are
poor, transesophageal echocardiography may be safely used in infants on ECMO to obtain diagnostic information
(308). Cardiac catheterization may also be performed safely with infants on ECMO to evaluate for the presence
of residual lesions that may be amenable to transcatheter or surgical intervention, or to decompress the left
heart (309).
Rapid-deployment ECMO using a modified circuit and an organized response team may be used to facilitate the
resuscitation of infants in full cardiopulmonary arrest following congenital heart surgery. One single center
report described a series of 11 infants and children following a mean cardiopulmonary resuscitation time of 55
minutes, with 7 patients (64%) surviving to hospital discharge (294). Other centers have reported similar
outcomes for emergent ECMO initiated during a postoperative cardiac arrest (299,303).
Survival to hospital discharge following pediatric cardiac ECMO has been shown to be approximately 40% to
60% in several series (296,299,301,303,307,310,311). Children with viral myocarditis requiring mechanical
circulatory support have a higher survival rate (80%) (312). Although the intermediate-term cardiac follow-up is
favorable for infants who survive postpericardiotomy ECMO, this patient population may have an increased
incidence of significant neurologic deficits when compared with older children who required mechanical support
(313,314). The etiology for these neurological deficits is unclear, but may be related to intraoperative variables
such as the use of circulatory arrest, or the complex ECMO circuit that may cause thromboembolic events and
requires higher levels of anticoagulation when compared to ventricular assist devices used in older patients.
Pulmonary Complications
Pulmonary dysfunction occurs following CPB as a result of a variety of factors, including the diffuse inflammatory
response, increased capillary permeability, fluid overload and microemboli. In an infant animal model of CPB,
postpump pulmonary dysfunction includes an increased pulmonary vascular resistance, alveolar-arterial oxygen
gradient, and decreased pulmonary compliance (256). Following CPB, infants have decreased lung compliance
and increased total lung resistance (315). A variety of strategies designed to counteract these adverse effects of
CPB have been discussed earlier, including the use of MUF, steroids, and strategies to prevent and treat
pulmonary hypertension.
The infant who cannot be weaned from mechanical ventilation or extubated in a timely fashion may have a
residual cardiac lesion and/or a noncardiac etiology as an explanation for the lack of progress. Non-cardiac
reasons for difficult weaning are numerous and include abnormalities in respiratory drive (e.g., central nervous
system injury, sedation), respiratory pump (e.g., critical illness polyneuropathy, phrenic nerve injury), gas
exchange (e.g., parenchymal lung injury) or increased ventilatory load (e.g., overfeeding, increased dead space).
External compression of the central airways occasionally may complicate the pre- or postoperative course in
infants with congenital heart disease. For example, aortic enlargement causing central airway compression may
occur in lesions with increased fetal aortic blood flow such as pulmonary atresia (316). Some neonates with
tetralogy of Fallot and absent pulmonary valve syndrome have significant pulmonary artery dilation that leads to
tracheal and mainstem bronchial compression in the preoperative period, and tracheobronchial malacia may
persist following surgery (317). Tracheal stenosis or malacia may develop postoperatively as a result of airway
mucosal injury related to the endotracheal tube. Upper airway obstruction will make intrathoracic pressure
significantly more negative in the spontaneously breathing patient, and the resultant increased ventricular
afterload may be poorly tolerated in the infant with limited cardiac reserve.
The phrenic nerves are at risk for paresis or transection during many infant cardiac surgeries, and paralysis of a
hemidiaphragm may make it difficult to wean infants from mechanical ventilation (318). The diagnosis may be
confirmed by echocardiogram or fluoroscopy while the infant is spontaneously breathing. Diaphragmatic plication
may be required to facilitate weaning from the ventilator (319,320). The recurrent laryngeal nerve may be
injured during aortic arch surgery or PDA ligation, resulting in unilateral vocal cord paralysis. Postoperatively,
such infants
P.746
may have problems with maintenance of functional residual capacity and aspiration.
Tracheobronchial ischemia, leading to obstruction of the lower airways, has been reported to occur following
unifocalization procedures in infants with pulmonary atresia, ventricular septal defect, and MAPCAs. The
tracheobronchial ischemia is probably related to interruption of peribronchial arterial blood supply during
mobilization of MAPCAs (321).
The development of a postoperative chylothorax is a relatively infrequent complication (approximately 1%)
following congenital heart surgery, but when it occurs it adds morbidity and prolongs hospitalization (322). A
chylothorax may develop following thoracic duct injury, and elevated systemic venous pressures or thrombosis
of the subclavian veins may exacerbate the condition (323,324). The diagnosis is confirmed by finding elevated
triglyceride levels and lymphocyte counts in fluid drained from the pleural cavity (325). Treatment strategies
initiate with chest tube drainage and dietary modification with a medium chain triglyceride (MCT) based formula,
although many weeks may be required for drainage to stop (325). Prolonged drainage may lead to depletion of
plasma proteins (albumin, coagulation factors, immunoglobulins) and lymphocytes, and consideration should be
given to replacement with albumin, fresh frozen plasma, and/or intravenous immunoglobulins (IVIG) (323,326).
A trial of total parenteral nutrition is warranted for infants not responding to a MCT formula (323). A small
randomized animal model and several human case reports and series suggest that chylous drainage may be
decreased using medications that decrease splanchnic blood and lymph flow, including somatostatin or
octreotide (a synthetic somatostatin analogue) (324,327-331). Surgical options for patients with refractory
chylothoraces include thoracic duct ligation, pleuroperitoneal shunt or pleurodesis (325,332,333).
Gastrointestinal Complications
Feeding problems are common in neonates following congenital heart surgery. Many of these patients never had
the opportunity to learn to suck and swallow before surgery, and had prolonged periods of oral-tracheal
intubation. Nasogastric feedings may be initially required, and input from a speech therapist may be useful.
NEC may occur either before or after congenital heart surgery, particularly in neonates with lesions that produce
diastolic runoff, such as single ventricle patients with a systemic to pulmonary artery shunt (50). Because of
bowel dysfunction related to anesthesia and narcotics, and the concern for NEC, total parenteral nutrition is
commonly started on the first postoperative day in young infants until the intestinal motility returns.
Although significant gastrointestinal bleeding as a result of peptic ulcers or gastritis is rather uncommon
following CPB in infants, histamine-2 antagonists are routinely administered until enteral feeding has been
established.
Infectious Complications
A variety of factors predispose infants to infection following cardiac surgery. CPB generates a complex pro- and
antiinflammatory response; the latter contributes to a generalized state of immunosuppression (334,335).
Neonates in particular may have prolonged pre and postoperative ICU stays following cardiac surgery, during
which they are exposed to a variety of indwelling catheters and tubes that predispose to infection. The risk of
catheter-related blood stream infection in children increases with the number of central lines used, the duration
of catheter use, and the use of ECMO (336). Guidelines for the prevention and management of catheter related
infections in children have recently been published (337,338). Prophylactic antibiotics that provide coverage for
gram-positive organisms (e.g., a first generation cephalosporin) are administered in the early postoperative
period, generally until all of the chest tubes are removed from the patient (339).
The use of delayed sternal closure may be a risk factor for bacterial infections. In a series of children who had
delayed sternal closure, the vast majority of whom were infants, the reported incidences are 1% to 7% for
surgical site infection, and 0% to 4% for mediastinitis (189,190). Treatment of mediastinitis includes surgical
debridement and prolonged parenteral antibiotics.
Renal Insufficiency
CPB is associated with significant fluid retention in infants, and the problem is usually exacerbated by a decline
in urine output in the first 12 to 24 hours following surgery. Loop diuretics are used to improve urine output in
the early postoperative period. A continuous infusion of furosemide may provide a gentler diuresis in those
patients with labile hemodynamics. Spironolactone may be useful for its potassium sparing effect if the enteral
route of administration is available.
Significant renal insufficiency may occur in infants with prolonged CPB times or as a result of low cardiac output
in the early postoperative period (207). The development of overt renal failure requiring dialysis is associated
with 40% to 60% risk of mortality (340,341). Renal perfusion pressure may be calculated by the following
equation: renal perfusion pressure =mean arterial pressurecentral venous pressure. In general, a renal
perfusion pressure of 40 to 50 mmHg in infants is required to maintain urine output. Some surgeons electively
place temporary peritoneal dialysis catheters at the time of complex cardiac surgery in infants to allow drainage
of postoperative ascites and minimize external compression on the renal veins. Peritoneal dialysis may be
performed if needed. Renal function usually recovers following the improvement in cardiac output (340).
Neurologic Complications
Several preoperative, intraoperative, and postoperative factors may contribute to suboptimal neurologic
outcomes
P.747
following cardiac surgery in infancy. Recent reports demonstrate that preoperative neurological abnormalities
may be detected on physical examination and brain MRI in some term neonates with congenital heart disease
who otherwise had no known risk factors for neurologic injury (342,343).
The use of CPB may result in the embolization of gas bubbles or thrombus, which may contribute to the
development of seizures, infarcts, periventricular leukomalacia, and intracranial hemorrhage (343,344). The use
of deep hypothermic circulatory arrest (DHCA) for complex neonatal arch reconstruction has been practiced
since the 1960s. Data from the 1990s suggests that the use of longer periods of circulatory arrest, when
compared to low flow CPB, is associated with an increased incidence of early postoperative seizures (176,345).
Both early postoperative seizures and longer circulatory arrest times are associated with neurological deficits
following the arterial switch and Norwood operations (346,347,348,349,350,351). In order to minimize
neurologic injury related to the use of circulatory arrest, several centers have recently reported the use of
regional low-flow perfusion for complex aortic arch reconstructions (4,130,352,353). Concurrently, however,
there have been many advances in the pH and hematocrit strategies used in circulatory arrest that appear to
minimize adverse neurologic sequelae (354,355,356). For example, the incidence of early postoperative clinical
seizures following DHCA has decreased dramatically in patients from the current era (2%) as compared to the
cohort reported in the 1988-1992 Boston Circulatory Arrest Study (8%) (176,344). Thus it is unclear whether
neurological outcomes will be improved using regional perfusion techniques rather than DHCA.
Infants may have additional hypotensive or hypoxic episodes in the intensive care unit following surgery. These
postoperative events may exacerbate neurologic insults that occurred prior to or during surgery. Infants with a
longer postoperative length of stay appear to have a worse cognitive outcome when assessed at 8 years of age
(357).
Early Postoperative Physiology and Management: Specific Lesions

Palliation of the Neonate with a Single Ventricle
Neonates who have cardiac anatomy unsuitable for a two ventricular repair typically have atresia or significant
hypoplasia of either of the atrioventricular valves, and/or significant hypoplasia of either ventricle (358). The
majority of neonates with single ventricle physiology will eventually undergo Fontan completion as toddlers.
Because the Fontan circulation is characterized by passive pulmonary blood flow, this physiology is dependent
on low pulmonary artery pressure, and the absence of pulmonary artery distortion, pulmonary venous
obstruction, atrioventricular valve regurgitation and ventricular hypertrophy. These requirements dictate the
surgical strategy in the newborn period (358). To limit ventricular hypertrophy, unobstructed systemic blood
flow must exist, and if not present at birth, may be achieved using either a Norwood operation or Damus-KayeStansel procedure. Pulmonary blood flow must be tightly regulated to prevent ventricular volume overload,
atrioventricular valve regurgitation and pulmonary artery hypertension. This usually requires the placement of a
Blalock-Taussig shunt or pulmonary artery band. Care must be taken with both of these procedures to minimize
pulmonary artery distortion. Finally, pulmonary venous return must not be obstructed, as this may cause
pulmonary artery hypertension. An atrial septectomy is required in some neonates, as is done during the
Norwood operation.
Systemic to Pulmonary Artery Shunt

The most common type of systemic to pulmonary artery shunt operation refers to a procedure in which a GoreTex tube graft is surgically placed between the subclavian or innominate artery and the pulmonary artery,
typically without the use of CPB. This operation, also known as the modified Blalock-Taussig shunt, is widely
used as the operation of choice to palliate infants with single ventricle physiology and decreased pulmonary
blood flow (Fig. 34-15). The operation is also used in selected infants with two-ventricular physiology and
obstructed pulmonary blood flow whose complete repair will be deferred. A systemic to pulmonary artery shunt
may be performed in isolation or as a part of a larger procedure, such as the Norwood operation for HLHS. In a
retrospective, nonrandomized study, Odim and associates reported that performance of the modified BlalockTaussig operation via a midline sternotomy was technically easier, and was associated with a lower rate of shunt
failure when compared with a classic thoracotomy approach (359).
P.748
Another type of systemic to pulmonary artery shunt is a central shunt, which involves the placement of a GoreTextube between the ascending aorta and pulmonary artery. A central shunt is typically placed only when
anatomical issues preclude a modified Blalock-Taussig shunt. Several postoperative complications may occur
following the placement of a systemic to pulmonary artery shunt. If the shunt is too large, the systemic ventricle
will be subjected to volume overload and congestive heart failure may develop. In patients with single ventricle
physiology, elevated pulmonary vascular resistance may exist with the large shunt that may complicate future
staging operations. A small shunt will result in excessive postoperative cyanosis. The average shunt diameter is
3.5 mm for a term infant. Distortion or scarring of the pulmonary arteries may also develop at the shunt
insertion site. Thrombosis of the shunt is a dreaded postoperative complication that results in the rapid
development of severe cyanosis. Most infants are started on aspirin once their chest tubes and lines have been
removed. Reported neurologic complications with this shunt include Horner's syndrome, recurrent laryngeal
nerve injury, and phrenic nerve injury (359).
Figure 34-15 The modified Blalock-Taussig shunt operation. Reprinted from Wernovsky G, Bove EL. Single
ventricle lesions. In: Chang A, Hanley FL, Wernovsky G, et al., editors. Pediatric cardiac intensive care.
Baltimore: Williams and Wilkins, 1998:271-287, with permission.
Pulmonary Artery Band

In selected infants with single ventricular physiology and unobstructed pulmonary and systemic blood flow, or
occasionally in infants with two-ventricular physiology and pulmonary overcirculation, a restrictive band is placed
on the main pulmonary artery to decrease Qp/Qs and thus limit symptoms of congestive heart failure, the
adverse effects of volume loading on the ventricle(s), and the development of pulmonary vascular obstructive
disease (Fig. 34-16). CPB is not required for this procedure. It would be inappropriate to place a pulmonary
artery band in an infant who has, or is likely to develop, systemic outflow tract obstruction, as the resultant
doubly obstructed physiology will lead to the rapid development of ventricular hypertrophy. Placement of a
pulmonary artery band, although conceptually simple, requires a great deal of surgical judgment. If the band is
placed in an infant with high pulmonary vascular resistance, the patient's systemic oxygen saturation in the
operating room may indicate a well-balanced circulation (saturation approximately 80%). However, as the
pulmonary vascular resistance falls over the coming weeks, pulmonary blood flow may increase, systemic
oxygen saturation will rise, and signs of congestive heart failure will develop. If the pulmonary artery band
migrates distally on the main pulmonary artery, distortion of the branch pulmonary arteries may develop, which
would have to be corrected before or during the bi-directional Glen/hemi-Fontan operation. The band is removed
at the time of complete surgical repair or subsequent palliation.
Damus-Kaye-Stansel Procedure
In children with single ventricular physiology and subaortic stenosis with or without distal aortic arch
obstruction, a Damus-Kaye-Stansel palliation may be performed to provide unobstructed systemic blood flow.
This procedure involves anastomosis of the proximal main pulmonary artery and ascending aorta, and aortic
arch reconstruction if needed. The distal main pulmonary artery is over sewn and a modified Blalock-Taussig
shunt is placed to provide pulmonary blood flow. The postoperative physiology and complications are similar to
those seen following the Norwood operation, which are discussed in detail below.
Figure 34-16 Pulmonary artery band. Reprinted from Wernovsky G, Bove EL. Single ventricle lesions. In: Chang
A, Hanley FL, Wernovsky G, et al., editors. Pediatric cardiac intensive care. Baltimore: Williams and Wilkins,
1998:271-287, with permission.
Bi-directional Glenn or Hemi-Fontan Operation

The bi-directional Glenn or hemi-Fontan operation involves redirection of the superior vena cava blood flow to
both pulmonary arteries (Fig. 34-17). One of these two operations is commonly performed in infants with single
ventricle physiology before or at approximately six months of age, in anticipation of a Fontan completion, which
is typically done at 2 to 3 years of age (360). Because passive pulmonary blood flow is present following these
operations, they cannot be performed when pulmonary vascular resistance is high as in the newborn period
(361). The benefits of performing a bidirectional Glenn/hemi-Fontan operation as an intermediate staging
procedure before Fontan completion include volume-unloading of the single ventricle, decreased atrioventricular
valve regurgitation, and improved systemic oxygen saturation, and decreased
P.749
pleural effusions and mortality following the Fontan completion (360,362-364).
Figure 34-17 Bi-directional Glenn operation. Reprinted from Wernovsky G, Bove EL. Single ventricle lesions. In:
Chang A, Hanley FL, Wernovsky G, et al., editors. Pediatric cardiac intensive care. Baltimore: Williams and
Wilkins, 1998:271-287, with permission.
The bi-directional Glenn operation, also known as the bi-directional cavopulmonary shunt operation, involves an
end-to-side anastomosis between the superior vena cava and the pulmonary artery. The term bi-directional
refers to the fact that blood from the superior vena cava perfuses both the left and right pulmonary arteries. The
hemi-Fontan operation involves the end-to-side anastomosis between the superior vena cava and the pulmonary
artery as described above, but in addition, the proximal superior vena cava is anastomosed to the inferior
surface of the pulmonary artery, and a patch is placed at the junction of the superior vena cava and the right
atrium. The functional result of this operation is the same as the bi-directional Glenn operation, but it makes the
completion Fontan operation simpler if the surgeon ultimately uses a lateral tunnel type of Fontan. There may be
an increase risk of sinus node dysfunction following the hemi-Fontan operation related to surgical manipulation.
Surgeons that complete the Fontan using an extracardiac conduit perform the bi-directional Glenn shunt as a
staging procedure. Any central pulmonary artery stenosis is addressed during the bi-directional Glenn operation.
The bi-directional Glenn operation is associated with minimal morbidity or mortality, and most infants have an
unremarkable postoperative course (361,362). Superior vena cava syndrome, identified by the presence of
cerebral and upper extremity venous congestion, may develop related to increased pulmonary vascular
resistance, and therapies to reduce pulmonary vascular resistance should then be applied. Infants who are 2 to
3 months of age may have significant cyanosis immediately following the bi-directional Glenn/hemi-Fontan,
which usually improves in 24 to 48 hours (360). The evaluation and management of the postoperative bidirectional Glenn/hemi-Fontan patient with excessive hypoxemia is discussed in detail in the previous section on
cyanosis.
Fontan completion is generally performed at approximately two years of age. During this operation, the inferior
vena cava blood is channeled directly to the pulmonary arteries using several different techniques, functionally
bypassing the heart, and thus separating the systemic and pulmonary circulations and essentially eliminating
cyanosis. Factors that have been associated with a good outcome following the Fontan operation include low
pulmonary vascular resistance, undistorted pulmonary arteries and unobstructed pulmonary veins, and the
absence of significant atrioventricular valve regurgitation or ventricular hypertrophy. These factors are greatly
influenced be decision making in the neonatal period, as discussed earlier. As this chapter pertains to infancy,
the Fontan operation will not be discussed further.
Right Ventricular Outflow Tract Obstruction

Critical Pulmonary Valve Stenosis
The initial intervention in neonates with isolated critical pulmonary valve stenosis is usually a balloon
valvuloplasty, performed in the cardiac catheterization laboratory. Infants with critical pulmonary valve stenosis
may have very poor right ventricular compliance following balloon valvuloplasty, perhaps related to myocardial
ischemia and endocardial fibroelastosis (54). Rather than fill the right ventricle, some systemic venous blood will
cross the foramen ovale to the left atrium, resulting in significant cyanosis. To ensure adequate pulmonary blood
flow following the procedure, these infants may require several days of a PGE1 infusion (56,57). Occasionally a
systemic to pulmonary artery shunt will be required until right ventricular compliance improves (54,59).
Placement of a stent in the ductus arteriosus during cardiac catheterization is an alternative. The systemic to
pulmonary artery shunt or the stented ductus arteriosus can then be coil embolized or surgically removed at a
later date, once the right ventricular compliance has improved. These infants are also at risk for developing
infundibular spasm (suicide right ventricle) following balloon valvuloplasty that manifests as cyanosis and low
cardiac output (365). Volume expansion and beta-blockers are the initial treatments should infundibular spasm
occur.
Tetralogy of Fallot
Surgical repair of tetralogy of Fallot includes closure of the VSD(s), and reduction or elimination of obstruction to
blood flow across the right ventricular outflow tract and pulmonary
P.750
arteries. Accomplishing this often involves resection of muscle bundles in the right ventricular outflow tract, a
pulmonary valvotomy or leaflet resection, and patch augmentation of the proximal pulmonary arteries. In those
infants with favorable anatomy, the operation may be accomplished using a transatrial-transpulmonary
approach. However, right ventriculotomy with transannular patch is required in some cases to enlarge the right
ventricular outflow tract. In infants with tetralogy of Fallot and pulmonary atresia, a ventriculotomy is required if
a complete repair is performed using a right ventricle to pulmonary artery conduit.
Right ventricular function and compliance are abnormal following complete repair of tetralogy of Fallot. This is
attributable in part to right ventricular hypertrophy that was present prior to surgery, and also to right
ventricular myocardial injury related to the subpulmonic resection and VSD closure. Pulmonary regurgitation, if
present, will exacerbate the right heart failure. Some surgeons leave the foramen ovale patent so that the right
heart may decompress into the left atrium, thus maintaining systemic cardiac output at the expense of mild
cyanosis. If the surgeon closed the atrial septum in the operating room, and significant right heart failure exists
in the early postoperative period, an atrial communication may be created in the cardiac catheterization
laboratory. Inotropic support of the right ventricle is commonly required immediately following surgery, and low
dose dopamine and milrinone are commonly used in this regard. Strategies to maintain low pulmonary vascular
resistance should be employed. JET is a common rhythm disturbance seen following tetralogy of Fallot repair
(242). Following complete repair of tetralogy of Fallot, residual lesions that may complicate the postoperative
course include a VSD, right ventricular outflow tract obstruction, and pulmonary artery stenosis.
Neonates with tetralogy of Fallot and significant preoperative cyanosis who have a palliative systemic to
pulmonary artery shunt placed may be well saturated but develop congestive heart failure in the postoperative
period. The pulmonary blood flow from the shunt plus the native flow across the right ventricular outflow tract
may be excessive.
Surgical repair for tetralogy of Fallot with absent pulmonary valve syndrome includes a reduction pulmonary
arterioplasty, VSD closure and placement of a valved homograft in the right ventricular outflow tract (317).
Replacement of the dilated central pulmonary arteries with bifurcated valved pulmonary homograft is a surgical
modification that has been associated in one report with improved survival in neonates who present with
respiratory distress (366). A more recently described surgical approach includes translocation of the pulmonary
artery anterior to the aorta (Lecompte maneuver) and away from the tracheobronchial tree (367). In addition to
the aforementioned issues following tetralogy of Fallot repair, the postoperative course for infants with absent
pulmonary valve syndrome may be significantly complicated by airway obstruction as a result of
tracheobronchomalacia that persists even after plication and reduction of the aneurysmal pulmonary arteries
(317,368). Case reports suggest that those with refractory tracheobronchomalacia may improve with metallic
stenting of the central airways (317). Some of these infants will require a tracheostomy and long-term positivepressure ventilation until they outgrow the malacia.
Pulmonary Atresia, Ventricular Septal Defect, Major

Aortopulmonary Collateral Arteries
The goal of surgical intervention in infants with pulmonary atresia, a VSD and MAPCAs is to incorporate blood
vessels that supply as many lung segments as possible into the repair, while eliminating any MAPCAs that
represent dual blood supply to these segments. The interventional management must be individualized
depending on the pulmonary artery anatomy and often involves a series of surgical procedures and
transcatheter interventions with balloon dilation of the stenotic branch pulmonary arteries. Surgical options
include a staged approach involving bilateral unifocalizations followed by repair using a right ventricular to
pulmonary artery conduit (369). A single stage, complete repair is advocated for many of these infants in early
infancy at some centers (370). In selected infants with diminutive but confluent pulmonary arteries, a
transannular patch to open the right ventricular outflow tract, placement of an right ventricle to pulmonary
artery conduit, or an aortopulmonary window may be used to encourage pulmonary artery growth (371).
Complete repair including VSD closure is then performed at a later date provided that the pulmonary vascular
bed is adequate to accept the entire right ventricular output without the development of severe pulmonary
hypertension. The VSD patch may be fenestrated in selected cases to prevent the development of postoperative
suprasystemic pulmonary hypertension and right ventricular failure (372).
The postoperative course for infants with pulmonary atresia, ventricular septal defect and MAPCAs is dependent
on the operative strategy. Those infants undergoing complete repair including closure of the VSD and placement
of a conduit from the right ventricle to the pulmonary artery are at risk for right ventricular failure, secondary to
the VSD closure, right ventricular outflow tract muscle bundle resection, ventriculotomy required for conduit
placement, and branch pulmonary stenosis. Often the VSD patch is fenestrated, or a foramen ovale is left
patient in these patients, which leads to mild postoperative cyanosis, but serves to decompress the right heart
and preserve systemic cardiac output (372). Congestive heart failure may persist in the postoperative period if
all of the MAPCAs were not incorporated into the repair. Following unifocalization procedures, severe airflow
limitation may occur secondary to tracheobronchial ischemia, which may originate from interruption of
peribronchial arterial blood supply during mobilization of MAPCAs (321). The intermediate term prognosis is
guarded for infants with pulmonary atresia, VSD and MAPCAs secondary to pulmonary hypertension
P.751
and right ventricular failure (66). Serial cardiac catheterizations with balloon angioplasty of the pulmonary
arteries are required for many patients.

The postoperative course for infants with pulmonary atresia and intact ventricular septum is highly dependent
upon the operation or transcatheter procedure performed. Neonates undergoing a two-ventricular repair may
have poor right ventricular compliance, right to left shunting at the foramen and subsequent cyanosis, similar to
that seen following repair of tetralogy of Fallot. Those with right ventricular dependent coronary circulation
may have problems with myocardial ischemia following a systemic to pulmonary artery shunt, even though the
right ventricle was not decompressed.

Neonates with Ebstein's anomaly who require early surgical intervention are at high risk for mortality as a result
of pulmonary hypoplasia, pulmonary hypertension and right heart failure. Surgical options for neonates with
refractory symptoms include a primary complete repair, consisting of a reduction atrioplasty, fenestrated closure
of the atrial septum and tricuspid valvuloplasty (73). Another approach is to stage the patient toward Fontan
palliation with a systemic to pulmonary artery shunt, over sewing of the tricuspid valve annulus and atrial
septectomy (74,75). Regardless of the surgical procedure, infants recovering from surgery for Ebstein's anomaly
are likely to have problems with pulmonary hypertension related in part to pulmonary hypoplasia. In addition to
the usual strategies used to treat pulmonary hypertension, care should be taken to ensure that excessive
positive airway pressure is not used to attempt to expand lungs, which may appear small on chest radiograph.
Overdistension of the hypoplastic lungs will only lead to further increases in pulmonary vascular resistance.
Residual tricuspid regurgitation and right heart failure may also complicate the postoperative course.
Atrioventricular reentry tachycardia may occur at any time before or after surgery.
Left Ventricular Outflow Tract Obstruction

Critical Aortic Valve Stenosis
Following the alleviation of critical aortic valve stenosis, either by surgical valvotomy or balloon valvuloplasty,
most infants will have a marked improvement in clinical status with recovery of left ventricular function. Note
that the gradient across the aortic valve may be minimal immediately following intervention, but will often
increase as myocardial function and cardiac output improve. Inotropic support is often required for a few days
following the procedure. The PGE1 infusion may be discontinued once left to right flow is seen at the ductal level
by echocardiography. A small number of infants will do poorly in the postoperative period, as a result of
inadequate size of left heart structures or extensive endocardial fibroelastosis. Such patients may require a
Norwood palliation, but historically, infants who had a failed attempt at a two-ventricular repair and subsequent
attempted Norwood operation have a high mortality (373).
Critical Coarctation of the Aorta

Several operative techniques are available for coarctation repair, including resection with end-to-end
anastomoses, patch aortoplasty, subclavian flap aortoplasty. Resection and extended end-to-end anastomosis is
a technique that also addresses aortic arch hypoplasia and appears to be associated with a low recoarctation
rate (374,375). Due to a significant recoarctation rate and the potential for aneurysm formation, primary balloon
angioplasty is generally not recommended in infancy at most congenital heart centers (376,377).
Following repair of an isolated critical coarctation of the aorta, accurate four extremity blood pressure
measurements must be obtained to evaluate for the presence of residual aortic arch obstruction. A residual armleg blood pressure gradient may be secondary to an inadequate repair, hypoplasia of the aortic arch, or
constriction of residual ductal tissue. If a residual coarctation exists, the timing of intervention depends upon the
clinical course. Infants who have adequate left ventricular function and who are easily weaned from support may
return in 2 to 3 months, after the suture lines have healed, for cardiac catheterization and possible balloon
dilation of the obstruction. Those who are unable to be weaned from the ventilator require another procedure
during the same hospitalization. Infants who presented in shock may be at risk for mesenteric ischemia and
enteral feedings are withheld in the early postoperative period. Infants appear to be at much lower risk for
spinal cord ischemia or significant postoperative hypertension following coarctation repair when compared with
older children and adults. As with any surgical intervention on the aortic arch, the phrenic and recurrent
laryngeal nerves are at risk for injury during coarctation repair.

During the Norwood operation for HLHS, the distal pulmonary artery is over sewn, the proximal pulmonary
artery and the aorta are anastomosed, and the aortic arch is reconstructed with pulmonary artery homograft
material to allow for unobstructed systemic blood flow (378). An atrial septectomy is performed to allow
pulmonary venous return to pass easily to the right atrium, and a modified Blalock-Taussig shunt is placed to
provide pulmonary blood flow (Fig. 34-18). Thus following the Norwood operation, the single right ventricle
pumps blood to the systemic circulation and coronary arteries, via the reconstructed aorta (neoaorta), and to
the pulmonary circulation, via the systemic to pulmonary artery shunt.
Figure 34-18 Norwood operation for hypoplastic left heart syndrome. The purpose of the Norwood operation for
hypoplastic left heart syndrome is to establish reliable, unobstructed outflow to the systemic circulation and to
balance the systemic and pulmonary circulations. This is achieved by ligating and dividing the ductus arteriosus
and detaching the central and branch pulmonary arteries from the main pulmonary artery as shown in A. The
hypoplastic aorta is then opened from the descending aorta retrograde to the level of the aortic valve and
augmented with a patch of homograft arterial wall. The augmented aorta is then connected to the cardiac end
of the main pulmonary artery stump as shown in B and C. This achieves unobstructed outflow from the single
right ventricle through the pulmonary valve to the aorta and systemic vascular bed. Additionally, an atrial
septectomy is performed to ensure unobstructed pulmonary venous outflow, and the pulmonary circulation is
supplied through the use of a right-sided modified Blalock-Taussig shunt. Reprinted from Wernovsky G, Bove
EL. Single ventricle lesions. In: Chang A, Hanley FL, Wernovsky G, et al., editors. Pediatric cardiac intensive
care. Baltimore: Williams and Wilkins, 1998:271-287, with permission.
P.752
A variety of problems may arise following the Norwood operation that often are poorly tolerated as a result of
the limited physiologic reserve possessed by these infants. Excessive pulmonary blood flow with inadequate
systemic perfusion, similar to that seen in a typical preoperative neonate with HLHS, may be seen but is less
common than in prior years due the smaller diameter Blalock-Taussig shunts that are currently used. Others
may have problems with myocardial dysfunction, pulmonary hypertension, or the systemic to pulmonary artery
shunt.
Ideally, the circulations will be balanced following the Norwood operation such that the Qp/Qs ratio will be
approximately 1:1. Pulmonary blood flow will be determined by the pulmonary vascular resistance and the
resistance provided by the systemic to pulmonary artery shunt, assuming an unrestrictive, surgically created
ASD and unobstructed pulmonary venous blood flow. Systemic blood flow will be determined by the systemic
vascular resistance and if present, any residual aortic arch obstruction. The ratio of systemic and pulmonary
blood flow may be calculated by the following modified Fick equation: Qp/Qs =(SaO2-SvO2)/(PvO2-PaO2), in
which SaO2 =aortic saturation; SvO2 =mixed venous saturation (estimated by the superior vena cava oxygen
saturation); PvO2 =pulmonary venous oxygen saturation (assumed to be >95% if not directly measured); and
PaO2 > pulmonary artery oxygen saturation (assumed to equal the aortic saturation). Estimating Qp/Qs using
only the arterial PaO2 or SaO2 may be misleading without knowledge of SvO2, estimated by the superior vena
cava oxygen saturation, and PvO2 (Fig. 34-19) (191,379-383). One study documented pulmonary venous
desaturation in 11/12 infants following the Norwood operation at some point in the early postoperative period,
despite the absence of radiographic abnormalities on chest radiograph (191). Many centers monitor SvO2
following the Norwood operation, but few monitor PvO2. Caution must be used when using saturation data for
clinical decision-making following the Norwood operation. The CXR must be carefully examined to ensure that
the tips of the catheters from which blood gas samples are drawn are in the proper location. The oximetry data
cannot be used in isolation and the entire clinical picture must be considered.
Signs of low cardiac output following the Norwood palliation may be attributed to pulmonary overcirculation,
myocardial dysfunction, or a combination of the two. An imbalance between the systemic and pulmonary blood
flows, particularly a high Qp/Qs, has been implicated as a major factor in the high early postoperative mortality
following the Norwood operation. Such infants will have high SaO2, low SvO2, and other clinical and laboratory
evidence for low cardiac output. Investigators have thus spent
P.753
considerable effort to study the cardiovascular physiology following the Norwood operation, and several
interventions are available that may serve to better balance the pulmonary and systemic circulations. In an
experimental animal model of postoperative single ventricle physiology with a systemic to pulmonary shunt,
addition of carbon dioxide (PiCO2) to the inspired gas resulted in increase pulmonary vascular resistance, even
when systemic pH was controlled by administration of tromethamine buffer (Tham) (85). Other investigators
have reported the ability to manipulate pulmonary vascular resistance, Qp/Qs, and oxygen delivery in single
ventricle animal models by varying the inspired concentrations of oxygen or providing inhaled carbon dioxide or
NO (380,384).
Figure 34-19 Regression analysis of SaO2 against true Qp/Qs following the Norwood operation. SaO2 is a poor
predictor of Qp/Qs (R2 = 0.08, p < 0.05). Variability in Qp/Qs is most pronounced at SaO2 values in range of
65% to 85%, usual target range for patients after Norwood palliation. Reprinted from Taeed R, Schwartz SM,
Pearl JM, et al. Unrecognized pulmonary venous desaturation early after Norwood palliation confounds Qp/Qs
assessment and compromises oxygen delivery. Circulation 2001; 103:2699-2704, with permission.
The findings from preoperative studies in neonates and postoperative animal models that have evaluated the
use of sub-ambient oxygen or inspired carbon dioxide to alter pulmonary vascular resistance should not be
extrapolated to neonates recovering from the Norwood operation. Doing so would not account for several
important variables, including the fixed resistance provided by the relatively small (3.5 to 4 mm) systemic to
pulmonary shunts used in most Norwood operations in the current era as compared to the 6 mm shunts used in
some animal models (4,380,385). Although early reports suggest that an imbalance of systemic to pulmonary
blood flow may be life threatening to infants following the Norwood operation, several recent postoperative
studies suggest that supplemental oxygen or alkalosis do not have deleterious effects on the hemodynamics of
most patients with Norwood physiology, presumably as a result of the fixed resistance provided by the systemic
to pulmonary shunt (94,191,382,383,386,387). With the above considerations in mind, an infant recovering
from the Norwood operation with signs of isolated pulmonary overcirculation (i.e., high calculated Qp/Qs,
pulmonary edema on chest radiograph) should have measures taken to lower systemic vascular resistance and
increase pulmonary vascular resistance. Residual aortic arch obstruction must also be excluded in this situation,
as such an obstruction will limit systemic blood flow and direct more blood through the shunt.
Several recent reports suggest that myocardial dysfunction may be the dominant hemodynamic disturbance in
those infants with low cardiac output following the Norwood operation (201,383,388). Tricuspid valve
regurgitation may also contribute to low cardiac output. Isolated myocardial dysfunction is present when the Qp/
Qs is calculated to be approximately 1:1, and the SvO2 is low. An anaerobic threshold is reached when the SvO2
falls below 30% following the Norwood operation, and efforts to maintain SvO2 above this value have been
associated with very low early mortality (389). Inotropic support and afterload reduction will be beneficial for
such patients. A study examining the effect of various inotropic agents in an animal model of single ventricle
physiology found that low-dose dopamine and epinephrine had favorable effects on Qp/Qs, and epinephrine
improved oxygen delivery, whereas dobutamine increased Qp/Qs and decreased oxygen delivery (390).
The use of phenoxybenzamine, an alpha-adrenergic blocker, may address both the problems of pulmonary
overcirculation and myocardial dysfunction following the Norwood operation. In a prospective, nonrandomized
study, Tweddell and associates report favorable hemodynamic effects of administering phenoxybenzamine to
neonates undergoing the Norwood palliation. Phenoxybenzamine produced lower systemic vascular resistance,
resulting in a Qp/Qs that approached unity, and a higher SvO2 (Fig. 34-20 and 34.21)
P.754
(382). In a retrospective study, Tweddell et al found several variables in addition to the use of
phenoxybenzamine that may be associated with improved survival following the Norwood operation, including
the use of a modified operative technique designed to avoid kinking of the ascending aorta, use of modified
ultrafiltration and aprotinin, and postoperative continuous monitoring of SvO2 (4). Phenoxybenzamine has a long
half-life, which makes some clinicians reluctant to use the drug.
Figure 34-20 Superior vena cava saturation (SvO2) during the first 48 hours after the Norwood procedure. The
SvO2 was significantly higher in the phenoxybenzamine group than in the control group during hours 1 to 10
and hour 25 (p < 0.05, RM ANOVA). Reprinted from Tweddell JS, Hoffman GM, Fedderly RT, et al.
Phenoxybenzamine improves systemic oxygen delivery after the Norwood procedure. Ann Thorac Surg
1999;67:161-167, with permission from The Society of Thoracic Surgeons.
Figure 34-21 Pulmonary to systemic flow ratio (Qp/Qs) during the first 48 hours after the Norwood procedure.
The Qp/Qs was significantly higher in the control group than in the phenoxybenzamine group during hours 1 to
10, 30 to 32, and 36 (p < 0.05, RM ANOVA). Reprinted from Tweddell JS, Hoffman GM, Fedderly RT, et al.
Phenoxybenzamine improves systemic oxygen delivery after the Norwood procedure. Ann Thorac Surg
1999;67:161-167, with permission from The Society of Thoracic Surgeons.
Excessive desaturation may occur following the Norwood operation, and the differential diagnosis and
management strategies are discussed previously in the postoperative cyanosis section. One particular cause for
cyanosis following the Norwood operation is inadequate surgical resection of the atrial septum. The septum
primum is often displaced posteriorly and to the left. If it is not completely resected, pulmonary venous
obstruction may occur.
Recent interest has developed in a surgical variant of the Norwood operation, commonly known as the Sano
modification (391). In this operation, a right ventricular to pulmonary artery conduit instead of a systemic to
pulmonary artery shunt is used to provide pulmonary blood flow, which may improve myocardial function by
increasing diastolic blood pressure and thus coronary perfusion pressure (352,392). The postoperative
physiology following the Sano modification is somewhat different from the standard Norwood operation and
deserves comment. Due to the absence of diastolic runoff through a Blalock-Taussig shunt, infants with Sano
physiology do not have pulmonary overcirculation (or a large Qp/Qs) and thus, in contrast to patients with
typical Norwood physiology, usually will not benefit from aggressive afterload reduction. In fact, excessive
systemic vasodilation may be harmful by creating inadequate pulmonary blood flow and severe cyanosis, similar
to the physiology of a patient with unrepaired tetralogy of Fallot. Early, nonrandomized reports suggest that the
Sano modification may improve early survival following stage I palliation of HLHS (391,393). Risks related to
this modification center upon the effects of performing a ventriculotomy in an infant with a single right ventricle,
and include myocardial dysfunction, arrhythmias, and false aneurysm formation. A multicenter, randomized,
clinical trail is planned to compare outcomes following the Sano modification with the standard Norwood
operation.
Interrupted Aortic Arch

Complete repair, including VSD closure and reconstruction of the aortic arch, is the preferred operation for
newborns with IAA (101). If significant subaortic stenosis is present, a Norwood-type operation may be used to
stage the patient for a later biventricular repair (394). Following IAA repair, one must ensure that a residual
VSD, sub-aortic obstruction, or aortic arch obstruction does not exist (101,395). Patients with respiratory
compromise may have compression of the left mainstem bronchus by the reconstructed aortic arch.
Complications related to DiGeorge syndrome, such as hypocalcemia and infection, may also occur.
Mixing Lesions
Transposition of the Great Arteries
In neonates with TGA and no significant outflow tract obstruction, the arterial switch operation is performed and
the VSD, if present, is closed (1). The arterial switch operation involves transecting the aorta and pulmonary
arteries above the semilunar valves, and anastomosing the aorta to the neoaortic root, such that the left
ventricle ejects into the systemic circulation (1). The pulmonary artery is brought anterior to the aorta such that
its branches drape over the aorta (Lecompte maneuver), and the pulmonary artery is anastomosed such that it
receives blood from the right ventricular outflow tract. The coronary arteries are mobilized with a button of
tissue around the ostia and reimplanted into the neoaorta. The VSD (if present) is closed (Fig. 34-22). If
significant right ventricular outflow tract obstruction is present in TGA, a Damus-Kaye-Stansel procedure may be
performed as an initial palliation, or as a part of the complete repair including closure of the VSD and placement
of a right ventricle to pulmonary artery conduit. If significant left ventricular outflow tract obstruction exists, the
classic surgical repair is the Rastelli procedure, which involves using a baffle to close the VSD to the aorta and
placement a right ventricle to pulmonary artery conduit (396). Other options that do not require a conduit
include the Nikaidoh operation (aortic root translocation into the surgically enlarged left ventricular outflow tract,
VSD closure, and right ventricular outflow tract reconstruction) or the Rparation l'Etage ventriculaire (REV
procedure; involving infundibular resection to enlarge the VSD,
P.755
baffle closure of the VSD directing left ventricular flow to the aorta, the Lecompte maneuver, and right
ventricular to pulmonary artery connection using a patch) (397,398).
Figure 34-22 Arterial switch operation for transposition of the great arteries (TGA). A. The external anatomy is
shown. The procedure is performed using cardiopulmonary bypass and either moderate or deep hypothermia
with or without circulatory arrest. The broken lines show the sites of transection of the two great vessels.The
aorta and main pulmonary arteries have been surgically transected and the coronary ostia have been removed
from the native aortic root.The coronary buttons are in the process of being transferred to the native aortic root.
The coronary transfer has been completed and the neoaortic root has been anastomosed to the ascending
aorta. The coronary explantation sites on the neopulmonary root have been repaired with a patch and the
neopulmonary artery is in the process of anastomosis to the distal pulmonary artery. Note, the distal
pulmonary artery has been moved anterior to the ascending aorta as described by Lecompte. The procedure
also involves closing the atrial septal defect and dividing the patent ductus arteriosus. Reprinted from
Wernovsky G, Jonas RA. Other conotruncal lesions. In: Chang A, Hanley FL, Wernovsky G, et al., editors.
Pediatric cardiac intensive care. Baltimore: Williams and Wilkins, 1998:289-301, with permission.
The expected mortality for the arterial switch operation is very low (399), and many infants will have an
unremarkable postoperative course. A period of early low cardiac output has been documented in approximately
25% of these patients (140). Problems with coronary ischemia may occur if there is kinking or stenosis of a
coronary artery following reimplantation into the neoaorta. Volume overload in the immediate postoperative
period can cause cardiac distension that stretches the newly implanted coronary arteries, resulting in myocardial
ischemia and/or infarction. Coronary insufficiency may manifest as low cardiac output with ischemic
electrocardiographic changes or ventricular arrhythmias. Significant obstruction of the aortic or pulmonary
anastomosis is unusual in the early postoperative period. Residual VSDs and ventricular outflow obstruction can
be problematic in complex cases.
Infants with TGA and intact ventricular septum who are referred after 1 to 2 months of age for the arterial
switch operation often require placement of a pulmonary artery band to prepare the left ventricle before the
arterial switch operation (400). A systemic to pulmonary artery
P.756
shunt is also placed at the time of the pulmonary artery band to ensure adequate pulmonary blood flow. These
patients are often critically ill following this operation with biventricular failure (right ventricular failure as a
result of the acute volume overload created by the shunt and left ventricular failure related to the acute increase
in afterload from the pulmonary artery band) and low cardiac output (400,401). Inotropic support and measures
to decrease pulmonary overcirculation are often required for several days until the hemodynamics stabilize.
Once the left ventricle is prepared, the patient is returned to the operating room, in which the systemic to
pulmonary artery shunt and pulmonary band are removed and the arterial switch operation is performed.
Tricuspid Atresia
The postoperative course following palliation for tricuspid atresia depends upon the operation performed. Infants
who require surgery in early infancy will have either a systemic to pulmonary artery shunt or a pulmonary artery
band, and the complications associated with these procedures were described earlier. Infants with tricuspid
atresia and TGA often require a Damus-Kaye-Stansel procedure, and the postoperative problems are similar to
those seen following the Norwood operation as described above. A bi-directional Glenn or hemi-Fontan operation
is then performed at around six months of life, and the postoperative course for these operations is also
discussed earlier in the chapter.
Total Anomalous Pulmonary Venous Return

In infants with supracardiac and infracardiac total anomalous pulmonary venous return, the pulmonary venous
confluence is anastomosed to the back of the left atrium using circulatory arrest, the primitive vertical vein is
ligated, and the ASD is closed. If the pulmonary veins drain to the coronary sinus, surgical intervention involves
unroofing of the coronary sinus, and closure of the ASD.
Following repair of TAPVR, infants may be predisposed to pulmonary hypertensive crises or persistent pulmonary
vein obstruction, in part related to abnormal muscularity of the pulmonary arteries and veins which develops in
utero (106,402,403). Pulmonary edema and poor lung compliance may be seen in infants whose pulmonary
veins were obstructed preoperatively. The left atrium may be small and poorly compliant, and rapid volume
infusions may exacerbate pulmonary edema and pulmonary hypertension in this setting. Deep sedation and the
judicious use of positive end-expiratory pressure (PEEP) are indicated in the early postoperative period to
facilitate gas exchange. The pulmonary hypertension seen in some infants following repair of TAPVR is very
responsive to INO (270).
Truncus Arteriosus
Surgical repair of truncus arteriosus involves VSD closure, removal of the pulmonary arteries from the arterial
trunk, and placement of a conduit from the right ventricle to the pulmonary arteries. If present, surgical
attention is also given to the IAA and/or truncal valve insufficiency (404). Care must be taken to identify the
origin and course of the coronary arteries so that they are not injured during the operation.
Following repair of truncus arteriosus, infants are at risk for developing right ventricular failure related to
pulmonary hypertension, the VSD closure, and the ventriculotomy required for the right ventricle to pulmonary
artery conduit. Pulmonary hypertension may be problematic, particularly in patients referred for surgery beyond
the neonatal period. Surgeons may leave an atrial communication to decompress the right heart, at the expense
of mild postoperative cyanosis. The truncal valve functions as the aortic valve following surgical repair. The
severity of preoperative truncal valve stenosis usually is decreased in the postoperative period as the volume
overload has been eliminated. Severe truncal valve insufficiency will be poorly tolerated in the postoperative
period and may require a surgical valvuloplasty or replacement (109,404). The left ventricle is exposed to
increased afterload following surgery as a result of the elimination of the low vascular resistance pulmonary
circulation, and the effects of CPB, which may create temporary left ventricular failure. A residual VSD will add a
volume load to the left ventricle and contribute to pulmonary hypertension. A state of low cardiac output may
exist related to one or more of the above problems. If an IAA was repaired, care must be taken to ensure that
residual arch obstruction does not exist and exacerbate myocardial dysfunction and truncal valve insufficiency.
Left to Right Shunts

Atrial Septal Defect
Secundum ASDs may be surgically closed primarily or with a pericardial or Gore-Tex patch. Primum ASDs
typically require patch closure, and the commonly associated cleft in the mitral valve is usually sutured closed.
Sinus venosus ASDs are typically associated with anomalous pulmonary venous drainage of the right pulmonary
veins, which must be redirected to the left atrium.
Most patients may be extubated in the operating room or soon after returning to the intensive care unit
following ASD closure (405). Atrial arrhythmias may occur following any ASD closure. There is an increase
incidence of sinus node dysfunction following sinus venosus ASD repair that is usually transient, however,
temporary atrial pacing may be required (406).

The surgical approach to a VSD depends on its location. Perimembranous and inlet VSDs are approached
through the right atrium. Tricuspid valve detachment may be required for adequate exposure of the defect
(407). Conal VSDs are approached through the pulmonary artery
P.757
(408). Muscular VSDs may be approached through the atrium, although apical defects often require a small
ventriculotomy (409). Selected infants with large VSDs who are not thought to be candidates from primary
repair because of patient size, the presence of multiple VSDs with anticipated difficult surgical exposure, or
noncardiac co-morbidities may be palliated with a pulmonary artery band.
The postoperative course following repair of large, isolated VSDs may be complicated by pulmonary
hypertension, although this is quite uncommon in the current era as a result of earlier repair. The incidence of a
significant residual VSD, which also will cause pulmonary hypertension, is minimized with the routine use of
transesophageal echocardiography. Patch dehiscence is a rare complication of VSD closure that may occur once
the infant leaves the operating room and thus will be not detected by the transesophageal echocardiogram. Data
that would suggest a significant residual VSD include elevated left atrial and pulmonary artery pressures, a stepup of greater than 10% in oxygen saturation from the superior vena cava or right atrium to the pulmonary
artery, the presence of low cardiac out syndrome, persistent cardiomegaly and increased pulmonary vascular
markings on chest radiograph, and difficulty weaning from the ventilator. JET or surgical heart block may occur
following repair of a VSD. Right bundle branch block commonly occurs but does not appear to be of long-term
significance.
Atrioventricular Canal Defect

Incomplete and transitional atrioventricular canal defects typically are repaired with a patch closure of the ASD
and, commonly, suture closure of the cleft in the mitral valve. The postoperative course is similar to that seen
following ASD closure as described above.
Complete atrioventricular septal defect repair may be accomplished using a one or two patch technique to close
the primum atrial and inlet ventricular septal defects, dependent on patient anatomy and surgical preference
(410). The common atrioventricular valve is divided to create separate tricuspid and mitral valves. All of the
postoperative concerns following VSD closure as discussed above also apply to the infant following
atrioventricular canal repair. These include pulmonary hypertension, residual VSDs, and arrhythmias
(specifically, JET and complete heart block). Additionally, mitral valve regurgitation may occur, causing
ventricular volume overload, congestive heart failure, and pulmonary hypertension (411). Mitral regurgitation
may be suspected by auscultation and the identification of large ventricular waves on the left atrial line tracing,
and further quantified by echocardiography. Afterload reduction should be initiated when significant mitral
regurgitation is identified, and volume overload should be avoided. Mitral stenosis is less common, but will cause
elevated left atrial pressures and large atrial waves on the left atrial pressure tracing.

PDAs may be ligated surgically via a thoracotomy or using a video-assisted thoracoscopic technique (VATS).
Specific complications that are uncommonly encountered are related to the injury of nearby structures, including
the recurrent laryngeal and phrenic nerves. Coils and occluding devices may be used to close PDAs in the
cardiac catheterization laboratory. Hemolysis or obstructions by the coil of the aorta or the proximal left
pulmonary artery are rare complications following transcatheter occlusion of a PDA.
Although several surgical techniques have been used in the past for aortopulmonary windows, currently the
transaortic patch repair is favored by many centers (119). Any associated cardiac lesions are repaired at the
same operation. Postoperative problems that may be encountered include pulmonary hypertension and residual
left to right shunting. A related and somewhat rare lesion in which one pulmonary artery arises from the
ascending aorta and the other from the right ventricle deserves special consideration. Preoperatively, the
pulmonary artery that arises from the ascending aorta is exposed to systemic pressures, and the other
pulmonary artery receives the entire cardiac output from the right ventricle. This unique combination of high
pressure to one pulmonary vascular bed and high flow to the other predisposes these infants to having
significant problems with pulmonary hypertension in the immediate postoperative period.
Heart Transplantation
The classical surgical technique used in heart transplantation involves a bi-atrial anastomosis, and connection of
the great vessels at the level of the ascending aorta and main pulmonary artery. Many surgeons are now
performing a bicaval anastomosis, which eliminates suture lines in the right atrium, includes the donor sinus
node in the transplantation, and may reduce the incidence of atrial arrhythmias (412).
Several unique issues must be considered during the early postoperative period following heart transplantation.
The characteristics of the donor heart may impact on the postoperative course. The circumstances leading to
brain death of the donor, the amount of inotropic support required prior to the organ harvest, and the duration
of total ischemic time all impact on myocardial function following transplantation. The size of the donor heart in
relation the recipient is also important. For example, a larger donor heart may be useful to overcome pulmonary
hypertension in the recipient but may make sternal closure difficult. The condition of the recipient going into the
transplant will certainly impact the postoperative course. For example, inadequate nutrition, pulmonary
hypertension or renal insufficiency may have been present before surgery and will influence postoperative
management.
P.758
Immunosuppression is required to prevent rejection and is based on a triple drug regimen including
corticosteroids, a calcineurin inhibitor (cyclosporine or tacrolimus) and an antiproliferative agent (azathioprine or
mycophenolate mofetil) (413). This strategy suppresses the immune system at multiple different levels while
minimizing the toxicity of individual drugs. The use of induction immunotherapy, which refers to the
administration of anti-T cell antibodies, may be considered immediately following transplant in patients at high
risk for rejection or renal insufficiency (414,415). Induction immunotherapy may allow for dose reduction or
delayed introduction of calcineurin inhibitors, allowing time for recovery of renal function. The risk of rejection is
highest immediately following transplantation, and thus intense immunosuppression is initiated and then tapered
over time. Serial echocardiograms and myocardial biopsies are needed for close surveillance. Signs and
symptoms of severe rejection include tachycardia, hypotension, tachypnea, poor perfusion and arrhythmias.
Treatment of rejection centers on the augmentation of immune suppression and supportive care until recovery
of myocardial function occurs.
Severe pulmonary hypertension may lead to right ventricular failure, and aggressive measures may be required
to lower the pulmonary vascular resistance and support the right ventricular systolic function. In patients who
are known to be at risk for postoperative hypertension, it may be advantageous to transplant a slightly
oversized donor heart. If the donor heart is too large, however, the sternum may need to be left open for
several days.
Systemic hypertension is common following heart transplantation related in part to fluid retention, and the side
effects of corticosteroids and cyclosporine. Treatment is based on the severity of hypertension and the
availability of the enteral route of delivering medications. Intravenous vasodilators and oral antihypertensive
agents, such as calcium channel blockers and angiotensin-converting enzyme inhibitors, are commonly used.
Seizures may also occur, sometimes related to severe hypertension or cyclosporine toxicity.
The risk of infection following heart transplantation is increased due to a variety of factors. Many infants have
prolonged preoperative ICU stays, prolonged exposure to central venous catheters, and malnutrition.
Immunosuppression is intense in the early postoperative period. Thus vigilance for infection is high immediately
following surgery. In addition to standard surgical antibiotic prophylaxis against gram-positive organisms,
additional prophylactic drugs are commonly prescribed to minimize the incidence of infection against
cytomegalovirus, herpes viruses and Pneumocystis carinii pneumonia.
CONCLUSION
Neonates and infants with critical congenital heart disease present a tremendous challenge to all medical
personnel who are involved in their care. The general pathophysiology, presentations, and management
strategies describe above serve as a starting point to guide physicians toward providing state of the art care.
Subtle variations in cardiac anatomy and physiology exist within each type of heart defect, and the importance
of developing individualized medical management and interventional plans for each patient cannot be
overstated. Optimal outcomes are achieved when problems are anticipated and addressed in a timely fashion
(proactive, not reactive management). Frequent reevaluation of response to treatment is essential. Detailed
communication across all subspecialties is equally important to achieve optimal outcomes. Perioperative care for
infants with congenital heart disease continues to evolve, as new medications, treatment strategies, and surgical
and interventional techniques are reported on a routine basis. Fetal heart surgery and transcatheter
interventional procedures are still in the developmental stages but are likely to play an expanding role in this
patient population, and this trend will certainly present a new set of challenges in the years to come (92,416418).
REFERENCES
1. Jatene AD, Fontes VF, Paulista PP, et al. Anatomic correction of transposition of the great vessels. J Thorac
2. Parry AJ, McElhinney DB, Kung GC, et al. Elective primary repair of acyanotic tetralogy of Fallot in early
infancy: overall outcome and impact on the pulmonary valve. J Am Coll Cardiol 2000;36:2279.
3. Hanley FL, Heinemann MK, Jonas RA, et al. Repair of truncus arteriosus in the neonate. J Thorac Cardiovasc
Surg 1993; 105:1047.
4. Tweddell JS, Hoffman GM, Mussatto KA, et al. Improved survival of patients undergoing palliation of
hypoplastic left heart syndrome: lessons learned from 115 consecutive patients. Circulation 2002;106:I82.
5. Newburger JW, Silbert AR, Buckley LP, et al. Cognitive function and age at repair of transposition of the
great arteries in children. N Engl J Med 1984;310:1495.
6. Clapp S, Perry BL, Farooki ZQ, et al. Down's syndrome, complete atrioventricular canal, and pulmonary
vascular obstructive disease. J Thorac Cardiovasc Surg 1990;100:115.
7. Castaneda AR, Mayer JE Jr, Jonas RA, et al. The neonate with critical congenital heart disease: repaira
surgical challenge. J Thorac Cardiovasc Surg 1989;98:869.
8. Bando K, Turrentine MW, Sharp TG, et al. Pulmonary hypertension after operations for congenital heart
disease: analysis of risk factors and management. J Thorac Cardiovasc Surg 1996; 112:1600.
9. Friedman AH, Copel JA, Kleinman CS. Fetal echocardiography and fetal cardiology: indications, diagnosis
and management. Semin Perinatol 1993;17:76.
10. Verheijen PM, Lisowski LA, Stoutenbeek P, et al. Prenatal diagnosis of congenital heart disease affects
preoperative acidosis in the newborn patient. J Thorac Cardiovasc Surg 2001;121:798.
11. Eapen RS, Rowland DG, Franklin WH. Effect of prenatal diagnosis of critical left heart obstruction on
perinatal morbidity and mortality. Am J Perinatol 1998;15:237.
12. Bonnet D, Coltri A, Butera G, et al. Detection of transposition of the great arteries in fetuses reduces
neonatal morbidity and mortality. Circulation 1999;99:916.
13. Kumar RK, Newburger JW, Gauvreau K, et al. Comparison of outcome when hypoplastic left heart
syndrome and transposition of the great arteries are diagnosed prenatally versus when diagnosis of these two
conditions is made only postnatally. Am J Cardiol 1999;83:1649.
14. Tworetzky W, McElhinney DB, Reddy VM, et al. Improved surgical outcome after fetal diagnosis of
hypoplastic left heart syndrome. Circulation 2001;103:1269.
P.759
15. Kern JH, Hayes CJ, Michler RE, et al. Survival and risk factor analysis for the Norwood procedure for
hypoplastic left heart syndrome. Am J Cardiol 1997;80:170.
16. Mahle WT, Clancy RR, McGaurn SP, et al. Impact of prenatal diagnosis on survival and early neurologic
morbidity in neonates with the hypoplastic left heart syndrome. Pediatrics 2001;107:1277.
17. Snider A, Serwer G, Ritter S. Specialized echocardiographic techniques. In: Snider A, Serwer G, Ritter S,
eds. Echocardiography in pediatric heart disease, 2nd ed. St. Louis: Mosby-Year Book, 1997:76.
18. Sharland GK, Chan KY, Allan LD. Coarctation of the aorta: difficulties in prenatal diagnosis. Br Heart J
1994;71:70.
19. Rudolph A. The fetal circulation and postnatal adaptation. In: Rudolph A, ed. Congenital diseases of the
heart: clinical-pathophysiological considerations, 2nd ed. Armonk, NY: Futura, 2001:3.
20. Edelstone DI, Rudolph AM. Preferential streaming of ductus venosus blood to the brain and heart in fetal
lambs. Am J Physiol 1979;237:H724.
21. Edelstone DI. Regulation of blood flow through the ductus venosus. J Dev Physiol 1980;2:219.
22. Rudolph AM. Fetal and neonatal pulmonary circulation. Am Rev Respir Dis 1977;115:11.
23. Teitel DF, Iwamoto HS, Rudolph AM. Changes in the pulmonary circulation during birth-related events.
24. Stanger P, Silverman NH, Foster E. Diagnostic accuracy of pediatric echocardiograms performed in adult
laboratories. Am J Cardiol 1999;83:908.
25. Heymann MA, Berman W Jr, Rudolph AM, et al. Dilatation of the ductus arteriosus by prostaglandin E1 in
aortic arch abnormalities. Circulation 1979;59:169.
26. Neutze JM, Starling MB, Elliott RB, et al. Palliation of cyanotic congenital heart disease in infancy with Etype prostaglandins. Circulation 1977;55:238.
27. Kramer HH, Sommer M, Rammos S, et al. Evaluation of low dose prostaglandin E1 treatment for ductus
dependent congenital heart disease. Eur J Pediatr 1995;154:700.
28. Lewis AB, Freed MD, Heymann MA, et al. Side effects of therapy with prostaglandin E1 in infants with
critical congenital heart disease. Circulation 1981;64:893.
29. Hallidie-Smith KA. Prostaglandin E1 in suspected ductus dependent cardiac malformation. Arch Dis Child
1984;59:1020.
30. Lim DS, Kulik TJ, Kim DW, et al. Aminophylline for the prevention of apnea during prostaglandin E1
infusion. Pediatrics 2003; 112:e27.
31. Tworetzky W, McElhinney DB, Brook MM, et al. Echocardiographic diagnosis alone for the complete repair
of major congenital heart defects. J Am Coll Cardiol 1999;33:228.
32. Kaulitz R, Jonas RA, van der Velde ME. Echocardiographic assessment of interrupted aortic arch. Cardiol
Young 1999;9:562.
33. Need LR, Powell AJ, del Nido P, et al. Coronary echocardiography in tetralogy of Fallot: diagnostic
accuracy, resource utilization and surgical implications over 13 years. J Am Coll Cardiol 2000;36:1371.
34. Rashkind WJ, Miller WW. Creation of an atrial septal defect without thoracotomy. A palliative approach to
complete transposition of the great arteries. JAMA 1966;196:991.
35. Atz AM, Feinstein JA, Jonas RA, et al. Preoperative management of pulmonary venous hypertension in
hypoplastic left heart syndrome with restrictive atrial septal defect. Am J Cardiol 1999; 83:1224.
36. Chung T. Assessment of cardiovascular anatomy in patients with congenital heart disease by magnetic
resonance imaging. Pediatr Cardiol 2000;21:18.
37. Lardo AC. Real-time magnetic resonance imaging: diagnostic and interventional applications. Pediatr
Cardiol 2000;21:80.
38. Khoury MJ, Cordero JF, Mulinare J, et al. Selected midline defect associations: a population study.
39. Murugasu B, Yip WC, Tay JS, et al. Sonographic screening for renal tract anomalies associated with
congenital heart disease. J Clin Ultrasound 1990;18:79.
40. Towbin J, Greenberg F. Genetic syndromes and clinical molecular genetics. In: Garson A, Bricker J, Fisher
D, et al, eds. The science and practice of pediatric cardiology. Baltimore, MD: Williams & Wilkins, 1998:2627.
41. Goldmuntz E, Clark BJ, Mitchell LE, et al. Frequency of 22q11 deletions in patients with conotruncal
defects. J Am Coll Cardiol 1998;32:492.
42. Frohn-Mulder IM, Wesby Swaay E, Bouwhuis C, et al. Chromosome 22q11 deletions in patients with
selected outflow tract malformations. Genet Couns 1999;10:35.
43. Chang AC, Hanley FL, Lock JE, et al. Management and outcome of low birth weight neonates with
congenital heart disease. J Pediatr 1994;124:461.
44. Reddy VM, McElhinney DB, Sagrado T, et al. Results of 102 cases of complete repair of congenital heart
defects in patients weighing 700 to 2500 grams. J Thorac Cardiovasc Surg 1999;117:324.
45. Reddy VM, Hanley FL. Cardiac surgery in infants with very low birth weight. Semin Pediatr Surg 2000;9:91.
46. Rossi AF, Seiden HS, Sadeghi AM, et al. The outcome of cardiac operations in infants weighing two
kilograms or less. J Thorac Cardiovasc Surg 1998;116:28.
47. Bacha EA, Almodovar M, Wessel DL, et al. Surgery for coarctation of the aorta in infants weighing less
than 2 kg. Ann Thorac Surg 2001;71:1260.
48. Weinstein S, Gaynor JW, Bridges ND, et al. Early survival of infants weighing 2.5 kilograms or less
undergoing first-stage reconstruction for hypoplastic left heart syndrome. Circulation 1999;100:II167.
49. Pizarro C, Davis DA, Galantowicz ME, et al. Stage I palliation for hypoplastic left heart syndrome in low
birth weight neonates: can we justify it? Eur J Cardiothorac Surg 2002;21:716.
50. McElhinney DB, Hedrick HL, Bush DM, et al. Necrotizing enterocolitis in neonates with congenital heart
disease: risk factors and outcomes. Pediatrics 2000;106:1080.
51. Eronen M, Siren MK, Ekblad H, et al. Short- and long-term outcome of children with congenital complete
heart block diagnosed in utero or as a newborn. Pediatrics 2000;106:86.
52. Gregoratos G, Abrams J, Epstein AE, et al. ACC/AHA/NASPE 2002 Guideline Update for Implantation of
Cardiac Pacemakers and Antiarrhythmia Devicessummary article: a report of the American College of
Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2002;40:1703.
53. Kovalchin JP, Forbes TJ, Nihill MR, et al. Echocardiographic determinants of clinical course in infants with
critical and severe pulmonary valve stenosis. J Am Coll Cardiol 1997;29:1095.
54. Gournay V, Piechaud JF, Delogu A, et al. Balloon valvotomy for critical stenosis or atresia of pulmonary
valve in newborns. J Am Coll Cardiol 1995;26:1725.
55. Colli AM, Perry SB, Lock JE, et al. Balloon dilation of critical valvar pulmonary stenosis in the first month of
life. Cathet Cardiovasc Diagn 1995;34:23.
56. Gildein HP, Kleinert S, Goh TH, et al. Treatment of critical pulmonary valve stenosis by balloon dilatation
in the neonate. Am Heart J 1996;131:1007.
57. Tabatabaei H, Boutin C, Nykanen DG, et al. Morphologic and hemodynamic consequences after
percutaneous balloon valvotomy for neonatal pulmonary stenosis: medium-term follow-up. J Am Coll Cardiol
1996;27:473.
58. Weber HS. Initial and late results after catheter intervention for neonatal critical pulmonary valve stenosis
and atresia with intact ventricular septum: a technique in continual evolution. Catheter Cardiovasc Interv
2002;56:394.
59. Hanley FL, Sade RM, Freedom RM, et al. Outcomes in critically ill neonates with pulmonary stenosis and
intact ventricular septum: a multiinstitutional study. Congenital Heart Surgeons Society. J Am Coll Cardiol
1993;22:183.
60. Fedderly RT, Lloyd TR, Mendelsohn AM, et al. Determinants of successful balloon valvotomy in infants with
critical pulmonary stenosis or membranous pulmonary atresia with intact ventricular septum. J Am Coll Cardiol
1995;25:460.
61. Fraser CD Jr, McKenzie ED, Cooley DA. Tetralogy of Fallot: surgical management individualized to the
patient. Ann Thorac Surg 2001;71:1556.
62. Pigula FA, Khalil PN, Mayer JE, et al. Repair of tetralogy of Fallot in neonates and young infants.
Circulation 1999;100:II157.
63. Johnson MC, Strauss AW, Dowton SB, et al. Deletion within chromosome 22 is common in patients with
absent pulmonary valve syndrome. Am J Cardiol 1995;76:66.
64. Lakier JB, Stanger P, Heymann MA, et al. Tetralogy of Fallot with absent pulmonary valve. Natural history
and hemodynamic considerations. Circulation 1974;50:167.
65. Heinemann MK, Hanley FL. Preoperative management of neonatal tetralogy of Fallot with absent
pulmonary valve syndrome. Ann Thorac Surg 1993;55:172.
P.760
66. Bull K, Somerville J, Ty E, et al. Presentation and attrition in complex pulmonary atresia. J Am Coll Cardiol
1995;25:491.
67. Minich LL, Tani LY, Ritter S, et al. Usefulness of the preoperative tricuspid/mitral valve ratio for predicting
outcome in pulmonary atresia with intact ventricular septum. Am J Cardiol 2000; 85:1325.
68. Satou GM, Perry SB, Gauvreau K, et al. Echocardiographic predictors of coronary artery pathology in
pulmonary atresia with intact ventricular septum. Am J Cardiol 2000;85:1319.
69. Daubeney PE, Delany DJ, Anderson RH, et al. Pulmonary atresia with intact ventricular septum: range of
morphology in a population-based study. J Am Coll Cardiol 2002;39:1670.
70. Jahangiri M, Zurakowski D, Bichell D, et al. Improved results with selective management in pulmonary
atresia with intact ventricular septum. J Thorac Cardiovasc Surg 1999;118:1046.
71. Alwi M, Geetha K, Bilkis AA, et al. Pulmonary atresia with intact ventricular septum percutaneous
radiofrequency-assisted valvotomy and balloon dilation versus surgical valvotomy and Blalock Taussig shunt. J
Am Coll Cardiol 2000;35:468.
72. Celermajer DS, Cullen S, Sullivan ID, et al. Outcome in neonates with Ebstein's anomaly. J Am Coll Cardiol
1992;19:1041.
73. Knott-Craig CJ, Overholt ED, Ward KE, et al. Repair of Ebstein's anomaly in the symptomatic neonate: an
evolution of technique with 7-year follow-up. Ann Thorac Surg 2002;73:1786.
74. Starnes VA, Pitlick PT, Bernstein D, et al. Ebstein's anomaly appearing in the neonate. A new surgical
approach. J Thorac Cardiovasc Surg 1991;101:1082.
75. Stellin G, Santini F, Thiene G, et al. Pulmonary atresia, intact ventricular septum, and Ebstein anomaly of
the tricuspid valve. Anatomic and surgical considerations. J Thorac Cardiovasc Surg 1993;106:255.
76. Leoni F, Huhta JC, Douglas J, et al. Effect of prostaglandin on early surgical mortality in obstructive lesions
of the systemic circulation. Br Heart J 1984;52:654.
77. Lofland GK, McCrindle BW, Williams WG, et al. Critical aortic stenosis in the neonate: a multi-institutional
study of management, outcomes, and risk factors. Congenital Heart Surgeons Society. J Thorac Cardiovasc
Surg 2001;121:10.
78. Cowley CG, Dietrich M, Mosca RS, et al. Balloon valvuloplasty versus transventricular dilation for neonatal
critical aortic stenosis. Am J Cardiol 2001;87:1125.
79. Gildein HP, Kleinert S, Weintraub RG, et al. Surgical commissurotomy of the aortic valve: outcome of open
valvotomy in neonates with critical aortic stenosis. Am Heart J 1996;131:754.
80. McCrindle BW, Blackstone EH, Williams WG, et al. Are outcomes of surgical versus transcatheter balloon
valvotomy equivalent in neonatal critical aortic stenosis? Circulation 2001;104:I152.
81. Barnea O, Austin EH, Richman B, et al. Balancing the circulation: theoretic optimization of pulmonary/
systemic flow ratio in hypoplastic left heart syndrome. J Am Coll Cardiol 1994;24:1376.
82. Barnea O, Santamore WP, Rossi A, et al. Estimation of oxygen delivery in newborns with a univentricular
circulation. Circulation 1998;98:1407.
83. Day RW, Tani LY, Minich LL, et al. Congenital heart disease with ductal-dependent systemic perfusion:
Doppler ultrasonography flow velocities are altered by changes in the fraction of inspired oxygen. J Heart Lung
Transplant 1995;14:718.
84. Day RW, Barton AJ, Pysher TJ, et al. Pulmonary vascular resistance of children treated with nitrogen
during early infancy. Ann Thorac Surg 1998;65:1400.
85. Mora GA, Pizarro C, Jacobs ML, et al. Experimental model of single ventricle. Influence of carbon dioxide
on pulmonary vascular dynamics. Circulation 1994;90:II43.
86. Tabbutt S, Ramamoorthy C, Montenegro LM, et al. Impact of inspired gas mixtures on preoperative infants
with hypoplastic left heart syndrome during controlled ventilation. Circulation 2001;104:I159.
87. Ramamoorthy C, Tabbutt S, Kurth CD, et al. Effects of inspired hypoxic and hypercapnic gas mixtures on
cerebral oxygen saturation in neonates with univentricular heart defects. Anesthesiology 2002;96:283.
88. Rychik J, Rome JJ, Collins MH, et al. The hypoplastic left heart syndrome with intact atrial septum: atrial
morphology, pulmonary vascular histopathology and outcome. J Am Coll Cardiol 1999;34:554.
89. Graziano JN, Heidelberger KP, Ensing GJ, et al. The influence of a restrictive atrial septal defect on
pulmonary vascular morphology in patients with hypoplastic left heart syndrome. Pediatr Cardiol 2002;23:146.
90. Canter CE, Moorehead S, Huddleston CB, et al. Restrictive atrial septal communication as a determinant of
outcome of cardiac transplantation for hypoplastic left heart syndrome. Circulation 1993;88:II456.
91. Vlahos AP, Lock JE, McElhinney DB, et al. Hypoplastic left heart syndrome with intact or highly restrictive
atrial septum: outcome after neonatal transcatheter atrial septostomy. Circulation 2004; 109:2326.
92. Marshall AC, van der Velde ME, Tworetzky W, et al. Creation of an atrial septal defect in utero for fetuses
with hypoplastic left heart syndrome and intact or highly restrictive atrial septum. Circulation 2004;110:253.
93. Barber G, Helton JG, Aglira BA, et al. The significance of tricuspid regurgitation in hypoplastic left-heart
syndrome. Am Heart J 1988;116:1563.
94. Pigott JD, Murphy JD, Barber G, et al. Palliative reconstructive surgery for hypoplastic left heart syndrome.
Ann Thorac Surg 1988;45:122.
95. Breymann T, Kirchner G, Blanz U, et al. Results after Norwood procedure and subsequent cavopulmonary
anastomoses for typical hypoplastic left heart syndrome and similar complex cardiovascular malformations.
Eur J Cardiothorac Surg 1999; 16:117.
96. Gutgesell HP, Gibson J. Management of hypoplastic left heart syndrome in the 1990s. Am J Cardiol
2002;89:842.
97. Razzouk AJ, Chinnock RE, Gundry SR, et al. Transplantation as a primary treatment for hypoplastic left
heart syndrome: intermediate-term results. Ann Thorac Surg 1996;62:1.
98. Morrow WR, Naftel D, Chinnock R, et al. Outcome of listing for heart transplantation in infants younger
than six months: predictors of death and interval to transplantation. J Heart Lung Transplant 1997;16:1255.
99. Ikle L, Hale K, Fashaw L, et al. Developmental outcome of patients with hypoplastic left heart syndrome
treated with heart transplantation. J Pediatr 2003;142:20.
100. Bauer J, Thul J, Kramer U, et al. Heart transplantation in children and infants: short-term outcome and
long-term follow-up. Pediatr Transplant 2001;5:457.
101. Sell JE, Jonas RA, Mayer JE, et al. The results of a surgical program for interrupted aortic arch. J Thorac
102. Backer CL, Mavroudis C. Congenital Heart Surgery Nomenclature and Database Project: patent ductus
arteriosus, coarctation of the aorta, interrupted aortic arch. Ann Thorac Surg 2000;69:S298.
103. Ashfaq M, Houston AB, Gnanapragasam JP, et al. Balloon atrial septostomy under echocardiographic
control: six years' experience and evaluation of the practicability of cannulation via the umbilical vein. Br
Heart J 1991;65:148.
104. Tandon R, Edwards JE. Tricuspid atresia. A re-evaluation and classification. J Thorac Cardiovasc Surg
1974;67:530.
105. Epstein M. Tricuspid atresia. In: Allen H, Gutgesell H, Clark E, et al, eds. Moss and Adam's heart disease
in infants, children, and adolescents. Philadelphia: Lippincott, Williams & Wilkins, 2001:799.
106. Lupinetti FM, Kulik TJ, Beekman RH 3rd, et al. Correction of total anomalous pulmonary venous
connection in infancy. J Thorac Cardiovasc Surg 1993;106:880.
107. McElhinney DB, Driscoll DA, Emanuel BS, et al. Chromosome 22q11 deletion in patients with truncus
arteriosus. Pediatr Cardiol 2003;24:569.
108. Jacobs ML. Congenital heart surgery nomenclature and database project: truncus arteriosus. Ann Thorac
Surg 2000;69:S50.
109. Bove EL, Lupinetti FM, Pridjian AK, et al. Results of a policy of primary repair of truncus arteriosus in the
neonate. J Thorac Cardiovasc Surg 1993;105:1057.
110. Thompson LD, McElhinney DB, Reddy M, et al. Neonatal repair of truncus arteriosus: continuing
improvement in outcomes. Ann Thorac Surg 2001;72:391.
111. Du ZD, Hijazi ZM, Kleinman CS, et al. Comparison between transcatheter and surgical closure of
secundum atrial septal defect in children and adults: results of a multicenter nonrandomized trial. J Am Coll
Cardiol 2002;39:1836.
112. Lun K, Li H, Leung MP, et al. Analysis of indications for surgical closure of subarterial ventricular septal
defect without associated aortic cusp prolapse and aortic regurgitation. Am J Cardiol 2001;87:1266.
P.761
113. Puchalski MD, Brook MM, Silverman NH. Simplified echocardiographic criteria for decision making in
perimembranous ventricular septal defect in childhood. Am J Cardiol 2002; 90:569.
114. Ross R. Medical management of chronic heart failure in children. Am J Cardiovasc Drugs 2001;1:37.
115. Kumar K, Lock JE, Geva T. Apical muscular ventricular septal defects between the left ventricle and the
right ventricular infundibulum. Diagnostic and interventional considerations. Circulation 1997;95:1207.
116. McElhinney DB, Reddy VM, Silverman NH, et al. Atrioventricular septal defect with common valvar orifice
and tetralogy of Fallot revisited: making a case for primary repair in infancy. Cardiol Young 1998;8:455.
117. Van Overmeire B, Smets K, Lecoutere D, et al. A comparison of ibuprofen and indomethacin for closure
of patent ductus arteriosus. N Engl J Med 2000;343:674.
118. McElhinney DB, Reddy VM, Tworetzky W, et al. Early and late results after repair of aortopulmonary
septal defect and associated anomalies in infants <6 months of age. Am J Cardiol 1998; 81:195.
119. Backer CL, Mavroudis C. Surgical management of aortopulmonary window: a 40-year experience. Eur J
Cardiothorac Surg 2002;21:773.
120. Johnsrude CL, Perry JC, Cecchin F, et al. Differentiating anomalous left main coronary artery originating
from the pulmonary artery in infants from myocarditis and dilated cardiomyopathy by electrocardiogram. Am J
Cardiol 1995;75:71.
121. Chang RR, Allada V. Electrocardiographic and echocardiographic features that distinguish anomalous
origin of the left coronary artery from pulmonary artery from idiopathic dilated cardiomyopathy. Pediatr
Cardiol 2001;22:3.
122. Schwartz ML, Jonas RA, Colan SD. Anomalous origin of left coronary artery from pulmonary artery:
recovery of left ventricular function after dual coronary repair. J Am Coll Cardiol 1997; 30:547.
123. Lee KJ, McCrindle BW, Bohn DJ, et al. Clinical outcomes of acute myocarditis in childhood. Heart
1999;82:226.
124. Boucek MM, Edwards LB, Keck BM, et al. The Registry of the International Society for Heart and Lung
Transplantation: fifth Official Pediatric Report-2001 to 2002. J Heart Lung Transplant 2002;21:827.
125. West LJ, Pollock-Barziv SM, Dipchand AI, et al. ABO-incompatible heart transplantation in infants. N Engl
J Med 2001;344:793.
126. Gajarski R, Mosca R, Ohye R, et al. Use of extracorporeal life support as a bridge to pediatric cardiac
transplantation. J Heart Lung Transplant 2003;22:28.
127. Bourke KD, Sondheimer HM, Ivy DD, et al. Improved pretransplant management of infants with
hypoplastic left heart syndrome enables discharge to home while waiting for transplantation. Pediatr Cardiol
2003;24:538.
128. Mitchell MB, Campbell DN, Boucek MM, et al. Mechanical limitation of pulmonary blood flow facilitates
heart transplantation in older infants with hypoplastic left heart syndrome. Eur J Cardiothorac Surg
2003;23:735.
129. Mayer J. Cardiopulmonary bypass. In: Chang A, Hanley F, Wernovsky G, et al, eds. Pediatric cardiac
intensive care. Baltimore, MD: Williams & Wilkins, 1998.
130. Pigula FA, Nemoto EM, Griffith BP, et al. Regional low-flow perfusion provides cerebral circulatory support
during neonatal aortic arch reconstruction. J Thorac Cardiovasc Surg 2000;119:331.
131. Lehot JJ, Villard J, Piriz H, et al. Hemodynamic and hormonal responses to hypothermic and
normothermic cardiopulmonary bypass. J Cardiothorac Vasc Anes 1992;6:132.
132. Ationu A, Singer DR, Smith A, et al. Studies of cardiopulmonary bypass in children: implications for the
regulation of brain natriuretic peptide. Cardiovasc Res 1993;27:1538.
133. Stewart JM, Gewitz MH, Clark BJ, et al. The role of vasopressin and atrial natriuretic factor in
postoperative fluid retention after the Fontan procedure. J Thorac Cardiovasc Surg 1991;102:821.
134. Burch M, Lum L, Elliott M, et al. Influence of cardiopulmonary bypass on water balance hormones in
children. Br Heart J 1992; 68:309.
135. Philbin DM, Levine FH, Emerson CW, et al. Plasma vasopressin levels and urinary flow during
cardiopulmonary bypass in patients with valvular heart disease: effect of pulsatile flow. J Thorac Cardiovasc
Surg 1979;78:779.
136. Kross J, Dries DJ, Kumar P, et al. Atrial natriuretic peptide may not play a role in diuresis and natriuresis
after cardiac operations. J Thorac Cardiovasc Surg 1992;103:1168.
137. Kirshbom PM, Tsui SS, DiBernardo LR, et al. Blockade of endothelin-converting enzyme reduces
pulmonary hypertension after cardiopulmonary bypass and circulatory arrest. Surgery 1995;118:440.
138. Townsend GE, Wynands JE, Whalley DG, et al. Role of renin-angiotensin system in cardiopulmonary
bypass hypertension. Can Anaesth Soc J 1984;31:160.
139. Bailey DR, Miller ED Jr, Kaplan JA, et al. The reninangiotensinaldosterone system during cardiac
surgery with morphinenitrous oxide anesthesia. Anesthesiology 1975;42:538.
140. Wernovsky G, Wypij D, Jonas RA, et al. Postoperative course and hemodynamic profile after the arterial
switch operation in neonates and infants. A comparison of low-flow cardiopulmonary bypass and circulatory
arrest. Circulation 1995;92:2226.
141. Wessel DL, Adatia I, Giglia TM, et al. Use of inhaled nitric oxide and acetylcholine in the evaluation of
pulmonary hypertension and endothelial function after cardiopulmonary bypass. Circulation 1993;88:2128.
142. Schulze-Neick I, Li J, Penny DJ, et al. Pulmonary vascular resistance after cardiopulmonary bypass in
infants: effect on postoperative recovery. J Thorac Cardiovasc Surg 2001;121:1033.
143. Butler J, Rocker GM, Westaby S. Inflammatory response to cardiopulmonary bypass. Ann Thorac Surg
1993;55:552.
144. Mainwaring RD, Lamberti JJ, Hugli TE. Complement activation and cytokine generation after modified
Fontan procedure. Ann Thorac Surg 1998;65:1715.
145. Kirklin JK, Westaby S, Blackstone EH, et al. Complement and the damaging effects of cardiopulmonary
bypass. J Thorac Cardiovasc Surg 1983;86:845.
146. Tanaka K, Takao M, Yada I, et al. Alterations in coagulation and fibrinolysis associated with
cardiopulmonary bypass during open heart surgery. J Cardiothorac Anesth 1989;3:181.
147. Riegel W, Spillner G, Schlosser V, et al. Plasma levels of main granulocyte components during
cardiopulmonary bypass. J Thorac Cardiovasc Surg 1988;95:1014.
148. Kondo C, Tanaka K, Takagi K, et al. Platelet dysfunction during cardiopulmonary bypass surgery. With
special reference to platelet membrane glycoproteins. ASAIO J 1993;39:M550.
149. Butler J, Parker D, Pillai R, et al. Effect of cardiopulmonary bypass on systemic release of neutrophil
elastase and tumor necrosis factor. J Thorac Cardiovasc Surg 1993;105:25.
150. Cavarocchi NC, England MD, Schaff HV, et al. Oxygen free radical generation during cardiopulmonary
bypass: correlation with complement activation. Circulation 1986;74:III130.
151. Cavarocchi NC, Pluth JR, Schaff HV, et al. Complement activation during cardiopulmonary bypass.
Comparison of bubble and membrane oxygenators. J Thorac Cardiovasc Surg 1986; 91:252.
152. Kern FH, Morana NJ, Sears JJ, et al. Coagulation defects in neonates during cardiopulmonary bypass.
Ann Thorac Surg 1992; 54:541.
153. Chaturvedi RR, Lincoln C, Gothard JW, et al. Left ventricular dysfunction after open repair of simple
congenital heart defects in infants and children: quantitation with the use of a conductance catheter
immediately after bypass. J Thorac Cardiovasc Surg 1998;115:77.
154. England MD, Cavarocchi NC, O'Brien JF, et al. Influence of antioxidants (mannitol and allopurinol) on
oxygen free radical generation during and after cardiopulmonary bypass. Circulation 1986;74:III134.
155. Bronicki RA, Backer CL, Baden HP, et al. Dexamethasone reduces the inflammatory response to
cardiopulmonary bypass in children. Ann Thorac Surg 2000;69:1490.
156. Schroeder VA, Pearl JM, Schwartz SM, et al. Combined steroid treatment for congenital heart surgery
improves oxygen delivery and reduces postbypass inflammatory mediator expression. Circulation
2003;107:2823.
157. Mojcik CF, Levy JH. Aprotinin and the systemic inflammatory response after cardiopulmonary bypass.
Ann Thorac Surg 2001; 71:745.
158. Dietrich W, Mossinger H, Spannagl M, et al. Hemostatic activation during cardiopulmonary bypass with
different aprotinin dosages in pediatric patients having cardiac operations. J Thorac Cardiovasc Surg
1993;105:712.
P.762
159. Carrel TP, Schwanda M, Vogt PR, et al. Aprotinin in pediatric cardiac operations: a benefit in complex
malformations and with high-dose regimen only. Ann Thorac Surg 1998;66:153.
160. D'Errico CC, Shayevitz JR, Martindale SJ, et al. The efficacy and cost of aprotinin in children undergoing
reoperative open heart surgery. Anesth Analg 1996;83:1193.
161. Penkoske PA, Entwistle LM, Marchak BE, et al. Aprotinin in children undergoing repair of congenital heart
defects. Ann Thorac Surg 1995;60:S529.
162. Miller BE, Tosone SR, Tam VK, et al. Hematologic and economic impact of aprotinin in reoperative
pediatric cardiac operations. Ann Thorac Surg 1998;66:535.
163. Jaquiss RD, Ghanayem NS, Zacharisen MC, et al. Safety of aprotinin use and re-use in pediatric
cardiothoracic surgery. Circulation 2002;106:I90.
164. Costello JM, Backer CL, de Hoyos A, et al. Aprotinin reduces operative closure time and blood product
use after pediatric bypass. Ann Thorac Surg 2003;75:1261.
165. Mossinger H, Dietrich W, Braun SL, et al. High-dose aprotinin reduces activation of hemostasis,
allogeneic blood requirement, and duration of postoperative ventilation in pediatric cardiac surgery. Ann
166. Tassani P, Richter JA, Barankay A, et al. Does high-dose methylprednisolone in aprotinin-treated patients
attenuate the systemic inflammatory response during coronary artery bypass grafting procedures? J
Cardiothorac Vasc Anes 1999;13:165.
167. Clancy RR, McGaurn SA, Goin JE, et al. Allopurinol neurocardiac protection trial in infants undergoing
heart surgery using deep hypothermic circulatory arrest. Pediatrics 2001;108:61.
168. Schermerhorn ML, Tofukuji M, Khoury PR, et al. Sialyl lewis oligosaccharide preserves cardiopulmonary
and endothelial function after hypothermic circulatory arrest in lambs. J Thorac Cardiovasc Surg
2000;120:230.
169. Naik SK, Knight A, Elliott M. A prospective randomized study of a modified technique of ultrafiltration
during pediatric open-heart surgery. Circulation 1991;84:III422.
170. Draaisma AM, Hazekamp MG, Frank M, et al. Modified ultrafiltration after cardiopulmonary bypass in
pediatric cardiac surgery. Ann Thorac Surg 1997;64:521.
171. Bando K, Turrentine MW, Vijay P, et al. Effect of modified ultrafiltration in high-risk patients undergoing
operations for congenital heart disease. Ann Thorac Surg 1998;66:821.
172. Davies MJ, Nguyen K, Gaynor JW, et al. Modified ultrafiltration improves left ventricular systolic function
in infants after cardiopulmonary bypass. J Thorac Cardiovasc Surg 1998;115:361.
173. Chaturvedi RR, Shore DF, White PA, et al. Modified ultrafiltration improves global left ventricular systolic
function after open-heart surgery in infants and children. Eur J Cardiothorac Surg 1999;15:742.
174. Bando K, Vijay P, Turrentine MW, et al. Dilutional and modified ultrafiltration reduces pulmonary
hypertension after operations for congenital heart disease: a prospective randomized study. J Thorac
175. Duke T, Butt W, South M, et al. Early markers of major adverse events in children after cardiac
operations. J Thorac Cardiovasc Surg 1997;114:1042.
176. Newburger JW, Jonas RA, Wernovsky G, et al. A comparison of the perioperative neurologic effects of
hypothermic circulatory arrest versus low-flow cardiopulmonary bypass in infant heart surgery. N Engl J Med
1993;329:1057.
177. Stevenson JG, Sorensen GK, Gartman DM, et al. Transesophageal echocardiography during repair of
congenital cardiac defects: identification of residual problems necessitating reoperation. J Am Soc
Echocardiogr 1993;6:356.
178. Rosenfeld HM, Gentles TL, Wernovsky G, et al. Utility of intraoperative transesophageal
echocardiography in the assessment of residual cardiac defects. Pediatr Cardiol 1998;19:346.
179. Lang P, Chipman CW, Siden H, et al. Early assessment of hemodynamic status after repair of tetralogy of
Fallot: a comparison of 24 hour (intensive care unit) and 1 year postoperative data in 98 patients. Am J
Cardiol 1982;50:795.
180. Flori HR, Johnson LD, Hanley FL, et al. Transthoracic intracardiac catheters in pediatric patients
recovering from congenital heart defect surgery: associated complications and outcomes. Crit Care Med
2000;28:2997.
181. Gold JP, Jonas RA, Lang P, et al. Transthoracic intracardiac monitoring lines in pediatric surgical patients:
a ten-year experience. Ann Thorac Surg 1986;42:185.
182. Sellden H, Nilsson K, Larsson LE, et al. Radial arterial catheters in children and neonates: a prospective
study. Crit Care Med 1987; 15:1106.
183. Graves PW, Davis AL, Maggi JC, et al. Femoral artery cannulation for monitoring in critically ill children:
prospective study. Crit Care Med 1990;18:1363.
184. Kocis KC, Vermilion RP, Callow LB, et al. Complications of femoral artery cannulation for perioperative
monitoring in children. J Thorac Cardiovasc Surg 1996;112:1399.
185. Yabek SM, Akl BF, Berman W Jr, et al. Use of atrial epicardial electrodes to diagnose and treat
postoperative arrhythmias in children. Am J Cardiol 1980;46:285.
186. Li J, Schulze-Neick I, Lincoln C, et al. Oxygen consumption after cardiopulmonary bypass surgery in
children: determinants and implications. J Thorac Cardiovasc Surg 2000;119:525.
187. Tibby SM, Hatherill M, Murdoch IA. Capillary refill and core-peripheral temperature gap as indicators of
haemodynamic status in paediatric intensive care patients. Arch Dis Child 1999;80:163.
188. Quasney MW, Goodman DM, Billow M, et al. Routine chest radiographs in pediatric intensive care units.
189. McElhinney DB, Reddy VM, Parry AJ, et al. Management and outcomes of delayed sternal closure after
cardiac surgery in neonates and infants. Crit Care Med 2000;28:1180.
190. Tabbutt S, Duncan BW, McLaughlin D, et al. Delayed sternal closure after cardiac operations in a
pediatric population. J Thorac Cardiovasc Surg 1997;113:886.
191. Taeed R, Schwartz SM, Pearl JM, et al. Unrecognized pulmonary venous desaturation early after Norwood
palliation confounds Qp:Qs assessment and compromises oxygen delivery. Circulation 2001;103:2699.
192. Main E, Elliott MJ, Schindler M, et al. Effect of delayed sternal closure after cardiac surgery on respiratory
function in ventilated infants. Crit Care Med 2001;29:1798.
193. Jenkins J, Lynn A, Edmonds J, et al. Effects of mechanical ventilation on cardiopulmonary function in
children after open-heart surgery. Crit Care Med 1985;13:77.
194. Chang AC, Zucker HA, Hickey PR, et al. Pulmonary vascular resistance in infants after cardiac surgery:
role of carbon dioxide and hydrogen ion. Crit Care Med 1995;23:568.
195. Cheifetz IM, Craig DM, Quick G, et al. Increasing tidal volumes and pulmonary overdistention adversely
affect pulmonary vascular mechanics and cardiac output in a pediatric swine model. Crit Care Med
1998;26:710.
196. Kloth RL, Baum VC. Very early extubation in children after cardiac surgery. Crit Care Med 2002;30:787.
197. Seear M, Wensley D, MacNab A. Oxygen consumption-oxygen delivery relationship in children. J Pediatr
1993;123:208.
198. Lister G, Hellenbrand WE, Kleinman CS, et al. Physiologic effects of increasing hemoglobin concentration
in left-to-right shunting in infants with ventricular septal defects. N Engl J Med 1982; 306:502.
199. Siegel LB, Dalton HJ, Hertzog JH, et al. Initial postoperative serum lactate levels predict survival in
children after open heart surgery. Intensive Care Med 1996;22:1418.
200. Cheifetz IM, Kern FH, Schulman SR, et al. Serum lactates correlate with mortality after operations for
complex congenital heart disease. Ann Thorac Surg 1997;64:735.
201. Charpie JR, Dekeon MK, Goldberg CS, et al. Postoperative hemodynamics after Norwood palliation for
hypoplastic left heart syndrome. Am J Cardiol 2001;87:198.
202. Rossi AF, Seiden HS, Gross RP, et al. Oxygen transport in critically ill infants after congenital heart
operations. Ann Thorac Surg 1999;67:739.
203. Singh NC, Kissoon N, al Mofada S, et al. Comparison of continuous versus intermittent furosemide
administration in postoperative pediatric cardiac patients. Crit Care Med 1992;20:17.
204. Hickey PR, Hansen DD, Wessel DL, et al. Pulmonary and systemic hemodynamic responses to fentanyl in
infants. Anesth Analg 1985;64:483.
205. Hickey PR, Hansen DD, Wessel DL, et al. Blunting of stress responses in the pulmonary circulation of
infants by fentanyl. Anesth Analg 1985;64:1137.
206. Anand KJ, Sippell WG, Aynsley-Green A. Randomised trial of fentanyl anaesthesia in preterm babies
undergoing surgery: effects on the stress response. Lancet 1987;1:62.
207. Rigden SP, Barratt TM, Dillon MJ, et al. Acute renal failure complicating cardiopulmonary bypass surgery.
Arch Dis Child 1982; 57:425.
P.763
208. Adatia I, Atz AM, Jonas RA, et al. Diagnostic use of inhaled nitric oxide after neonatal cardiac operations.
J Thorac Cardiovasc Surg 1996;112:1403.
209. Chang AC, Vetter JM, Gill SE, et al. Accuracy of prospective two-dimensional/Doppler echocardiography
in the assessment of reparative surgery. J Am Coll Cardiol 1990;16:903.
210. Wolfe LT, Rossi A, Ritter SB. Transesophageal echocardiography in infants and children: use and
importance in the cardiac intensive care unit. J Am Soc Echocardiogr 1993;6:286.
211. Friedman WF. The intrinsic physiologic properties of the developing heart. Prog Cardiovasc Dis
1972;15:87.
212. Romero T, Covell J, Friedman WF. A comparison of pressure-volume relations of the fetal, newborn, and
adult heart. Am J Physiol 1972;222:1285.
213. Zaritsky A, Chernow B. Use of catecholamines in pediatrics. J Pediatr 1984;105:341.
214. Friedman WF, George BL. Treatment of congestive heart failure by altering loading conditions of the
heart. J Pediatr 1985;106:697.
215. Kouchoukos NT, Kirklin JW, Sheppard LC, et al. Effect of left atrial pressure by blood infusion on stroke
volume early after cardiac operations. Surg Forum 1971;22:126.
216. Feneck RO. Intravenous milrinone following cardiac surgery. II. Influence of baseline hemodynamics and
patient factors on therapeutic response. The European Milrinone Multicentre Trial Group. J Cardiothorac Vasc
Anes 1992;6:563.
217. Schulze-Neick I, Bultmann M, Werner H, et al. Right ventricular function in patients treated with inhaled
nitric oxide after cardiac surgery for congenital heart disease in newborns and children. Am J Cardiol
1997;80:360.
218. Grace MP, Greenbaum DM. Cardiac performance in response to PEEP in patients with cardiac dysfunction.
Crit Care Med 1982;10:358.
219. Pinsky MR, Summer WR, Wise RA, et al. Augmentation of cardiac function by elevation of intrathoracic
pressure. J Appl Physiol 1983;54:950.
220. Rosenzweig EB, Starc TJ, Chen JM, et al. Intravenous arginine-vasopressin in children with vasodilatory
shock after cardiac surgery. Circulation 1999;100:II182.
221. Burrows FA, Williams WG, Teoh KH, et al. Myocardial performance after repair of congenital cardiac
defects in infants and children. Response to volume loading. J Thorac Cardiovasc Surg 1988;96:548.
222. Driscoll DJ, Gillette PC, Duff DF, et al. The hemodynamic effect of dopamine in children. J Thorac
223. Driscoll DJ, Gillette PC, Duff DF, et al. Hemodynamic effects of dobutamine in children. Am J Cardiol
1979;43:581.
224. Bohn DJ, Poirier CS, Edmonds JF, et al. Hemodynamic effects of dobutamine after cardiopulmonary
bypass in children. Crit Care Med 1980;8:367.
225. Berg RA, Donnerstein RL, Padbury JF. Dobutamine infusions in stable, critically ill children:
pharmacokinetics and hemodynamic actions. Crit Care Med 1993;21:678.
226. Hoffman TM, Wernovsky G, Atz AM, et al. Efficacy and safety of milrinone in preventing low cardiac
output syndrome in infants and children after corrective surgery for congenital heart disease. Circulation
2003;107:996.
227. Chang AC, Atz AM, Wernovsky G, et al. Milrinone: systemic and pulmonary hemodynamic effects in
neonates after cardiac surgery. Crit Care Med 1995;23:1907.
228. Caspi J, Coles JG, Benson LN, et al. Age-related response to epinephrine-induced myocardial stress. A
functional and ultrastructural study. Circulation 1991;84:III394.
229. Buchhorn R, Hulpke-Wette M, Ruschewski W, et al. Beta-receptor downregulation in congenital heart

disease: a risk factor for complications after surgical repair? Ann Thorac Surg 2002;73:610.
230. Shore S, Nelson DP, Pearl JM, et al. Usefulness of corticosteroid therapy in decreasing epinephrine
requirements in critically ill infants with congenital heart disease. Am J Cardiol 2001;88:591.
231. Klein I, Ojamaa K. Thyroid hormone and the cardiovascular system. N Engl J Med 2001;344:501.
232. Mitchell IM, Pollock JC, Jamieson MP, et al. The effects of cardiopulmonary bypass on thyroid function in
infants weighing less than five kilograms. J Thorac Cardiovasc Surg 1992;103:800.
233. Mainwaring RD, Healy RM, Meier FA, et al. Reduction in levels of triiodothyronine following the first stage
of the Norwood reconstruction for hypoplastic left heart syndrome. Cardiol Young 2001;11:295.
234. Chowdhury D, Ojamaa K, Parnell VA, et al. A prospective randomized clinical study of thyroid hormone
treatment after operations for complex congenital heart disease. J Thorac Cardiovasc Surg 2001;122:1023.
235. Bettendorf M, Schmidt KG, Grulich-Henn J, et al. Tri-iodothyronine treatment in children after cardiac
surgery: a double-blind, randomised, placebo-controlled study. Lancet 2000;356:529.
236. Portman MA, Fearneyhough C, Ning XH, et al. Triiodothyronine repletion in infants during
cardiopulmonary bypass for congenital heart disease. J Thorac Cardiovasc Surg 2000;120:604.
237. Pfammatter JP, Wagner B, Berdat P, et al. Procedural factors associated with early postoperative
arrhythmias after repair of congenital heart defects. J Thorac Cardiovasc Surg 2002;123:258.
238. Humes RA, Porter CJ, Puga FJ, et al. Utility of temporary atrial epicardial electrodes in postoperative
pediatric cardiac patients. Mayo Clin Proc 1989;64:516.
239. Deal BJ, Wolff GS, Gelband H, eds. Current concepts in diagnosis and management of arrhythmias in
infants and children. Armonk, NY: Futura Publishing, 1998.
240. Rosales AM, Walsh EP, Wessel DL, et al. Postoperative ectopic atrial tachycardia in children with
congenital heart disease. Am J Cardiol 2001;88:1169.
241. Hoffman TM, Bush DM, Wernovsky G, et al. Postoperative junctional ectopic tachycardia in children:
incidence, risk factors, and treatment. Ann Thorac Surg 2002;74:1607.
242. Dodge-Khatami A, Miller OI, Anderson RH, et al. Surgical substrates of postoperative junctional ectopic
tachycardia in congenital heart defects. J Thorac Cardiovasc Surg 2002;123:624.
243. Dodge-Khatami A, Miller OI, Anderson RH, et al. Impact of junctional ectopic tachycardia on
postoperative morbidity following repair of congenital heart defects. Eur J Cardiothorac Surg 2002;21:255.
244. Laird WP, Snyder CS, Kertesz NJ, et al. Use of intravenous amiodarone for postoperative junctional
ectopic tachycardia in children. Pediatr Cardiol 2003;24:133.
245. Pfammatter JP, Paul T, Ziemer G, et al. Successful management of junctional tachycardia by
hypothermia after cardiac operations in infants. Ann Thorac Surg 1995;60:556.
246. Walsh EP, Saul JP, Sholler GF, et al. Evaluation of a staged treatment protocol for rapid automatic
junctional tachycardia after operation for congenital heart disease. J Am Coll Cardiol 1997;29:1046.
247. Perry JC, Fenrich AL, Hulse JE, et al. Pediatric use of intravenous amiodarone: efficacy and safety in
critically ill patients from a multicenter protocol. J Am Coll Cardiol 1996;27:1246.
248. Leinbach RC, Chamberlain DA, Kastor JA, et al. A comparison of the hemodynamic effects of ventricular
and sequential A-V pacing in patients with heart block. Am Heart J 1969;78:502.
249. Hopkins RA, Bull C, Haworth SG, et al. Pulmonary hypertensive crises following surgery for congenital
heart defects in young children. Eur J Cardiothorac Surg 1991;5:628.
250. Miller OI, Tang SF, Keech A, et al. Inhaled nitric oxide and prevention of pulmonary hypertension after
congenital heart surgery: a randomised double-blind study. Lancet 2000;356:1464.
251. Lindberg L, Olsson AK, Jogi P, et al. How common is severe pulmonary hypertension after pediatric
cardiac surgery? J Thorac Cardiovasc Surg 2002;123:1155.
252. Reddy VM, Hendricks-Munoz KD, Rajasinghe HA, et al. Post-cardiopulmonary bypass pulmonary
hypertension in lambs with increased pulmonary blood flow. A role for endothelin 1. Circulation 1997;95:1054.
253. Koul B, Willen H, Sjoberg T, et al. Pulmonary sequelae of prolonged total venoarterial bypass: evaluation
with a new experimental model. Ann Thorac Surg 1991;51:794.
254. Beghetti M, Silkoff PE, Caramori M, et al. Decreased exhaled nitric oxide may be a marker of
cardiopulmonary bypass-induced injury. Ann Thorac Surg 1998;66:532.
255. Schulze-Neick I, Penny DJ, Rigby ML, et al. L-arginine and substance P reverse the pulmonary
endothelial dysfunction caused by congenital heart surgery. Circulation 1999;100:749.
256. Chai PJ, Williamson JA, Lodge AJ, et al. Effects of ischemia on pulmonary dysfunction after
cardiopulmonary bypass. Ann Thorac Surg 1999;67:731.
257. Anand KJ, Hickey PR. Halothane-morphine compared with high-dose sufentanil for anesthesia and
postoperative analgesia in neonatal cardiac surgery. N Engl J Med 1992;326:1.
P.764
258. Hiramatsu T, Imai Y, Kurosawa H, et al. Effects of dilutional and modified ultrafiltration in plasma
endothelin-1 and pulmonary vascular resistance after the Fontan procedure. Ann Thorac Surg 2002;73:861.
259. Schulze-Neick I, Li J, Reader JA, et al. The endothelin antagonist BQ123 reduces pulmonary vascular
resistance after surgical intervention for congenital heart disease. J Thorac Cardiovasc Surg 2002;124:435.
260. Russell IA, Zwass MS, Fineman JR, et al. The effects of inhaled nitric oxide on postoperative pulmonary
hypertension in infants and children undergoing surgical repair of congenital heart disease. Anesth Analg
1998;87:46.
261. Segar JL, Merrill DC, Chapleau MW, et al. Hemodynamic changes during endotracheal suctioning are
mediated by increased autonomic activity. Pediatr Res 1993;33:649.
262. Wessel DL. Managing low cardiac output syndrome after congenital heart surgery. Crit Care Med
2001;29:S220.
263. Atz AM, Adatia I, Lock JE, et al. Combined effects of nitric oxide and oxygen during acute pulmonary
vasodilator testing. J Am Coll Cardiol 1999;33:813.
264. Morray JP, Lynn AM, Mansfield PB. Effect of pH and PCO2 on pulmonary and systemic hemodynamics
after surgery in children with congenital heart disease and pulmonary hypertension. J Pediatr 1988;113:474.
265. Morris K, Beghetti M, Petros A, et al. Comparison of hyperventilation and inhaled nitric oxide for
pulmonary hypertension after repair of congenital heart disease. Crit Care Med 2000;28:2974.
266. Ignarro LJ, Buga GM, Wood KS, et al. Endothelium-derived relaxing factor produced and released from
artery and vein is nitric oxide. Proc Natl Acad Sci U S A 1987;84:9265.
267. Frostell C, Fratacci MD, Wain JC, et al. Inhaled nitric oxide. A selective pulmonary vasodilator reversing
hypoxic pulmonary vasoconstriction. Circulation 1991;83:2038.
268. Rimar S, Gillis CN. Selective pulmonary vasodilation by inhaled nitric oxide is due to hemoglobin
inactivation. Circulation 1993;88:2884.
269. Miller OI, Tang SF, Keech A, et al. Rebound pulmonary hypertension on withdrawal from inhaled nitric
oxide. Lancet 1995;346: 51.
270. Atz AM, Adatia I, Wessel DL. Rebound pulmonary hypertension after inhalation of nitric oxide. Ann
271. Young JD, Sear JW, Valvini EM. Kinetics of methaemoglobin and serum nitrogen oxide production during
inhalation of nitric oxide in volunteers. Br J Anaesth 1996;76:652.
272. Atz AM, Wessel DL. Sildenafil ameliorates effects of inhaled nitric oxide withdrawal. Anesthesiology
1999;91:307.
273. Day RW, Hawkins JA, McGough EC, et al. Randomized controlled study of inhaled nitric oxide after
operation for congenital heart disease. Ann Thorac Surg 2000;69:1907.
274. Journois D, Pouard P, Mauriat P, et al. Inhaled nitric oxide as a therapy for pulmonary hypertension after
operations for congenital heart defects. J Thorac Cardiovasc Surg 1994;107:1129.
275. Miller OI, Celermajer DS, Deanfield JE, et al. Very-low-dose inhaled nitric oxide: a selective pulmonary
vasodilator after operations for congenital heart disease. J Thorac Cardiovasc Surg 1994;108:487.
276. Curran RD, Mavroudis C, Backer CL, et al. Inhaled nitric oxide for children with congenital heart disease
and pulmonary hypertension. Ann Thorac Surg 1995;60:1765.
277. Atz AM, Adatia I, Jonas RA, et al. Inhaled nitric oxide in children with pulmonary hypertension and
congenital mitral stenosis. Am J Cardiol 1996;77:316.
278. Zobel G, Gamillscheg A, Schwinger W, et al. Inhaled nitric oxide in infants and children after open heart
surgery. J Cardiovasc Surg 1998;39:79.
279. Gamillscheg A, Zobel G, Urlesberger B, et al. Inhaled nitric oxide in patients with critical pulmonary
perfusion after Fontan-type procedures and bidirectional Glenn anastomosis. J Thorac Cardiovasc Surg
1997;113:435.
280. Goldman AP, Delius RE, Deanfield JE, et al. Nitric oxide might reduce the need for extracorporeal support
in children with critical postoperative pulmonary hypertension. Ann Thorac Surg 1996;62:750.
281. Bush A, Busst C, Booth K, et al. Does prostacyclin enhance the selective pulmonary vasodilator effect of
oxygen in children with congenital heart disease? Circulation 1986;74:135.
282. Bush A, Busst C, Knight WB, et al. Modification of pulmonary hypertension secondary to congenital heart
disease by prostacyclin therapy. Am Rev Respir Dis 1987;136:767.
283. Goldman AP, Delius RE, Deanfield JE, et al. Nitric oxide is superior to prostacyclin for pulmonary
hypertension after cardiac operations. Ann Thorac Surg 1995;60:300.
284. Rimensberger PC, Spahr-Schopfer I, Berner M, et al. Inhaled nitric oxide versus aerosolized iloprost in
secondary pulmonary hypertension in children with congenital heart disease: vasodilator capacity and cellular
mechanisms. Circulation 2001;103:544.
285. Doctor A, Walsh B, Doorley P, et al. Inhaled prostacycline for acute pulmonary hypertension complicating
congenital heart disease. Pediatr Crit Care Med 2003;4:A148.
286. Bradley SM, Simsic JM, Mulvihill DM. Hyperventilation impairs oxygenation after bidirectional superior
cavopulmonary connection. Circulation 1998;98:II372.
287. Bradley SM, Simsic JM, Mulvihill DM. Hypoventilation improves oxygenation after bidirectional superior
cavopulmonary connection. J Thorac Cardiovasc Surg 2003;126:1033.
288. Zonis Z, Seear M, Reichert C, et al. The effect of preoperative tranexamic acid on blood loss after cardiac
operations in children. J Thorac Cardiovasc Surg 1996;111:982.
289. Reid RW, Zimmerman AA, Laussen PC, et al. The efficacy of tranexamic acid versus placebo in
decreasing blood loss in pediatric patients undergoing repeat cardiac surgery. Anesth Analg 1997;84:990.
290. Chauhan S, Kumar BA, Rao BH, et al. Efficacy of aprotinin, epsilon aminocaproic acid, or combination in
cyanotic heart disease. Ann Thorac Surg 2000;70:1308.
291. Williams GD, Bratton SL, Riley EC, et al. Efficacy of epsilon-aminocaproic acid in children undergoing
cardiac surgery. J Cardiothorac Vasc Anesth 1999;13:304.
292. Rhodes JF, Blaufox AD, Seiden HS, et al. Cardiac arrest in infants after congenital heart surgery.
293. Anonymous. Guidelines 2000 for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care.
Part 10: pediatric advanced life support. Circulation 2000;102:I291.
294. Duncan BW, Ibrahim AE, Hraska V, et al. Use of rapid-deployment extracorporeal membrane
oxygenation for the resuscitation of pediatric patients with heart disease after cardiac arrest. J Thorac
295. Duncan B, ed. Mechanical support for cardiac and respiratory failure in pediatric patients. New York:
Marcel Dekker, 2001.
296. Walters HL 3rd, Hakimi M, Rice MD, et al. Pediatric cardiac surgical ECMO: multivariate analysis of risk
factors for hospital death. Ann Thorac Surg 1995;60:329.
297. Thuys CA, Mullaly RJ, Horton SB, et al. Centrifugal ventricular assist in children under 6 kg. Eur J
Cardiothorac Surg 1998;13: 130.
298. del Nido PJ, Duncan BW, Mayer JE Jr, et al. Left ventricular assist device improves survival in children
with left ventricular dysfunction after repair of anomalous origin of the left coronary artery from the
pulmonary artery. Ann Thorac Surg 1999; 67:169.
299. Dalton HJ, Siewers RD, Fuhrman BP, et al. Extracorporeal membrane oxygenation for cardiac rescue in
children with severe myocardial dysfunction. Crit Care Med 1993;21:1020.
300. Kirshbom PM, Bridges ND, Myung RJ, et al. Use of extracorporeal membrane oxygenation in pediatric
thoracic organ transplantation. J Thorac Cardiovasc Surg 2002;123:130.
301. Kulik TJ, Moler FW, Palmisano JM, et al. Outcome-associated factors in pediatric patients treated with
extracorporeal membrane oxygenator after cardiac surgery. Circulation 1996;94:II63.
302. Pizarro C, Davis DA, Healy RM, et al. Is there a role for extracorporeal life support after stage I Norwood?
Eur J Cardiothorac Surg 2001;19:294.
303. Jaggers JJ, Forbess JM, Shah AS, et al. Extracorporeal membrane oxygenation for infant postcardiotomy
support: significance of shunt management. Ann Thorac Surg 2000;69:1476.
304. Wessel DL, Almodovar MC, Laussen PC. Intensive care management of cardiac patients on extracorporeal
membrane oxygenation. In: Duncan BW, ed. Mechanical support for cardiac and respiratory failure in pediatric
patients. New York: Marcel Dekker, 2001:75.
305. Raithel SC, Pennington DG, Boegner E, et al. Extracorporeal membrane oxygenation in children after
cardiac surgery. Circulation 1992;86:II305.
306. Black MD, Coles JG, Williams WG, et al. Determinants of success in pediatric cardiac patients undergoing
extracorporeal membrane oxygenation. Ann Thorac Surg 1995;60:133.
P.765
307. Duncan BW, Hraska V, Jonas RA, et al. Mechanical circulatory support in children with cardiac disease. J
Thorac Cardiovasc Surg 1999;117:529.
308. Marcus B, Atkinson JB, Wong PC, et al. Successful use of transesophageal echocardiography during
extracorporeal membrane oxygenation in infants after cardiac operations. J Thorac Cardiovasc Surg
1995;109:846.
309. Booth KL, Roth SJ, Perry SB, et al. Cardiac catheterization of patients supported by extracorporeal
membrane oxygenation. J Am Coll Cardiol 2002;40:1681.
310. Delius RE, Bove EL, Meliones JN, et al. Use of extracorporeal life support in patients with congenital heart
disease. Crit Care Med 1992;20:1216.
311. Ziomek S, Harrell JE Jr, Fasules JW, et al. Extracorporeal membrane oxygenation for cardiac failure after
congenital heart operation. Ann Thorac Surg 1992;54:861.
312. Duncan BW, Bohn DJ, Atz AM, et al. Mechanical circulatory support for the treatment of children with
acute fulminant myocarditis. J Thorac Cardiovasc Surg 2001;122:440.
313. Ibrahim AE, Duncan BW, Blume ED, et al. Long-term follow-up of pediatric cardiac patients requiring
mechanical circulatory support. Ann Thorac Surg 2000;69:186.
314. Hamrick SEG, Gremmels DB, Keet CA, et al. Neurodevelopmental outcome of infants supported with
extracorporeal membrane oxygenation after cardiac surgery. Pediatrics 2003;111:e671.
315. DiCarlo JV, Raphaely RC, Steven JM, et al. Pulmonary mechanics in infants after cardiac surgery. Crit
Care Med 1992;20:22.
316. McElhinney DB, Reddy VM, Pian MS, et al. Compression of the central airways by a dilated aorta in
infants and children with congenital heart disease. Ann Thorac Surg 1999;67:1130.
317. Dodge-Khatami A, Backer CL, Holinger LD, et al. Complete repair of Tetralogy of Fallot with absent
pulmonary valve including the role of airway stenting. J Card Surg 1999;14:82.
318. Mok Q, Ross-Russell R, Mulvey D, et al. Phrenic nerve injury in infants and children undergoing cardiac
surgery. Br Heart J 1991;65:287.
319. Schwartz MZ, Filler RM. Plication of the diaphragm for symptomatic phrenic nerve paralysis. J Pediatr
Surg 1978;13:259.
320. de Leeuw M, Williams JM, Freedom RM, et al. Impact of diaphragmatic paralysis after cardiothoracic
surgery in children. J Thorac Cardiovasc Surg 1999;118:510.
321. Schulze-Neick I, Ho SY, Bush A, et al. Severe airflow limitation after the unifocalization procedure:
clinical and morphological correlates. Circulation 2000;102:III142.
322. Allen EM, van Heeckeren DW, Spector ML, et al. Management of nutritional and infectious complications
of postoperative chylothorax in children. J Pediatr Surg 1991;26:1169.
323. Bond SJ, Guzzetta PC, Snyder ML, et al. Management of pediatric postoperative chylothorax. Ann Thorac
Surg 1993;56:469.
324. Pratap U, Slavik Z, Ofoe VD, et al. Octreotide to treat postoperative chylothorax after cardiac operations
in children. Ann Thorac Surg 2001;72:1740.
325. Buttiker V, Fanconi S, Burger R. Chylothorax in children: guidelines for diagnosis and management.
Chest 1999; 116:682.
326. McWilliams BC, Fan LL, Murphy SA. Transient T-cell depression in postoperative chylothorax. J Pediatr
1981;99:595.
327. Markham KM, Glover JL, Welsh RJ, et al. Octreotide in the treatment of thoracic duct injuries. Am Surg
2000;66:1165.
328. Rimensberger PC, Muller-Schenker B, Kalangos A, et al. Treatment of a persistent postoperative

chylothorax with somatostatin. Ann Thorac Surg 1998;66:253.
329. Cheung Y, Leung MP, Yip M. Octreotide for treatment of postoperative chylothorax. J Pediatr
2001;139:157.
330. Ottinger JG. Octreotide for persistent chylothorax in a pediatric patient. Ann Pharmacother
2002;36:1106.
331. Rosti L, Bini RM, Chessa M, et al. The effectiveness of octreotide in the treatment of post-operative
chylothorax. Eur J Pediatr 2002;161:149.
332. Azizkhan RG, Canfield J, Alford BA, et al. Pleuroperitoneal shunts in the management of neonatal
chylothorax. J Pediatr Surg 1983;18:842.
333. Murphy MC, Newman BM, Rodgers BM. Pleuroperitoneal shunts in the management of persistent
chylothorax. Ann Thorac Surg 1989;48:195.
334. Tarnok A, Schneider P. Pediatric cardiac surgery with cardiopulmonary bypass: pathways contributing to
transient systemic immune suppression. Shock 2001;16:24.
335. Allen ML, Peters MJ, Goldman A, et al. Early postoperative monocyte deactivation predicts systemic
inflammation and prolonged stay in pediatric cardiac intensive care. Crit Care Med 2002;30:1140.
336. Odetola FO, Moler FW, Dechert RE, et al. Nosocomial catheter-related bloodstream infections in a
pediatric intensive care unit: risk and rates associated with various intravascular technologies. Pediatr Crit
Care Med 2003;4:432.
337. O'Grady NP, Alexander M, Dellinger EP, et al. Guidelines for the prevention of intravascular catheter-
related infections. Pediatrics 2002;110:e51.
338. Mermel LA, Farr BM, Sherertz RJ, et al. Guidelines for the management of intravascular catheter-related
infections. Clin Infect Dis 2001;32:1249.
339. Maher KO, VanDerElzen K, Bove EL, et al. A retrospective review of three antibiotic prophylaxis regimens
for pediatric cardiac surgical patients. Ann Thorac Surg 2002;74:1195.
340. Giuffre RM, Tam KH, Williams WW, et al. Acute renal failure complicating pediatric cardiac surgery: a
comparison of survivors and nonsurvivors following acute peritoneal dialysis. Pediatr Cardiol 1992;13:208.
341. Boigner H, Brannath W, Hermon M, et al. Predictors of mortality at initiation of peritoneal dialysis in
children after cardiac surgery. Ann Thorac Surg 2004;77:61.
342. Limperopoulos C, Majnemer A, Shevell MI, et al. Neurologic status of newborns with congenital heart
defects before open heart surgery. Pediatrics 1999;103:402.
343. Mahle WT, Tavani F, Zimmerman RA, et al. An MRI study of neurological injury before and after
congenital heart surgery. Circulation 2002;106:I109.
344. Menache CC, du Plessis AJ, Wessel DL, et al. Current incidence of acute neurologic complications after
open-heart operations in children. Ann Thorac Surg 2002;73:1752.
345. Clancy RR, McGaurn SA, Wernovsky G, et al. Risk of seizures in survivors of newborn heart surgery using
deep hypothermic circulatory arrest. Pediatrics 2003;111:592.
346. Bellinger DC, Jonas RA, Rappaport LA, et al. Developmental and neurologic status of children after heart
surgery with hypothermic circulatory arrest or low-flow cardiopulmonary bypass. N Engl J Med 1995;332:549.
347. Bellinger DC, Rappaport LA, Wypij D, et al. Patterns of developmental dysfunction after surgery during
infancy to correct transposition of the great arteries. J Dev Behav Pediatr 1997; 18:75.
348. Rappaport LA, Wypij D, Bellinger DC, et al. Relation of seizures after cardiac surgery in early infancy to
neurodevelopmental outcome. Circulation 1998;97:773.
349. Bellinger DC, Wypij D, Kuban KC, et al. Developmental and neurological status of children at 4 years of
age after heart surgery with hypothermic circulatory arrest or low-flow cardiopulmonary bypass. Circulation
1999;100:526.
350. Kern JH, Hinton VJ, Nereo NE, et al. Early developmental outcome after the Norwood procedure for
hypoplastic left heart syndrome. Pediatrics 1998;102:1148.
351. Mahle WT, Clancy RR, Moss EM, et al. Neurodevelopmental outcome and lifestyle assessment in schoolaged and adolescent children with hypoplastic left heart syndrome. Pediatrics 2000; 105:1082.
352. Imoto Y, Kado H, Shiokawa Y, et al. Experience with the Norwood procedure without circulatory arrest. J
Thorac Cardiovasc Surg 2001;122:879.
353. Reddy VM, Hanley FL. Techniques to avoid circulatory arrest in neonates undergoing repair of complex
heart defects. Semin Thorac Cardiovasc Surg 2001;4:277.
354. Shin'oka T, Shum-Tim D, Jonas RA, et al. Higher hematocrit improves cerebral outcome after deep
hypothermic circulatory arrest. J Thorac Cardiovasc Surg 1996;112:1610.
355. du Plessis AJ, Jonas RA, Wypij D, et al. Perioperative effects of alpha-stat versus pH-stat strategies for
deep hypothermic cardiopulmonary bypass in infants. J Thorac Cardiovasc Surg 1997;114:991.
356. Jonas RA. Deep hypothermic circulatory arrest: current status and indications. Semin Thorac Cardiovasc
Surg 2002;5:3.
357. Newburger JW, Wypij D, Bellinger DC, et al. Length of stay after infant heart surgery is related to
cognitive outcome at age 8 years. J Pediatr 2003;143:67.
358. Tweddell JS, Litwin SB, Thomas JP Jr, et al. Recent advances in the surgical management of the single
ventricle pediatric patient. Pediatr Clin North Am 1999;46:465.
P.766
359. Odim J, Portzky M, Zurakowski D, et al. Sternotomy approach for the modified Blalock-Taussig shunt.
360. Bradley SM, Mosca RS, Hennein HA, et al. Bidirectional superior cavopulmonary connection in young
infants. Circulation 1996; 94:II5.
361. Reddy VM, McElhinney DB, Moore P, et al. Outcomes after bidirectional cavopulmonary shunt in infants
less than 6 months old. J Am Coll Cardiol 1997;29:1365.
362. Chang AC, Hanley FL, Wernovsky G, et al. Early bidirectional cavopulmonary shunt in young infants.
Postoperative course and early results. Circulation 1993;88:II149.
363. Santamore WP, Barnea O, Riordan CJ, et al. Theoretical optimization of pulmonary-to-systemic flow ratio
after a bidirectional cavopulmonary anastomosis. Am J Physiol 1998;274:H694.
364. Zellers TM, Driscoll DJ, Humes RA, et al. Glenn shunt: effect on pleural drainage after modified Fontan
operation. J Thorac Cardiovasc Surg 1989;98:725.
365. Galal O, Kalloghlian A, Pittappilly BM, et al. Phentolamine improves clinical outcome after balloon
valvoplasty in neonates with severe pulmonary stenosis. Cardiol Young 1999;9:127.
366. Hew CC, Daebritz SH, Zurakowski D, et al. Valved homograft replacement of aneurysmal pulmonary
arteries for severely symptomatic absent pulmonary valve syndrome. Ann Thorac Surg 2002;73:1778.
367. Hraska V, Kantorova A, Kunovsky P, et al. Intermediate results with correction of tetralogy of Fallot with
absent pulmonary valve using a new approach. Eur J Cardiothorac Surg 2002;21:711.
368. Watterson KG, Malm TK, Karl TR, et al. Absent pulmonary valve syndrome: operation in infants with
airway obstruction. Ann Thorac Surg 1992;54:1116.
369. Puga FJ, Leoni FE, Julsrud PR, et al. Complete repair of pulmonary atresia, ventricular septal defect, and
severe peripheral arborization abnormalities of the central pulmonary arteries. Experience with preliminary
unifocalization procedures in 38 patients. J Thorac Cardiovasc Surg 1989;98:1018.
370. Reddy VM, McElhinney DB, Amin Z, et al. Early and intermediate outcomes after repair of pulmonary
atresia with ventricular septal defect and major aortopulmonary collateral arteries: experience with 85
patients. Circulation 2000;101:1826.
371. Rodefeld MD, Reddy VM, Thompson LD, et al. Surgical creation of aortopulmonary window in selected
patients with pulmonary atresia with poorly developed aortopulmonary collaterals and hypoplastic pulmonary
arteries. J Thorac Cardiovasc Surg 2002; 123:1147.
372. Marshall AC, Love BA, Lang P, et al. Staged repair of tetralogy of Fallot and diminutive pulmonary
arteries with a fenestrated ventricular septal defect patch. J Thorac Cardiovasc Surg 2003;126:1427.
373. Rhodes LA, Colan SD, Perry SB, et al. Predictors of survival in neonates with critical aortic stenosis.
Circulation 1991;84:2325.
374. Dodge-Khatami A, Backer CL, Mavroudis C. Risk factors for recoarctation and results of reoperation: a 40year review. J Card Surg 2000;15:369.
375. Backer CL, Mavroudis C, Zias EA, et al. Repair of coarctation with resection and extended end-to-end
anastomosis. Ann Thorac Surg 1998;66:1365.
376. Rao PS, Galal O, Smith PA, et al. Five- to nine-year follow-up results of balloon angioplasty of native
aortic coarctation in infants and children. J Am Coll Cardiol 1996;27:462.
377. Johnson MC, Canter CE, Strauss AW, et al. Repair of coarctation of the aorta in infancy: comparison of
surgical and balloon angioplasty. Am Heart J 1993;125:464.
378. Norwood WI, Lang P, Hansen DD. Physiologic repair of aortic atresia-hypoplastic left heart syndrome. N
Engl J Med 1983; 308:23.
379. Rossi AF, Sommer RJ, Lotvin A, et al. Usefulness of intermittent monitoring of mixed venous oxygen
saturation after stage I palliation for hypoplastic left heart syndrome. Am J Cardiol 1994; 73:1118.
380. Riordan CJ, Randsbeck F, Storey JH, et al. Effects of oxygen, positive end-expiratory pressure, and
carbon dioxide on oxygen delivery in an animal model of the univentricular heart. J Thorac Cardiovasc Surg
1996;112:644.
381. Riordan CJ, Locher JP Jr, Santamore WP, et al. Monitoring systemic venous oxygen saturations in the
hypoplastic left heart syndrome. Ann Thorac Surg 1997;63:835.
382. Tweddell JS, Hoffman GM, Fedderly RT, et al. Phenoxybenzamine improves systemic oxygen delivery
after the Norwood procedure. Ann Thorac Surg 1999;67:161.
383. Rychik J, Bush DM, Spray TL, et al. Assessment of pulmonary/systemic blood flow ratio after first-stage
palliation for hypoplastic left heart syndrome: development of a new index with the use of Doppler
echocardiography. J Thorac Cardiovasc Surg 2000;120:81.
384. Reddy VM, Liddicoat JR, Fineman JR, et al. Fetal model of single ventricle physiology: hemodynamic
effects of oxygen, nitric oxide, carbon dioxide, and hypoxia in the early postnatal period. J Thorac Cardiovasc
Surg 1996;112:437.
385. Wessel DL. Commentary: simple gases and complex single ventricles. J Thorac Cardiovasc Surg
1996;112:655.
386. Mosca RS, Bove EL, Crowley DC, et al. Hemodynamic characteristics of neonates following first stage
palliation for hypoplastic left heart syndrome. Circulation 1995;92:II267.
387. Bradley SM, Atz AM, Simsic JM. Redefining the impact of oxygen and hyperventilation after the Norwood
procedure. J Thorac Cardiovasc Surg 2004;127:473.
388. Strauss KM, Dongas A, Hein U, et al. Stage 1 palliation of hypoplastic left heart syndrome: implications
of blood gases. J Cardiothorac Vasc Anes 2001;15:731.
389. Hoffman GM, Ghanayem NS, Kampine JM, et al. Venous saturation and the anaerobic threshold in
neonates after the Norwood procedure for hypoplastic left heart syndrome. Ann Thorac Surg 2000;70:1515.
390. Riordan CJ, Randsbaek F, Storey JH, et al. Inotropes in the hypoplastic left heart syndrome: effects in an
animal model. Ann Thorac Surg 1996;62:83.
391. Sano S, Ishino K, Kawada M, et al. Right ventricle-pulmonary artery shunt in first-stage palliation of
hypoplastic left heart syndrome. J Thorac Cardiovasc Surg 2003;126:504.
392. Maher K, Pizarro C, Gidding S, et al. Improved hemodynamic profile following the Norwood procedure
with right ventricle to pulmonary artery conduit. Circulation 2002;106:II.
393. Pizarro C, Malec E, Maher K, et al. Right ventricle to pulmonary artery conduit improves outcome after
Norwood procedure for hypoplastic left heart syndrome. Circulation 2002;106:II.
394. Ohye RG, Kagisaki K, Lee LA, et al. Biventricular repair for aortic atresia or hypoplasia and ventricular
septal defect. J Thorac Cardiovasc Surg 1999;118:648.
395. Apfel HD, Levenbraun J, Quaegebeur JM, et al. Usefulness of preoperative echocardiography in predicting
left ventricular outflow obstruction after primary repair of interrupted aortic arch with ventricular septal
defect. Am J Cardiol 1998;82:470.
396. Moulton AL, de Leval MR, Macartney FJ, et al. Rastelli procedure for transposition of the great arteries,
ventricular septal defect, and left ventricular outflow tract obstruction. Br Heart J 1981; 45:20.
397. Nikaidoh H. Aortic translocation and biventricular outflow tract reconstruction. A new surgical repair for
transposition of the great arteries associated with ventricular septal defect and pulmonary stenosis. J Thorac
398. Borromee L, Lecompte Y, Batisse A, et al. Anatomic repair of anomalies of ventriculoarterial connection
associated with ventricular septal defect. Clinical results in 50 patients with pulmonary outflow tract
obstruction. J Thorac Cardiovasc Surg 1988;95:96.
399. Blume ED, Altmann K, Mayer JE, et al. Evolution of risk factors influencing early mortality of the arterial
switch operation. J Am Coll Cardiol 1999;33:1702.
400. Boutin C, Jonas RA, Sanders SP, et al. Rapid two-stage arterial switch operation. Acquisition of left
ventricular mass after pulmonary artery banding in infants with transposition of the great arteries. Circulation
1994;90:1304.
401. Wernovsky G, Giglia TM, Jonas RA, et al. Course in the intensive care unit after preparatory' pulmonary
artery banding and aortopulmonary shunt placement for transposition of the great arteries with low left
ventricular pressure. Circulation 1992; 86:II133.
402. Haworth SG. Total anomalous pulmonary venous return. Prenatal damage to pulmonary vascular bed
and extrapulmonary veins. Br Heart J 1982;48:513.
403. Lincoln CR, Rigby ML, Mercanti C, et al. Surgical risk factors in total anomalous pulmonary venous
connection. Am J Cardiol 1988;61:608.
404. Jahangiri M, Zurakowski D, Mayer JE, et al. Repair of the truncal valve and associated interrupted arch in
neonates with truncus arteriosus. J Thorac Cardiovasc Surg 2000;119:508.
P.767
405. Laussen PC, Reid RW, Stene RA, et al. Tracheal extubation of children in the operating room after atrial
septal defect repair as part of a clinical practice guideline. Anesth Analg 1996;82:988.
406. Walker RE, Mayer JE, Alexander ME, et al. Paucity of sinus node dysfunction following repair of sinus
venosus defects in children. Am J Cardiol 2001;87:1223.
407. Aebe R, Katogi T, Hashizume K, et al. Liberal use of tricuspid valve detachment for transatrial ventricular
septal defect closure. Ann Thorac Surg 2003;76:1073.
408. Backer CL, Idriss FS, Zales VR, et al. Surgical management of the conal (supracristal) ventricular septal
defect. J Thorac Cardiovasc Surg 1991;102:288.
409. Van Praagh S, Mayer JE Jr, Berman NB, et al. Apical ventricular septal defects: follow-up concerning
anatomic and surgical considerations. Ann Thorac Surg 2002;73:48.
410. Backer CL, Mavroudis C, Alboliras ET, et al. Repair of complete atrioventricular canal defects: results with
the two-patch technique. Ann Thorac Surg 1995;60:530.
411. Hanley FL, Fenton KN, Jonas RA, et al. Surgical repair of complete atrioventricular canal defects in
infancy. Twenty-year trends. J Thorac Cardiovasc Surg 1993;106:387.
412. Dreyfus G, Jebara V, Mihaileanu S, et al. Total orthotopic heart transplantation: an alternative to the
standard technique. Ann Thorac Surg 1991;52:1181.
413. Costello JM, Pahl E. Prevention and treatment of severe hemodynamic compromise in pediatric heart
transplant patients. Paediatric Drugs 2002;4:705.
414. Kobashigawa JA, Stevenson LW, Brownfield E, et al. Does short-course induction with OKT3 improve
outcome after heart transplantation? A randomized trial. J Heart Lung Transplant 1993; 12:205.
415. Boucek RJ Jr, Naftel D, Boucek MM, et al. Induction immunotherapy in pediatric heart transplant
recipients: a multicenter study. J Heart Lung Transplant 1999;18:460.
416. Reddy VM, McElhinney DB. Update on prospects for fetal cardiovascular surgery. Curr Opin Pediatr
1997;9:530.
417. Tulzer G, Arzt W, Franklin RC, et al. Fetal pulmonary valvuloplasty for critical pulmonary stenosis or
atresia with intact septum. Lancet 2002;360:1567.
418. Tworetzky W, Jennings RW, Wilkins-Haug LE, et al. Balloon dilation of severe aortic stenosis in the fetus:
technical advances. J Am Coll Cardiol 2003;41:496A.
Chapter 35
JAUNDICE
M. Jeffrey Maisels M.B., B.Ch.
Introduction
Jaundice is the most common and one of the most vexing problems that
can occur in the newborn. As Hansen points out in an elegant historical
review (1), neonatal jaundice must have been noticed by caregivers
throughout the centuries%, but the first documented scientific
description of neonatal jaundice occurred in the latter part of the
eighteenth century when Baumes was awarded a prize from the
University of Paris for his description of the clinical course of jaundice in
10 infants (1). Although most jaundiced infants are otherwise perfectly
healthy, they make us anxious because bilirubin is potentially toxic to the
central nervous system.
Jaundice occurs when the liver cannot clear a sufficient amount of
bilirubin from the plasma. When the problem is excessive bilirubin
formation or limited uptake and conjugation, unconjugated (i.e., indirectreacting) bilirubin appears in the blood. When bilirubin glucuronide
excretion is impaired (i.e., cholestasis), conjugated monoglucuronide and
diglucuronide (i.e., direct-reacting) bilirubin accumulate in plasma and,
because of their solubility, also appear in the urine. There is also a fourth
bilirubin fraction (unconjugated, monoglucuronide, and diglucuronide are
the first three) known as -bilirubin, which is formed nonenzymatically
from conjugated bilirubin and reacts directly with the diazo reagent (2).
In most jaundiced neonates, only unconjugated bilirubin is found in the
blood, and the accumulated bilirubin is distributed by the circulation
throughout the body and produces clinical jaundice. It generally is
assumed that to cross intact cell membrane barriers the bilirubin must be
free, or dissociated, from its albumin binding.
Formation, Structure, and

Properties of Bilirubin
Detailed reviews of the chemistry and metabolism of bilirubin are found
elsewhere (3,4,5,6,7).
Bilirubin is the end product of the catabolism of iron protoporphyrin or
heme, of which the major source is circulating hemoglobin (8). The
formation of bilirubin from hemoglobin involves removal of the iron and
protein moieties, followed by an oxidative process catalyzed by the
enzyme microsomal heme oxygenase, an enzyme found in the
reticuloendothelial system as well as many other tissues (9). The methane bridge of the heme porphyrin ring is opened and carbon
monoxide (CO) and biliverdin are formed (Fig. 35-1). One molecule of CO
and biliverdin (and, subsequently, bilirubin) is formed for each molecule
of heme degraded (4).
Figure 35-2 shows a linear representation of bilirubin. Although
conventionally illustrated as in Figs. 35-1 and 35-2, it is likely that the
prevalent structure of bilirubin in plasma has the ridge-tile conformation
shown in Fig. 35-3, because it is consistent with the biologic properties of
bilirubin. In this conformation, the bilirubin polar groups of the molecule
are involved in intramolecular hydrogen bonding that restricts solvation
and renders the pigment nearly insoluble in water at pH 7.4 but soluble in
nonpolar solvents such as chloroform (10). Under these circumstances,
bilirubin behaves like other lipophilic substances (e.g., dioxin,
polychlorinated biphenyls)it is difficult to excrete but crosses biologic
membranes, such as the placenta, blood-brain barrier, and hepatocyte
plasma membrane, easily (10,11). The addition of methanol or ethanol
interferes with hydrogen bonding and results in an immediate diazo
reaction, the basis for measurement of indirect-reacting bilirubin by the
van den Bergh test.
Fetal Bilirubin Metabolism

Bilirubin can be detected in normal amniotic fluid after about 12 weeks of
gestation, but it disappears by 36 to 37 weeks' gestation. The ability of human fetal
liver to remove bilirubin from the circulation and to conjugate it is severely limited.
Between 17 and 30 weeks of gestation, uridine diphosphoglucuronosyl transferase
(UDPGT) activity in fetal liver is only 0.1% of adult values, but it increases tenfold
to 1% of adult values between 30 and 40 weeks' gestation. After birth, activity
increases exponentially, reaching adult levels by 6 to 14 weeks' gestation (Fig. 354). This increase is independent of gestation (Table 35-1) (12,13).
The major route of fetal bilirubin excretion is across the placenta. Because virtually
all the fetal plasma bilirubin is
P.769
unconjugated, it is readily transferred across the placenta to the maternal
circulation, where it is excreted by the maternal liver. Thus, the newborn rarely is
born jaundiced, except in the presence of severe hemolytic disease, when there
may be accumulation of unconjugated bilirubin in the fetus. Conjugated bilirubin is
not transferred across the placenta, and it also may accumulate in the fetal plasma
and other tissues.
Figure 35-1 Biosynthesis of bilirubin. (From Lightner DA, McDonagh AF.

Molecular mechanisms of phototherapy for neonatal jaundice. Acc Chem Res
Maternal Hyperbilirubinemia and Its Effect on the

Fetus
Obstetricians and pediatricians occasionally are confronted with a pregnant mother
who has hyperbilirubinemia as a result of hemolytic anemia or liver disease.
Reported cases in the literature provide evidence for transfer of unconjugated
bilirubin from the mother to her fetus, but no clear guidelines for management
(14,15,16).
Figure 35-2 The chemical structure of bilirubin. (From McDonagh AF, Lightner
DA. Like a shrivelled blood orangebilirubin, jaundice and phototherapy.
Pediatrics 1985;75:443-455, with permission.)
It is possible that prolonged exposure of the fetus to a modest degree of

unconjugated hyperbilirubinemia in utero could lead to neurologic damage (16).
Neonatal Bilirubin Metabolism

Bilirubin Production
The normal destruction of circulating erythrocytes accounts for about 75% of the daily bilirubin production in the
newborn. Senescent erythrocytes are removed and destroyed in the reticuloendothelial system, where the heme
is catabolized and converted to bilirubin (Fig. 35-5). The catabolism of 1 g of hemoglobin yields 35 mg of
bilirubin.
A significant contribution (25% or more) to the daily production of bilirubin in the neonate comes from sources
other than effete erythrocytes (see Fig. 35-5). This bilirubin consists of two major components:
A nonerythropoietic component resulting from the turnover of nonhemoglobin heme protein and free
heme, primarily in the liver.
P.770
An erythropoietic component arising primarily from ineffective erythropoiesis and the destruction of
immature erythrocyte precursors, either in the bone marrow or soon after release into the circulation.
Figure 35-3 Preferred conformation of bilirubin. Chemical structure (left); bent paper clip analogy (middle);
space-filling molecular model (right). Each representation is asymmetric and has a nonsuperimposable mirror
image, like a D- or L-amino acid. Only one of the two possible mirror-image forms is shown in each
representation. (From McDonagh AF, Lightner DA. Like a shrivelled blood orange bilirubin, jaundice and
phototherapy. Pediatrics 1985;75:443-455 with permission.)
Figure 35-4 Developmental pattern of human hepatic uridine diphosphoglucuronosyl transferase (UDPGT)
activity. Samples were obtained from the livers of fetuses after elective abortions, at autopsy from premature
and full-term newborns who survived fewer than 7 days, and from liver biopsies of infants, children, and adults
undergoing laparotomy. Each point represents the activity of the liver homogenate of a single patient, but
results for patients older than 18 weeks of age are shown as a mean SD. (From Kawade N, Onishi S. The
prenatal and postnatal development of UDP-glucuronyl transferase activity toward bilirubin and the effect of
premature birth on this activity in the human liver. Biochem J 1981;196:257-260, with permission.)
Transport and Hepatic Uptake of Bilirubin

Once bilirubin leaves the reticuloendothelial system, it is transported in the plasma and bound reversibly to
albumin. Unconjugated bilirubin is bound tightly to albumin at a primary (high-affinity) binding site as well as a
secondary (low-affinity) site. Because binding affinity at the primary binding site is 107 to 108 L/mol (17), the
concentrations of free or unbound bilirubin in plasma are very low, even in the presence of significant
hyperbilirubinemia. (The magnitude of the affinity constant was recently questioned (18).) At pH 7.4, the
solubility of bilirubin is very low (about 4 nmol/L [0.24 mg/dL]).
TABLE 35-1 FETAL BILIRUBIN METABOLISM
Bilirubin detected in amniotic fluid at 12 weeks of gestation

Unconjugated bilirubin excreted across placenta
UDPGT activity at term ~1% of adult
UDPGT activity reaches adult levels by 614 weeks
Figure 35-5 Neonatal bile pigment metabolism. RBC, erythrocytes; R.E., reticuloendothelial.
The parenchymal cells of the liver have a selective and highly efficient capacity for removing unconjugated
bilirubin from the plasma. When the bilirubin-albumin complex reaches the plasma membrane of the hepatocyte,
a proportion of the bilirubin, but not the albumin, is transferred across the cell membrane into the hepatocyte, a
process that potentially involves four different transport proteins (3). In the hepatocyte, bilirubin is bound
principally to ligandin and possibly other cytosolic-binding proteins (Fig. 35-6) (3). A network of intracellular
microsomal membranes may also play an important role in the transfer of bilirubin within the cell and to the
endoplasmic reticulum (3,5).
Conjugation and Excretion of Bilirubin

Because of its hydrogen-bonded conformation (see Formation, Structure, and Properties of Bilirubin above),
unconjugated (i.e., indirect-reacting)
P.771
bilirubin is nonpolar and insoluble in aqueous solutions at pH 7.4 and must be converted to its water-soluble
conjugate (i.e., direct-reacting bilirubin) before it can be excreted (see Fig. 35-6). This is achieved when
bilirubin is combined enzymatically with a sugar, glucuronic acid, producing bilirubin monoglucuronide and
diglucuronide pigments that are more water soluble and sufficiently polar to be excreted into the bile or filtered
through the kidney.
Figure 35-6 Bilirubin transport and conjugation in the hepatocyte. Two mechanisms have been proposed for
uptake of bilirubin from the extracellular environment: by means of an albumin receptor or directly. In either
case, carrier protein may be involved in transmembrane passage (upper box). Transport into the endoplasmic
reticulum is facilitated by complexing to ligandin, but direct membrane-to-membrane transfer also may occur.
Cytosolic bilirubin is in equilibrium with endoplasmic reticulum bilirubin. The hypothesis of conjugation of
insoluble bilirubin to polar glucuronides (lower box). Conjugated bilirubin may enter the bile canaliculi by
either vesicular transport or carrier-mediated transport. (From Gollan JL, Knapp AB. Bilirubin metabolism and
congenital jaundice. Hosp Pract (Off Ed) 1985;Feb 15:83-106, with permission.)
The presence of elevated bilirubin concentrations in utero prematurely induces bilirubin uridine
diphosphoglucuronate glucuronosyltransferase (UGT) activity, which suggests that bilirubin plays an important
role in the initiation of its own conjugation after birth (19).
Structure and Function of the Uridine Diphosphoglucuronate

Glucuronosyltransferase 1A1 Gene
The process of conjugation is catalyzed by a specific hepatic enzyme isoform (1A1) belonging to the UGT family
of enzymes. These enzymes metabolize endogenous compounds and various food chemicals in most tissues.
Although the UGT1 family contains several isoforms, only the A1 isoform (UGT1A1) participates in the
conjugation of bilirubin (20). The glucuronosyltransferase enzyme is synthesized in the hepatocyte and its
structure is determined by the UGT1A1 gene (Fig. 35-7 and Table 35-2).
The gene encoding the UGT1 enzyme is located on chromosome 2 at 2q37 (21) and consists of 4 common exons
and 13 variable exons (see Fig. 35-7) (22,23). The gene also has a noncoding promoter area (see Fig. 35-7B),
which is an upstream regulatory region controlling gene expression. The UGT promoter contains a TATAA box,
which is a DNA deoxyribonucleic acid (DNA) sequence of thymine (T) and adenine (A). Mutations in the 1A1
exon or its promoter will affect bilirubin conjugation. Examples of this effect are seen in Gilbert syndrome and
the Crigler-Najjar syndromes (see Pathologic Causes of Jaundice: Decreased Bilirubin Clearance below).
Variations in the promoter sequence can also cause indirect-reacting hyperbilirubinemia.
TABLE 35-2 THE UGT 1A1 GENE
Encodes the UGT1 enzyme

Located on chromosome 2
Consists of 4 common exons (25) and 13 variable exons (A1A13)
Noncoding promoter area controls gene expression
UGT promoter contains TATAA box
Mutations in 1A1 exon or promoter affect bilirubin conjugation (e.g., Gilbert and Crigler-Najjar
syndromes)
Figure 35-7 The human UGT1 gene locus. A: Schematic of the genomic structure of the UGT1 gene complex.
B: Exploded view of exon 1A1 and common exons 2 to 5 of the gene complex that have been identified as sites
for genetic mutations associated with absent or decreased UGT activity that cause deficiencies of bilirubin
conjugation. (From Clarke DJ, Moghrabi N, Monaghan G, et al. Genetic defects of the UDPglucoronosyltransferase-1 (UGT1) gene that cause familial nonhaemolytic unconjugated hyperbilirubinemias.
Clinica Chim Acta 1997;266:63-74, with permission.)
P.772
Transfer of Bilirubin into Bile and Intestinal Transport

After conjugation, bilirubin is excreted rapidly into the bile canaliculi by the liver cell, a process that requires
metabolic work for the active transport of bilirubin across a large concentration gradient (see Fig. 35-6) (5).
Inter-ference with this process is probably responsible for the hyperbilirubinemia associated with hepatocellular
disorders such as hepatitis.
Once in the small intestine, conjugated bilirubin is not reabsorbed. In the healthy adult, it is largely reduced by
the action of colonic bacteria to a series of colorless tetra-pyrroles, collectively known as urobilinogen, and an
insignificant amount is hydrolyzed to unconjugated bilirubin and reabsorbed by way of the enterohepatic
circulation. In the newborn, however, this enterohepatic circulation of bilirubin is significant and important (see
Jaundice in the Healthy Newborn: Physiologic Jaundice below). In addition, in conditions involving high plasma
bilirubin levels and poor hepatic excretion, there is a gradient for unconjugated bilirubin from the plasma to the
intestinal lumen, and significant amounts of unconjugated bilirubin may be cleared by diffusion across the
intestinal wall (24). Figure 35-5 summarizes bile pigment metabolism in the newborn.
Physiologic Mechanisms of Neonatal

Jaundice
At any time in the infant's first few days after birth, the serum bilirubin level
reflects a combination of the effects of bilirubin production, conjugation, and
enterohepatic circulation. Using measurements of blood carboxyhemoglobin
(COHb) corrected for ambient CO (COHbc) as an index of bilirubin production and
high-performance liquid chromatography (HPLC) measurements of conjugated
bilirubin, Kaplan and associates (25) demonstrated that an imbalance between
bilirubin production and conjugation is fundamental in the pathogenesis of
neonatal bilirubinemia. Several processes account for the bilirubinemia that occurs
in virtually all newborns (Table 35-3).
Increased Bilirubin Load on the Liver Cell

Bilirubin Production
CO is produced in equimolar quantities with bilirubin and measurements of CO
production show that the normal
P.773
newborn produces an average of 8 to 10 mg/kg (13.7 to 17.1 mol/ kg) of
bilirubin per day (26,27). This is more than twice the rate of normal daily bilirubin
production in the adult and is explained by the fact that the neonate has a higher
circulating erythrocyte volume, a shorter mean erythrocyte lifespan, and a larger
early labeled bilirubin peak (see Table 35-3). Bilirubin production decreases with
increasing postnatal age but is still about twice the adult rate by age 2 weeks (26).
TABLE 35-3 PHYSIOLOGIC MECHANISMS OF NEONATAL JAUNDICE
Increased bilirubin load on liver cell

Increased erythrocyte volume
Decreased erythrocyte survival
Increased early labeled bilirubin
Increased enterohepatic circulation of bilirubin
Decreased hepatic uptake of bilirubin from plasma
Decreased ligandin
Decreased bilirubin conjugation
Decreased uridine diphosphoglucuronosyl transferase activity
Defective bilirubin excretion
Excretion impaired but not rate limiting
Enterohepatic Circulation
The newborn reabsorbs much larger quantities of unconjugated bilirubin by way of
the enterohepatic circulation, than does the adult. Infants have fewer bacteria in
the small and large bowel and greater activity of the deconjugating enzyme glucuronidase.(28) As a result, conjugated bilirubin, which is not reabsorbed, is
not converted to urobilinogen but is hydrolyzed to unconjugated bilirubin, which is
reabsorbed, thus increasing the bilirubin load on an already stressed liver (see
Fig. 35-5). Studies in newborn humans (29), monkeys (30), and Gunn rats (31)
suggest that the enterohepatic circulation of bilirubin is a significant contributor to
physiologic jaundice. In the first few days after birth, caloric intake is low, which
contributes to an increase in the enterohepatic circulation (32,33).
Decreased Clearance of Bilirubin from the Plasma

Uptake
Ligandin, the predominant bilirubin-binding protein in the human liver cell, is
deficient in the liver of newborn monkeys. It reaches adult levels in the monkey
by 5 days of age, coinciding with a fall in bilirubin levels, and administration of
phenobarbital increases the concentration of ligandin (34). Although this suggests
that impaired uptake may contribute to the pathogenesis of physiologic jaundice,
uptake does not appear to be rate limiting.
Conjugation
Deficient UGT1A1 activity, with resultant impairment of bilirubin conjugation, has
long been considered a major cause of physiologic jaundice. In human infants, the
early postnatal increase in serum bilirubin appears to play an important role in the
initiation of bilirubin conjugation. (19) In the first 10 days after birth, UGT1A1
activity in full-term and premature neonates usually is less than 1% of adult
values (see Fig. 35-4) (12,13). Thereafter, the activity increases at an exponential
rate, reaching adult values by 6 to 14 weeks of age (13). The postnatal increase
in UGT1A1 activity is independent of the infant's gestation.
Excretion
The absence of an elevated serum level of conjugated bilirubin in physiologic
jaundice suggests that, under normal circumstances, the neonatal liver cell is
capable of excreting the bilirubin that it has just conjugated. Nevertheless, the
ability of the newborn liver to excrete conjugated bilirubin and other anions (e.g.,
drugs, hormones) is more limited than that of the older child or adult and may
become rate limiting when the bilirubin load is significantly increased. Thus, when
intrauterine hyperbilirubinemia occurs, usually as a result of isoimmunization, it is
not uncommon to find an elevated serum level of conjugated bilirubin (19).
Physiologic Jaundice
Because at some point during the first week of life almost every newborn
has a total serum bilirubin (TSB) level that exceeds 1 mg/dL (17 mol/L,
the upper limit of normal for an adult), and two-thirds or more of
newborns will appear clinically jaundiced, this type of transient
hyperbilirubinemia has been called physiologic jaundice. This jaundice
results from the interaction of a number of factors (see Table 35-3). The
term physiologic jaundice generally is applied to newborns whose TSB
level falls within the normal range, but because of the significant
differences in TSB levels in different populations, it can be difficult to
define what is normal or abnormal, physiologic or nonphysiologic.
Additionally, defining the term normal is, in itself, a difficult task, and
depends on whether one chooses a gaussian, percentile, diagnostic, risk
factor, or therapeutic definition of the term (35).
In premature newborns, the term physiologic jaundice is of little value. If
untreated, low-birth-weight infants have exaggerated and prolonged
hyperbilirubinemia. Although this may be considered physiologic
because it occurs in all preterm infants, in very-low-birth-weight infants,
TSB levels well within the physiologic range are considered potentially
hazardous and are treated with phototherapy. Thus, the natural history of
hyperbilirubinemia in the very-low-birth-weight infant is never observed,
and defining certain bilirubin levels as physiologic in this population is
misleading and potentially dangerous. Using a diagnostic definition of
normal (35), a TSB level of 10 mg/dL (171 mol/ L) on day 4 in a 750-g
neonate would be considered completely physiologic, and no
investigation need be done to identify a cause for this jaundice.
Nevertheless, almost all neonatologists would treat this infant with
phototherapy, implying that this value exceeds the therapeutic definition
of normal (i.e., treatment is much more likely to do good than harm).
Thus, in today's neonatal intensive care population, the term physiologic
jaundice has no meaning and no utility and should be abandoned.
Bilirubin Toxicity
Kernicterus
Pathology
The first description of kernicterus (or brain jaundice) in newborns was provided by Hervieux in 1847 (1) and in
P.774
1875, Orth (36) observed bilirubin pigment at autopsy in the brains of infants who were severely jaundiced.
Schmorl (37) subsequently described two forms of brain icterus, the first characterized by a diffuse yellow
coloration of the entire brain substance, and a second form in which the jaundiced coloration appears to be
completely circumscribed andlimited to the so-called kern or nuclear region of the brain.
TABLE 35-4 COMPARATIVE NEUROPATHOLOGY OF KERNICTERUS
Premature Infants,
Full-term Infants,
Hyperbilirubinemia
Homozygous Gunn
Rats
Low Bilirubin
Levels
Globus pallidus
Subthalamus
Hypothalamus
+
+
+
+
+
-
+
+
-
Horn of Ammon
Reticular zone of the
substantia nigra
Cranial nerve nuclei
+
+
+
+
+
+
Reticular formation
Central pontine nuclei
Interstitial nucleus
Locus ceruleus
Lateral cuneate nucleus
of the medulla
+
+
+
+
Topography of Lesions
Cerebellum
Dentate nuclei
Nuclei of roof of fourth

+
+
+
ventricle
Purkinje cells
+
+
Spinal cord
+
+
+
+, Yellow pigment present; -, yellow pigment absent.
From Ahdab-Barmada M, Moossy J. The neuropathology of kernicterus in the premature neonate:
diagnostic problems. J Neuropathol Exp Neurol 1984;43:4556, with permission.
Topography
Full-term infants who die of kernicterus demonstrate bilirubin staining in a characteristic distribution (Table 354), although a variety of patterns have been described, grossly and microscopically (38). Kernicteric premature
infants and Gunn rats with inherited UGT deficiency display a similar topography of neuronal damage (see Table
35-4) (39). Those regions most commonly affected are the basal ganglia, particularly the subthalamic nucleus
and the globus pallidus; the hippocampus; the geniculate bodies; various brainstem nuclei, including the inferior
colliculus, oculomotor, vestibular, cochlear, and inferior olivary nuclei; and the cerebellum, especially the
dentate nucleus and the vermis (39,40). Ahdab-Barmada has provided a detailed review of the neuropathology
of kernicterus, and its anatomic, cytologic, and histologic characteristics (41).
Gross Anatomy
Yellow staining of the brain occurs when it is exposed to elevated levels of bilirubin. Table 35-5 lists the three
patterns of bilirubin staining of the brain seen in the newborn (41).
Histology and Cytology

Table 35-6 summarizes the neuropathologic findings of kernicterus. There can be some confusion regarding the
diagnosis of kernicterus in the presence of yellow discoloration of the central nervous system. Ahdab-Barmada
emphasizes that the diagnosis of kernicterus should only be applied to the incorporation of bilirubin pigment into
gangliosides or phospholipids of mature neurons with subsequent damage to the neuron, depending on the
amount of pigment trapped within the cell (41). The unique topographic pattern of nuclear involvement is
described above (see Topography) and the combination of the bright yellow-orange staining of these brain
nuclei, together with evidence of neuronal damage and degeneration within the nuclei, is necessary before a
diagnosis of kernicterus can be made (41).
TABLE 35-5 PATTERNS OF BILIRUBIN STAINING OF THE BRAIN IN HYPERBILIRUBINEMIA
Diffuse yellow staining of areas that normally lack a blood-brain barrier, e.g., leptomeninges,
ependyma, choroid plexus, cerebrospinal fluid.
Diffuse yellow staining of brain tissues in areas where blood-brain barrier integrity has been
compromised (as can occur following hypoxic ischemic encephalopathy, periventricular leukomalacia,
ischemic cerebral infarct).
Yellow staining of specific neuronal groups (kernicterus).
From Ahdab-Barmada M. The neuropathology of kernicterus: definitions and debate. In: Maisels MJ,
Watchko JF, eds. Neonatal Jaundice. London: Harwood Academic, 2000:7588.
TABLE 35-6 NEUROPATHOLOGIC FINDINGS OF KERNICTERUS

Early (25 Days)
Subacute (610 Days)
Late (>10 Days)
Yellow pigment in neuronal

cytoplasm
Moth-eaten appearance of
Spongy neuropil within involved

nuclei
Hypertrophic, bare astrocytic
Neuronal loss with astrocytosis

Granular mineralization of residual
neuronal and nuclear

membranes
nuclei
neuronal membranes in affected nuclei
Pyknotic nucleus
Basophilic neurons with increased Demyelinization of optic tracts and

nuclear density
fornix
Loss of Nissl substance
Cellular dissolution of some

Dysmyelination and degeneration of
neurons
globus pallidus and subthalamic
Granular mineralization of
nucleus
neuronal cytoplasmic membranes
Basophilic cytoplasm
Periodic acid-Schiff-positive
membrane-bound aggregates
within neurons
From Ahdab-Barmada M, Moossy J. The neuropathology of kernicterus in the premature neonate:
diagnostic problems. J Neuropathol Exp Neurol 1984;43:4556, with permission.
P.775
Autopsies on jaundiced infants reveal bilirubin staining of the aorta, pleural fluid, and ascitic fluid, or a
generalized yellow cast throughout the viscera. The staining usually is not considered a sign of tissue damage
unless other cytologic changes are found (38). Bilirubin staining also can be found in necrotic tissue anywhere in
the body and has been described in the gastrointestinal tract, lungs (hyaline membranes) (42), kidney,
adrenals, and gonads. In infants with hemolytic disease, bile plugs commonly are found in the canaliculi between
the hepatocytes, especially in the periportal areas. The kidneys may show bilirubin-stained tubular casts,
bilirubin crystals in the small vessels or in edematous interstitium, and renal tubular necrosis. The bilirubin
infarcts (i.e., patches of yellow staining in the renal medulla) are probably the result of focal areas of acute
tubular necrosis that have been stained by bilirubin (38).

Neuronal necrosis is the dominant histopathologic feature after 7 to 10 days of postnatal life (see Table 35-6).
For the most part, its distribution corresponds with the distribution of bilirubin staining, although there are some
exceptions to this rule. For example, intense staining develops in the olivary and dentate nuclei, but there is
little neuronal necrosis in these regions. The important areas of neuronal injury (as opposed to staining) include
the basal ganglia, brainstem oculomotor nuclei, and brainstem auditory (cochlear) nuclei (40). The involvement
of these regions explains some of the clinical sequelae of bilirubin encephalopathy (see Clinical Features of
Bilirubin Encephalopathy below).
Clinical and Pathologic Correlations

Originally a pathologic diagnosis and later a well-defined acute and chronic neurologic syndrome, kernicterus or
bilirubin encephalopathy appears to be a less-well-circumscribed entity that includes nuclear bilirubin staining of
very-low-birth-weight infants who died of other causes and, possibly, a subtle chronic encephalopathy in which
extrapyramidal motor disturbances and sensorineural hearing deficit are not the predominant features.
Most, but not all, full-term infants seen today with the pathologic changes described manifest the clinical
symptomatology of this disorder, including very high serum bilirubin levels (commonly higher than 30 mg/dL
[513 mol/L]). Exceptions have been described. Perlman and associates (43) recently reported kernicterus at
autopsy in two very sick near-term infants with maximum TSB levels of 5.2 and 14.4 mg/dL (89 to 246 mol/L).
Yellow staining of the brain also has been observed in premature infants who manifested none of the clinical
signs of kernicterus during life and in whom TSB levels remained low (44,45). Turkel and colleagues (46)
identified 32 infants with kernicterus at autopsy and compared them with 32 control infants of similar
gestational ages without kernicterus. In the kernicteric infants, although the gross pattern of staining followed
that of classic kernicterus, the typical histologic changes characteristic of kernicterus were found in only three
patients. These authors suggest that the bilirubin staining they observed probably was not the same
clinicopathologic entity as the kernicterus of posticteric encephalopathy. Instead of the neuronal degeneration
typically seen, they found spongy change and gliosis, which both imply nonspecific damage to the brain. This
suggests that prior diffuse injury may predispose the brain to bilirubin deposition at relatively low levels of
serum bilirubin.
Ahdab-Barmada and Moossy (39) found kernicterus in 97 autopsies of neonates (95 younger than 36 weeks of
gestation)
P.776
The neuropathology in these infants was strikingly similar to that of classic kernicterus in the full-term neonate
and in the Gunn rat (see Table 35-4). In the National Institute of Child Health and Human Development
(NICHHD) cooperative phototherapy study, four low-birth-weight infants had autopsy-proven kernicterus (47).
The neuropathologic findings in these infants were those of classic kernicterus. As Table 35-7 shows, the
neuropathology of kernicterus is different from that of hypoxic ischemic encephalopathy. Even though hypoxic
ischemic insults may predispose the brain to bilirubin deposition in some low-birth-weight infants, in others the
typical histologic features of kernicterus will be found.
TABLE 35-7 COMPARATIVE NEUROPATHOLOGY OF KERNICTERUS AND ANOXICISCHEMIC
ENCEPHALOPATHY IN THE PREMATURE NEONATE
a
Topography of Lesions
Kernicterus
Anoxic-Ischemic
Encephalopathy
Cerebral cortex
Periventricular white matter
Corpus striatum
Thalamus
Horn of Ammon
Absent
Absent
Globus pallidus
Subthalamus
Resistant sector (H23)
Present
Present
Putamen and caudate nuclei
Anterior and lateral nuclei
Midbrain
Interstitial nucleus
Inferior colliculi
Nuclei of nerve III
Reticular portion of
substantia nigra
Sommer sector (H1)

b
Nuclei of nerve III
Compact portion of
substantia nigra
Pons
Locus ceruleus
Nuclei of nerves VI, VII
b
Medulla
Cerebellum
Reticular formation
Vestibular and cochlear nuclei Inferior olivary nuclei
Superior olivary nuclei
b
Purkinje cells
Nuclei of roof of fourth

ventricle
a
Basal pontine nuclei

Superior olivary complex
Reticular formationb
Purkinje cells
Granular cells
Only topographic areas considered helpful for differential diagnosis are listed in this table.
Whenever neuronal damage was involved in the same structure in kernicterus and anoxic-ischemic
encephalopathy, the cytopathology was different.
From Ahdab-Barmada M. The neuropathology of kernicterus: definitions and debate. In: Maisels MJ,
Watchko JF, eds. Neonatal jaundice. London: Harwood Academic,2000:7588, with permission.
Pathophysiology of Bilirubin Toxicity

The pathogenesis of kernicterus is highly complex and the risk of developing kernicterus is related to a
multiplicity of factors (48), which are discussed below. The cellular and molecular mechanisms of bilirubin
toxicity were reviewed by Hansen (49,50) and Volpe (Fig. 35-8) (51).
Bilirubin Chemistry and Neurotoxicity

As discussed in Formation, Structure, and Properties of Bilirubin above, the polar groups of the bilirubin
molecule, in its most stable conformation, are involved in intramolecular hydrogen bonding that restricts
solvation and renders the pigment nearly insoluble in water at pH 7.4. When doubly ionized in alkaline medium,
the molecule is much more soluble. The low water solubility of bilirubin and its tendency to aggregate and
precipitate at physiologic pH, particularly acid pH, have long been thought to be key factors in its toxicity. Thus,
when the concentration of bilirubin acid exceeds its solubility, bilirubin may gradually aggregate and precipitate
from solution (17). Bilirubin crystals have been found in the brain cells of infants who died from kernicterus, and
bilirubin concentrations of 2 mg/dL (34 mol/L) have been observed in kernicteric brains (52). It is likely that
even higher local concentrations of pigment exist in the brain in kernicterus and may occur when aggregates
precipitate within brain cells (53,54). Wennberg (55) suggested that formation of reversible complexes between
bilirubin monoanion and membranes is also important in the development of bilirubin encephalopathy.
Although it is known that bilirubin uncouples oxidative phosphorylation and inhibits cellular respiration and
protein phosphorylation, there is no agreement that these are the key mechanisms of bilirubin toxicity in vivo
(see Fig. 35-8). Bilirubin also inhibits mitochondrial enzymes, interferes with DNA and protein synthesis (56),
and alters cerebral glucose metabolism (49,57). Unconjugated bilirubin will initiate a mitochondrial pathway of
apoptosis in developing brain neurons (58) and it inhibits the function of Nmethyl-aspartate-receptor ion
channels (59). Bilirubin also binds to lysine sites on albumin and ligandin, proteins that are essential for bilirubin
transport and metabolism (49). Consequently, lysine binding may have a role in the pathogenesis of bilirubin
toxicity.
It is not known why bilirubin is deposited preferentially in the basal ganglia, but it is possible that regional
differences in uptake, tissue binding, metabolism, or cellular
P.777
clearance play a role (49). It is also possible that there are regional differences in blood flow or blood brain
barrier permeability.
Figure 35-8 Effects of bilirubin on neurons and neuronal metabolic processes. Bilirubin affects a large number
of cellular functions and processes, both in vivo and in vitro, It decreases neuronal viability; increases
membrane permeability and decreases membrane potential; uncouples oxidative phosphorylation; inhibits
neurotransmitter release, synthesis, and uptake; inhibits synaptic activation; increases postsynaptic
excitability; decreases protein synthesis; and decreases synthesis and replication of DNA. Other bilirubinaffected functions not indicated in the illustration are inhibition of enzyme function and protein phosphorylation.
ER, endoplasmic reticulum; M, mitochondrial; N, nucleus; PSM, postsynaptic membrane; PSV, presynaptic
vesicles; SC, synaptic cleft. (From Hansen TWR. The pathophysiology of bilirubin toxicity. In: Maisels MJ,
Watchko JF, eds. Neonatal jaundice. London: Harwood Academic, 2000:89-104, with permission.)
Albumin Binding and the Concept of Free Bilirubin

Bilirubin is transported in the plasma as a dianion bound tightly, but reversibly, to serum albumin and the
portion that is unbound or loosely bound (sometimes termed free bilirubin) can more readily leave the
intravascular space and cross the intact blood brain barrier (60). Albumin has a primary, high-affinity binding
site at which the association constant, derived from the equilibrium concentrations of bound and free bilirubin, is
about 107 to 108 moles-1 (17). Because of this high binding affinity, equilibrium concentrations of unbound or
free bilirubin in plasma are very low. It has been widely accepted that bilirubin toxicity occurs when free bilirubin
enters the brain and binds to cell membranes (61) and that the presence of albumin mitigates the in vivo and in
vitro toxic effects of bilirubin (61,62).
The relationship between free bilirubin levels, kernicterus, and developmental outcome is discussed below in the
section on the clinical sequelae of hyperbilirubinemia (see Clinical Sequelae of Hyperbilirubinemia: Bilirubin
Binding Capacity, Kernicterus, and Developmental Outcome) Drugs that decrease albumin binding of bilirubin,
such as sulfisoxazole, increase the risk of kernicterus (63, 64). These observations are consistent with the
hypothesis that bilirubin is able to move freely across the blood-brain barrier, bind to tissues, and damage the
cells of the central nervous system.
Ostrow and colleagues have questioned the significance of some published data from in vitro models of bilirubin
cellular toxicity because of the high concentrations of unconjugated bilirubin that were used (18,65,66,67). They
suggest that bilirubin toxicity might occur at free bilirubin levels that are significantly lower than those
previously documented and that precipitation of unconjugated bilirubin in, or on, cells may not be required to
produce neurotoxicity (67). These issues cannot be resolved until clinical studies are performed on infants who
have had appropriate measurements of free bilirubin as newborns (68) and are then followed into childhood.
Measurement of Free Bilirubin

Although several techniques exist for measuring or estimating free or loosely bound bilirubin and the binding
capacity and affinity of bilirubin for albumin (17), changes in free bilirubin concentrations can be transient
because there is rapid equilibration and redistribution of bilirubin between the plasma (i.e., albumin) and the
tissues. Even under experimental conditions that lead to a significant increase in brain bilirubin content, the
differences in free
P.778
bilirubin concentrations in the serum between control and study animals is small (69) and Brodersen has
expressed doubt that the accurate measurement of free bilirubin is possible or clinically useful (17).
Nevertheless, there is some evidence to suggest that estimates of unbound bilirubin can provide clinically
relevant information regarding changes in auditory brainstem responses (ABRs) (70,71) and in frank kernicterus
(72).
There is an extensive literature dealing with bilirubin-binding tests, but no test is currently in general use in the
United States in clinical decision making, although a relatively simple semiautomated application of the
peroxidase oxidation test is in use in Japan (73) and has been approved in the United States by the Food and
Drug Administration (FDA) (74). However, the peroxidase method employs a 40-fold dilution of the sample that
can alter intrinsic bilirubin binding (68). Ahlfors developed a test that combines the peroxidase technique with a
diazo method for measuring conjugated and unconjugated bilirubin (68). This method uses minimal dilution and
should provide more reliable data on free bilirubin levels (68).
Because one molecule of albumin is capable of binding one molecule of bilirubin tightly at the primary binding
site, a bilirubin-albumin molar ratio of 1 represents about 8.5 mg of bilirubin per gram of albumin. Thus, a well,
full-term infant with a serum albumin concentration of 3 to 3.5 g/dL should be able to bind about 25 to 28 mg/
dL of bilirubin (428 to 479 mol/L) if no other endogenous or exogenous ligands compete for the same site. The
albumin-binding capacity of sick low-birth-weight infants is much less than that of full-term infants, and their
serum albumin levels often are lower so they are able to bind effectively much less bilirubin. Ahlfors has
suggested a range of bilirubin-to-albumin ratios (in mg/g) that can be used as a guide in the process of deciding
whether or not to perform an exchange transfusion in term and preterm infants at different levels of risk (75),
an approach that has been endorsed by the American Academy of Pediatrics (AAP) (76) (see Treatment below).
It should be noted, however, that there is significant variability in bilirubin-albumin binding among infants (77),
which can affect the validity of the bilirubin-to-albumin ratio. Because binding improves with increasing birth
weight, Ahlfors suggests a level of 1.3 g/dL per kg as a level of free bilirubin at which exchange transfusion
should be considered (72), although we have no data relating such free bilirubin levels to long-term outcome.
These data are urgently needed.
Factors Affecting the Binding of Bilirubin to Serum Albumin

This subject has been reviewed in detail (17); however, a few of the factors are discussed here.
Fatty Acids
Free fatty acids in plasma may compete with bilirubin for its binding to albumin, but significant interference with
bilirubin binding probably does not occur until molar ratios of free fatty acids-to-albumin (F:A) exceed 4:1 (17).
The infusion of 1 g/kg of intralipid over a 15-hour period in infants of less than 30 weeks of gestation produced
an F:A ratio of less than 3 and minimal increases in unbound bilirubin concentrations (78). With doses of 2 to 3
g/kg, however, higher ratios were found. Intravenous fat, given as a continuous infusion of 2 g/kg per day for 7
days to infants of 32 weeks or less of gestation (mean birth weight 1,200 g) produced F:A ratios of only 0.1 to
1.8. (79)
pH
The binding of bilirubin to albumin is thought to be unaffected by changes in the serum pH (80,81).
Nevertheless, the correction of neonatal acidosis in 11 sick newborns appeared to decrease the serum free
bilirubin concentration as measured by a peroxidase technique (82). The role of pH, on the other hand, may be
pivotal in determining the binding of bilirubin to cells and partitioning to extravascular tissue and, therefore, its
deposition in the central nervous system (55,60).
Drugs
The effect of numerous drugs on bilirubin-albumin binding has been tested in vitro using different methods. The
measured effect varies with the method used; some systems require much greater concentrations of the drug
than others to demonstrate an increase in unbound bilirubin. In vivo, of course, the effect of drugs and their
potential for inducing kernicterus depends, not simply on their ability to displace bilirubin from albumin, but also
on their route and mode of administration. Thus, a displacing drug is likely to be more dangerous when
administered intravenously as a bolus than as an infusion. Robertson and colleagues (83) reviewed the bilirubindisplacing effect of drugs used in neonatology (Table 35-8). They arbitrarily chose to consider an increase in the
free bilirubin concentration of 5% as potentially dangerous, and they consider a drug to be a potential displacer
if it occupies 5% or more of the available albumin. Knowledge of the usual peak serum bilirubin concentrations
and the percentage of albumin-bound drug also can be used to calculate the concentration of bound drug. If the
bound drug concentration is less than 15 mol/L, it is unlikely that this drug will cause significant displacement
of bilirubin (83).
Robertson and colleagues (83) calculated a maximal displacement factor, , from the KD value, using the
following equation:
where d is the concentration of free drug in the patient's plasma, and KD is the displacement constant, which
represents the competitive effect of the drug with bilirubin for albumin binding. If KD is 0, then =1, and the
drug does not displace bilirubin. If =1.2, there has been a 20% increase of free bilirubin concentration after
drug administration. Although an arbitrary value of 1.2 has been suggested as the upper permissible limit for
bilirubin displacement,
P.779
P.780
it is recommended that, as much as possible, drugs with the lowest values be selected. Table 35-8 lists the
effects of drugs used in neonatology on bilirubin-albumin binding. The free drug concentration is calculated from
the serum concentration and the percentage of bound drug as taken from existing data in the literature.
TABLE 35-8 EFFECT OF DRUGS USED IN NEONATOLOGY ON BILIRUBINALBUMIN BINDING
Agent
Anticonvulsants
Diazepam
Phenobarbital
Phenytoin
Valproate
Testing not required: Iorazepam
Antihypertensive agents
Diazoxide
1.00
1.04
1.02
1.09
Testing not required: hydralazine,

methyldopa, nitroprusside, reserpine
Cardiac drugs
Lidocaine
Procainamide
Testing not required: bretylium tosylate, digoxin, disopyramide, quinidine, verapamil
Diuretics
Acetazolamide
Bumetanide
Chlorothiazide
Ethacrynic acid
Furosemide
Hydrochlorothiazide
Testing not required: spironolactone
Infectious disease agents
Acyclovir
Amdinocillin
1.00
1.00
1.10
1.00
1.03
1.27
1.07
1.04
1.00
1.00
Ampicillin
Azlocillin
Aztreonam
Carbenicillin
Cefamandole
Cefazolin
1.08
1.33
1.12
1.35
1.07
1.17
Cefmenoxime
Cefmetazole
Cefonicid
Cefoperazone
Ceforanide
Cefotaxime
Cefotetan
1.10
2.01
1.71
1.18
1.04
1.05
1.74
Cefoxitin
Ceftazidime
?
1.02
Ceftizoxime
Ceftriaxone
Cefuroxime
Cephalothin
Cephapirin
Cephradine
1.03
3.00
1.02
1.03
1.03
1.02
Chloramphenicol
Chloroquine
Cilastatin
Clindamycin
Fusidate
1.02
1.00
1.00
1.00
1.00
Imipenem
Lincomycin
Methicillin
Metronidazole
1.00
1.17
1.00
1.11
Mezlocillin
Moxalactam
Nafcillin
Oxacillin
Penicillin G
Piperacillin
Polymyxin B
Quinine
Spiramycin
Streptomycin
Sulfadiazine
Sulfamethoxazole
Sulfisoxazole
Tazobactam
Ticarcillin
Trimethoprim
Vancomycin
Vidarabine
Testing not required: amphotericin B, ciprofloxacin, erythromycin, isoniazid, miconazole,
netilmicin, pyrimethamine, tobramycin
Miscellaneous
Calcium chloride
Calcium gluconate
Calcium lactate
1.63
1.05
1.07
1.06
1.03
1.00
?
1.00
1.00
1.18
1.69
2.43
1.00
1.27
1.01
1.01
1.00
1.00
1.00
1.00
Carnitine
Clofibrate
Diatrizoate
Indomethacin
Magnesium sulfate
Mannitol
Tin protoporphyrin
Tolazoline
Tromethamine
Testing not required: bicarbonate, cimetidine, dextran, enalapril, flumecinol, heparin,
ketamine, metoclopramide, naloxone, nicardipine, prostaglandin E1
Neuromuscular junction agents
Pancuronium
Testing not required: atracurium besylate, neostigmine, tubocurarine, vecuronium
Sedative and analgesic agents
Chloral hydrate
Paraldehyde
Pentobarbital
Thiopental
Testing not required: alfentanil, chlorpromazine, fentanyl, meperidine, midazolam,
morphine
Stimulants
Aminophylline
Doxapram
Sympathetic and parasympathetic agents
Edrophonium chloride
1.00
1.00
1.24
1.00
1.00
1.00
?
1.00
1.00
1.01
1.00
1.00
1.03
1.04
1.24
1.00
1.00
Testing not required: atropine dobutamine, dopamine, epinephrine, isoproterenol,

propranolol
Drugs are listed according to the category of their use. The symbol represents the maximal
displacement factor. If = 1.2, there is a 20% increase infree bilirubin concentration following drug
administration. Drugs listed as not requiring testing have low protein binding or low mean peak serum
concentrations. A question mark indicates that the drug requires testing but that the peroxidase
technique is not applicable.
From Robertson A, Carp W, Broderson R. Bilirubin displacing effect of drugs used in neonatology. Acta
Paediatr Scand 1991;80:11191127, with permission.
Other drugs used (or with potential for use) in the neonatal intensive care unit (NICU) have also been tested.
Aminoglycosides have minimal binding to albumin and should have no displacing effect. This has been confirmed
for gentamicin (84). The use of ibuprofen has been suggested as an alternative to indomethacin to close the
ductus arteriosus in premature newborns. Although indomethacin has some effect on bilirubin-albumin binding,
it does so only at plasma levels far greater than those that occur clinically (85). Using the peroxidase-diazo
method (68), Ahlfors found a significant increase in the unbound bilirubin concentration at serum ibuprofen
levels of 100 g/mL or greater. This effect was similar to that found with sulfisoxazole. On the first day of life,
serum bilirubin levels are very low, but ibuprofen has a long half-life and, over the next few days, could displace
bilirubin from its albumin binding and distribute it to the extravascular spaces and cause low bilirubin
kernicterus. Using a different technique, others have also found significant displacement of bilirubin by ibuprofen
(86). Following a dose of 10 mg/kg, plasma ibuprofen levels range from 181 g/mL on day 1 to 33 g/mL on
day 3 (87,88). These data argue against the use of ibuprofen in the sick, low-birth-weight infant who is already
at a greater risk for developing kernicterus.
Benzyl alcohol has been used as a bacteriostatic agent in multidose bottles. A significant association was found
between the use of benzyl alcohol and the development of kernicterus and intraventricular hemorrhage (89).
Benzyl alcohol is a membrane fluidizer (55) and might act by increasing the permeability of the blood-brain
barrier (90). On the other hand, benzoate, a metabolite of benzyl alcohol, is a powerful displacer of bilirubin and
leads to an increase in unbound bilirubin levels (91).
Robertson and associates (92) also evaluated the effect of drug combinations on bilirubin-albumin binding. This
is important because drug combinations are commonly administered to sick neonates, and the data show that
the bilirubin-displacing effect of these combinations cannot be predicted from each drug's effect. For example,
the administration of aminophylline with vancomycin increased the displacing effect when compared with either
drug alone, but the overall effect was still minimal. In the absence of published data, drugs should be selected
that have a low affinity for albumin and that have therapeutic concentrations much lower than the usual
concentration of albumin (about 2.8 g/dL in a very-low-birth-weight newborn). Simultaneous treatment with
several drugs should be limited as much as possible (92). Finally, as discussed in Blood-Brain and BloodCerebrospinal Fluid Barriers below, drugs that do not affect albumin binding of bilirubin may nevertheless affect
brain uptake of bilirubin by inhibiting P-glycoprotein (P-gp) function (93).
Other Competing Anions

It is possible that certain unidentified anions interfere with the albumin binding of bilirubin. Evidence for this is
suggested by the failure of exchange transfusion to alter significantly the albumin binding of infants' serum,
even though the exchange transfusion removes bilirubin, replaces much of the infant's serum albumin with
bilirubin-free albumin, and lowers the free bilirubin level (94). Anions that interfere with albumin binding might
be present in the extravascular compartment and might not be removed by exchange transfusion.
Clinical Status of the Infant

A relation may exist between bilirubin-binding capacity and the clinical condition and gestational age of the
infant. In some studies, very premature and sick infants were able to bind less bilirubin per mole of albumin
than were more mature infants (95,96). However, no relation was found between bilirubin-binding ability and
gestational age in other studies (97,98).
Finally, while a great deal of attention has focused on the effects of drugs on bilirubin binding and metabolism,
relatively little attention has been given to the effects of bilirubin on the pharmacokinetics, efficacy, and
toxicities of commonly used drugs.
Bilirubin and the Brain

Under normal circumstances, there is thought to be a constant movement of bilirubin in and out of the brain.
Experimentally, however, it is difficult to induce kernicterus or electrical central nervous system changes in
healthy animals by infusion of bilirubin, regardless of how much is infused. This may be related to the practical
difficulty of preparing homogenous infusates with high bilirubin concentrations, as well as the efficient
mechanisms for clearance of bilirubin in experimental animals (99,100). Gross staining of the brain and
electrophysiologic changes, however, occur readily in asphyxiated animals and in those subjected to
perturbations of the blood-brain barrier or of bilirubin-albumin binding (99,101,102,103,104,105,106).
Genetic factors may be involved in determining the susceptibility of patients to bilirubin-induced neurotoxicity.
For example, in different strains of Gunn rats exposed to similar bilirubin and albumin concentrations, there
were significant differences in susceptibility to kernicterus and mortality, suggesting that genetic or other factors
may be important in determining individual susceptibility to bilirubin toxicity (107).
Blood-Brain and Blood-Cerebrospinal Fluid Barriers

The blood-brain barrier (BBB) is the barrier that exists between the brain capillary endothelium and the brain
parenchyma; it limits the entry of certain substances into the central nervous system (CNS) (Fig. 35-9). The
choroid plexus is the barrier between the blood and the cerebrospinal
P.781
fluid (CSF) (67). The BBB consists of a continuous lining of endothelial cells connected by tight junctions that
restrict intercellular diffusion. The BBB normally excludes most water-soluble substances, proteins, and
macromolecules, but is permeable to low-molecular-weight lipid-soluble substances that are not highly protein
bound. Large molecules, such as albumin, are excluded from the brain but may enter when the BBB is made
permeable by the infusion of a hypertonic solution (62, 101,106). The endothelial cells of the choroid plexus lack
tight junctions and are fenestrated, allowing some proteins and their bound ligands to enter the CSF (67).
Although the total protein concentration of CSF in the newborn is only approximately 2% of the plasma protein
concentration, the ratios of TSB to protein are similar in CSF and plasma, suggesting that the steep
concentration gradient from plasma to CSF is the result of the much lower concentration of bilirubin binding
protein in the CSF (67).
Figure 35-9 Possible mechanisms for bilirubin entry into the brain and for binding to neuronal cell membranes.
The different factors affecting this process are indicated. A, albumin; AB, albumin-bilirubin complex; B=,
bilirubin monoanion; B=, bilirubin dianion; BBF, brain-blood flow. (From Bratlid D. How bilirubin gets into the
brain. Clin Perinatol 1990;17:449-465, with permission.)
P.782
Opening of the BBB allows albumin-bound bilirubin to bathe the neurons. The partitioning of sufficient bilirubin
from albumin to neuronal membranes will produce changes in the electroencephalogram.(62). Disruption of the
BBB and increased delivery of bilirubin to the brain may be important in the pathogenesis of bilirubin toxicity.
Protection of Cells from Neurotoxicity and Limiting Bilirubin Access

to the Brain.
Certain factorsincluding intracellular binding of unconjugated bilirubin to cytosolic proteins (67,108), neuronal
apoptosis inhibitor proteins (NAIPs), and bilirubin oxidation in brain cells (50)may protect cells of the central
nervous system during exposure to elevated concentrations of unconjugated bilirubin. Other factors play a role
in limiting bilirubin access to the brain. adenosine triphosphate binding cassette (ABC) transporters are proteins
that use adenosine triphosphate (ATP) to translocate substrates across biologic membranes (109) and they play
a role in limiting the accumulation of organic anions and other small molecules in the CNS and CSF (67). These
are multidrug resistance-associated proteins (MRPs) and multidrug resistance P-glycoproteins (MDR/P-gps). Pgp is a member of the ABC family of membrane transporters and is encoded for by a family of genes known as
the multidrug resistance (MDR) genes (MDR1 and MDR2 in humans; mdr1a, mdr1b, and mdr2 in rodents)
(67,109,110). P-gp is an ATP-dependent plasma membrane efflux pump expressed in brain capillary endothelial
cells and astrocytes of the blood-brain barrier. It limits the passage of lipophilic substrates and, possibly,
bilirubin, into the brain (110). Following a high-load bilirubin infusion, P-gp-deficient mice have higher brain
bilirubin content than do controls (110). Drugs that inhibit P-gp function increase brain uptake of bilirubin under
the same experimental conditions (93). P-gp and other transporters could have an important influence on the
deposition of bilirubin in the cells of the CNS (109), but their specific role has yet to be defined.
Lipid Solubility
Lipid-soluble substances that are not protein bound and gases, such as carbon dioxide and oxygen, cross the
BBB easily, by simple diffusion, whereas water-soluble substances, proteins, and polar compounds (i.e., ions) do
not.
Factors Affecting Blood-Brain Barrier Permeability

Anoxia, hypercarbia, and hyperosmolality open the BBB and increase the deposition of bilirubin and albumin in
the brain, producing neurophysiologic and biochemical changes as well as changes in brain physiology and
energy metabolism (see Fig. 35-9) (49,60,111). Thus, opening of the BBB is likely to be one important
mechanism in the pathogenesis of kernicterus, although other mechanisms undoubtedly exist. For example,
during. hypercarbia, the increased bilirubin that is deposited in the brain is predominantly in the unbound form,
although some is also albumin bound (69,104). The regional deposition of bilirubin in the brain of piglets occurs
in areas where hypercarbia produces the greatest increase in regional blood flow (104). Respiratory acidosis
increases bilirubin deposition in the brain but metabolic acidosis does not (103,104).
Effect of Maturity on Blood-Brain Barrier Permeability

Is the BBB of the neonate more permeable to bilirubin and albumin than in older children or adults? The
immature brain demonstrates greater passive permeability between the blood and central nervous system for
lipid-insoluble molecules (112). Studies in newborn piglets show that the BBB is more permeable to bilirubin in 2day-old than in 2-week-old piglets, whereas the permeability to albumin does not change (113). Other studies
have found that albumin transfer into the brain of newborn Gunn rats decreases with age (114).
Mechanisms by Which Bilirubin Enters the Brain

Bratlid (60) has provided a simplified scheme, not involving membrane transporters for bilirubin entry into the
brain, its binding to neuronal cell membranes, and the potential clinical signs that may follow (see Fig. 35-9).
Under normal circumstances, bilirubin can enter the brain unaccompanied by albumin. Clinical confirmation of
this fact is provided by the observation that even modest elevations of serum bilirubin can sometimes produce
clinical and electrophysiologic alterations in healthy full-term infants as demonstrated by changes in behavior
(115), characteristics of the cry (116), and changes in the brainstem auditory evoked response (BAER) (117).
These symptoms disappear as the bilirubin level decreases (117,118).
The probability of toxic levels of bilirubin entering the brain increases when the serum level of unbound pigment
increases, as depicted in the center panel of Fig. 35-9. Finally, if the BBB is disrupted, both albumin and bilirubin
can enter the brain; but even in the presence of a damaged BBB, more bilirubin than albumin, on a molar basis,
seems to be deposited (69,101,102,103). In all of these situations, acidosis would increase deposition of
bilirubin in brain cells.
Oxidation of Bilirubin in the Brain

There is some evidence that mitochondria in the brain and other tissues contain a bilirubin oxidase that converts
bilirubin to biliverdin and other nontoxic products (119, 120). The specificity and clinical relevance of this
putative enzyme remains to be established.
Clinical Features of Bilirubin

Encephalopathy
Terminology
Although originally a pathologic diagnosis characterized by bilirubin staining of the
brainstem nuclei and cerebellum, the term kernicterus has come to be used
interchangeably with both the acute and chronic findings of bilirubin
encephalopathy. Bilirubin encephalopathy describes the
P.783
clinical central nervous system findings caused by bilirubin toxicity to the basal
ganglia and various brainstem nuclei. To avoid confusion and encourage greater
consistency in the literature, the AAP recommends (76) that the term acute
bilirubin encephalopathy be used to describe the acute manifestations of bilirubin
toxicity seen in the first weeks after birth and that the term kernicterus be
reserved for the chronic and permanent clinical sequelae of bilirubin toxicity. This
is the terminology that is used in this chapter. Recently, Johnson and associates
suggested the use of the term bilirubin-induced neurologic dysfunction (BIND)
(121) to describe the changes associated with acute bilirubin encephalopathy; they
also proposed a scoring system to quantify the severity of the clinical
manifestations.
Acute Bilirubin Encephalopathy

In classic acute bilirubin encephalopathy, markedly jaundiced infants progress
through three fairly distinct clinical phases (Table 35-9) (51,122,123,124). In the
early phase, the infant becomes lethargic and hypotonic, and sucks poorly. The
intermediate phase is characterized by moderate stupor, irritability, and
hypertonia. The infant may develop a fever and a high-pitched cry (51), which
might alternate with drowsiness and hypertonia. The hypertonia involves the
extensor muscle groups, and most infants exhibit backward arching of the neck
(retrocollis) and trunk (opisthotonos) (Fig. 35-10). The fever may be caused by
diencephalic involvement. The advanced phase is characterized by pronounced
retrocollis-opisthotonos, shrill cry, refusal to feed, apnea, fever, deep stupor to
coma, and sometimes seizures and death (51,123,124).
Figure 35-10 This infant presented at age 30 days with a serum bilirubin level of
30 mg/dL (513 mol/L) secondary to the Crigler-Najjar syndrome type I. He
demonstrates retrocollis and opisthotonos, signs of the intermediate to advanced
stage of acute bilirubin encephalopathy.
TABLE 35-9 MAJOR CLINICAL FEATURES OF ACUTE BILIRUBIN

ENCEPHALOPATHY
Initial phase
Slight stupor (lethargic, sleepy)
Slight hypotonia, paucity of movement
Poor sucking; slightly high-pitched cry
Intermediate phase
Moderate stuporirritable
Tone variableusually increased; some with retrocollis-opisthotonos
Minimal feeding; high-pitched cry
Advanced phase
Deep stupor to coma
Tone usually increased; some with retrocollis-opisthotonos
No feeding; shrill cry
From Volpe JJ. Neurology of the newborn. 4th ed. Philadelphia: WB
Saunders, 2001, with permission.
Subsequently, usually after 1 week, hypertonia subsides and is replaced by
hypotonia. Infants who manifest hypertonia during the second phase almost
always develop the clinical features of chronic bilirubin encephalopathy (122,123)
although an emergent exchange transfusion might, in some cases, reverse the
CNS changes (125). Van Praagh (123) found that those who were consistently
neurologically normal during the first week of life never developed the features of
chronic encephalopathy, but other investigators found later evidence of brain
damage in some infants in whom no, or equivocal, manifestations of kernicterus

were apparent in the newborn (Table 35-10) (122,126,127).
TABLE 35-10 OCCURRENCE OF CLINICAL FEATURES IN ACUTE BILIRUBIN
ENCEPHALOPATHY
Clinical Features
No definite neurologic signs

Equivocal neurologic signs
Definite neurologic signs
Percent (%) of Cases
15
2030
5565
Adapted from Volpe JJ. Neurology of the newborn. 4th ed.

Philadelphia: WB Saunders, 2001; Van Praagh R. Diagnosis of kernicterus
in the neonatal period. Pediatrics 1961;28:870876; and Byers RK, Paine
RS, Crothers V. Extrapyramidal cerebral palsy with hearing loss following
erythroblastosis. Pediatrics 1955;15:248.
P.784
Kernicterus (Chronic Bilirubin Encephalopathy)

Temporal Evolution
There is a typical temporal evolution of chronic bilirubin encephalopathy after
neonatal erythroblastosis fetalis (128). In the first year, infants typically feed
poorly, develop a high-pitched cry, and are hypotonic, but have increased deep
tendon reflexes, a persistent tonic neck reflex, and motor delay. There is a delay
in acquisition of motor skills, although most infants walk alone by 5 years of age.
The other typical features of chronic bilirubin encephalopathy usually are not
apparent before 1 year of age and often not for several years (51). These children
generally are hypotonic at rest for the first 6 or 7 years. By the time they reach
their teens, hypertonia has replaced hypotonia (129).
Clinical Features
The classic sequelae of posticteric encephalopathy constitute a tetrad consisting of
extrapyramidal disturbances, auditory abnormalities, gaze palsies, and dental
dysplasia (Table 35-11) (129).
Extrapyramidal Disturbances
Athetosis (i.e., involuntary, sinuous, writhing movements) may develop as early as
18 months of age, but may be delayed until as late as 8 or 9 years of age (128). If
sufficiently severe, athetosis may prevent useful limb function. These movements
are described as uncontrollable, purposeless, involuntary, and incoordinate. They
may be rapid and jerky (choreiform), slow and worm-like (orthodox athetosis), or
so slowed by hypertonicity that the patient may assume momentarily fixed
attitudes with stiffness of the extremities (dystonia) (129). Occasionally,
extrapyramidal rigidity may predominate, rather than involuntary motion. In the
opinion of Perlstein (129), the absence of athetosis or of other forms of
extrapyramidal dyskinesia, makes the diagnosis of post-icteric encephalopathy
dubious, if not untenable. Severely affected children also may have dysarthria,
facial grimacing, drooling, and difficulty chewing and swallowing.
TABLE 35-11 MAJOR CLINICAL FEATURES OF CHRONIC
POSTKERNICTERIC BILIRUBIN ENCEPHALOPATHY
Extrapyramidal abnormalities, especially athetosis

Gaze abnormalities, especially of upward gaze
Auditory disturbance, especially sensorineural hearing loss
Intellectual deficits, but minority in mentally retarded range
From Volpe JJ. Neurology of the newborn. 4th ed. Philadelphia: WB
Saunders, 2001, with permission.
Auditory Abnormalities
Some degree of hearing loss is often found in children with kernicterus. Pathologic
studies and studies of BAERs indicate that injury to the brainstem, specifically to
the cochlear nuclei, is the principal cause of hearing loss, although occasional
studies suggest possible involvement of the peripheral auditory system as well
(39,73,117,130). It is noteworthy that in some frequently quoted studies, virtually
all of the infants who developed severe hearing loss had received prophylactic
streptomycin (an ototoxic antibiotic) prior to exchange transfusion (126,131).
Hearing loss is generally most severe in the high frequencies, and an association
between moderate hyperbilirubinemia and subsequent sensorineural hearing loss
has been described in low-birth-weight infants (see Clinical Sequelae of
Hyperbilirubinemia below).
Auditory neuropathy, or auditory dyssynchrony is a recently described entity that
is functionally defined as abnormal or absent BAER with normal inner ear function
as tested by cochlear microphonic responses or otoacoustic emissions (132). The

inner ear or cochlea is normal but the ascending auditory pathway in the nerve or
brainstem is abnormal (132). Shapiro has diagnosed auditory neuropathy in ten
children with classical kernicterus (132).
Gaze Abnormalities
There may be limitation of upward gaze and other gaze abnormalities although full
vertical eye movements during the doll's-eye maneuver are attained in most
affected children. This suggests that the lesion is above the level of the oculomotor
nuclei (51). Some patients have paralytic gaze palsies. Supranuclear palsies can
be explained by bilirubin deposition and neuronal injury in the rostral midbrain,
and nuclear palsies can be explained by damage to the oculomotor nuclei (39).
Dental Dysplasia
Approximately 75% of kernicterus children with posticteric encephalopathy have
some degree of dental enamel hypoplasia. A smaller percentage have green
discoloration of the teeth (129).
Magnetic Resonance Imaging

The diagnosis of acute bilirubin encephalopathy and kernicterus can now be
confirmed by magnetic resonance
P.785
imaging (MRI) (133,134,135,136 and 137). The most characteristic image is a
bilateral, symmetric, high-intensity signal in the globus pallidus seen on both T1and T2-weighted images (Fig. 35-11), although it should be noted that similar
findings in the globus pallidi have been described in mitochondrial disorders and in
pyruvate dehydrogenase deficiency (138). High signal intensity also has been seen
in the hippocampus and the thalamus (133). We do not yet know the natural
history or time course of the MRI changes of kernicterus, how early these changes
will occur, and whether or not they are seen in every case. They have been seen in
an 8-day-old infant as well as in a 12-year-old child with a history of severe
neonatal jaundice (133,134).
Figure 35-11 Magnetic resonance imaging scan of a 21-month-old male infant

who had erythroblastosis fetalis and presented with extreme hyperbilirubinemia
and clinical signs of kernicterus at age 54 hours. Note the symmetric, abnormally
high-intensity signal from the area of the globus pallidus on both sides (arrows).
(From Grobler JM, Mercer MJ. Kernicterus associated with elevated predominantly
direct-reacting bilirubin. S Afr Med J 1997;87:146, with permission.)
Clinical Sequelae of Hyperbilirubinemia

Although there is no doubt about the relationship between elevated TSB levels
and brain damage, the ability of a single peak bilirubin level to predict long-term
neurodevelopmental outcomes is poor (139,140). In addition, our conclusions
regarding hyperbilirubinemia and neurodevelopmental outcomes are limited by
the quality of the published data. In many case reports of kernicterus, analyzed in
detail by Ip and associates (139,140), it is impossible to know if the putative peak
TSB levels (measured in some cases more than 7 days after birth) were true peak
levels, and many cohort studies have significant problems with blinding of the
examiners and dropouts (139,140,141). For a detailed analysis of the published
literature from 1966 to 2001 dealing with these issues, see the evidence reports
of Ip and associates (139,140).
Hsia and colleagues (142) and Mollison and Cutbush (143) first established the
link between bilirubin levels and brain damage in the early 1950s, when they
demonstrated that the risk of kernicterus in infants with Rh hemolytic disease
increased dramatically with rising bilirubin levels and that exchange transfusion
could markedly reduce the risk. Subsequent studies suggested that, in untreated
infants with hemolytic disease, the incidence of kernicterus was much higher than
the incidence in markedly jaundiced infants without hemolytic disease (144,145).
For many years, there has been considerable disagreement about how jaundice in
infants without hemolytic disease should be treated (144,145,146,147,148,149).
Some investigators have described clinical and pathologic kernicterus in
premature infants who did not have hemolytic disease and whose TSB levels were
well below 20 mg/dL (342 mol/L) (150,151). In some small, sick infants, yellow
staining of the brain was seen at autopsy, with TSB levels less than 10 mg/dL
(171 mol/L) (151). In addition, the finding that gross bilirubin staining of the
brain can occur in the absence of the typical microscopic neuronal damage
described with kernicterus called into question the meaning of yellow staining of
the brain in some very-low-birth-weight infants (46). Unlike the classic form of
fatal kernicterus, these very-low-birth-weight infants, although they may have
had kernicterus, died of other causes.
Starting in 1967 and continuing into the late 1970s, reports from the Collaborative
Perinatal Project (CPP), a study of 53,000 pregnant women and their offspring,
linked moderate elevations of neonatal serum bilirubin to lower developmental
scores, lower IQ scores, and increased risk of neurologic abnormalities (152).
These findings occurred at TSB levels previously presumed to be safe and
suggested that acute bilirubin encephalopathy or classic kernicterus was only the
most obvious and extreme manifestation of a spectrum of bilirubin toxicity. At the
other end of the spectrum might lie more subtle forms of neurotoxicity that occur
at much lower bilirubin levels and in the absence of any obvious abnormal clinical
findings in the neonatal period; there are some data to support this view (140).
The following section deals with the clinical sequelae of hyperbilirubinemia in
different groups of infants.
Rh Hemolytic Disease of the Newborn

In the only controlled clinical trial conducted on the treatment of hemolytic
disease of the newborn, Mollison and Walker (153) demonstrated beyond
reasonable doubt that exchange transfusion in these infants improved their
chance of survival and decreased the risk of fatal kernicterus. The results of these
studies, combined with subsequent uncontrolled observations, established
exchange transfusion as the standard treatment for preventing
P.786
kernicterus in infants with erythroblastosis fetalis and showed that kernicterus
was unlikely to occur if TSB levels were kept below 20 mg/dL (342 mol/L), an
observation that has been amply confirmed by subsequent experience with the
treatment of hemolytic disease (142,143). It is interesting to recall, however, that
these studies were performed on infants born in the late 1940s and early 1950s.
These infants were commonly asphyxiated and seriously ill, and many were
delivered prematurely to prevent stillbirth. Streptomycin, an ototoxic drug was
frequently administered to those who received exchange transfusions.
Hsia and associates (142) reported that the incidence of kernicterus in their
infants was 8% for those with TSB levels of 19 to 24 mg/dL (325 to 410 mol/L),
33% for TSB levels of 25 to 29 mg/dL (428 to 496 mol/L), and 73% for those
with bilirubin levels higher than 30 mg/dL (513 mol/L). Later experience with Rh
hemolytic disease was far more encouraging. Johnston and associates (126)
studied 129 infants born between 1957 and 1958, all of whom had indirectreacting serum bilirubin levels higher than 20 mg/dL (342 mol/L). Ninety-two
infants with Rh hemolytic disease were followed to age 5 to 6 years and evaluated
with detailed psychometric, neurologic, and audiologic evaluations. One of 92
infants had a minimal sensorineural hearing loss (and a normal IQ) and 1 infant
had mild athetosis, moderate sensorineural hearing loss, and a normal IQ. Thus,
the risk of bilirubin encephalopathy in infants with Rh disease born from 1957 to
1958 with indirect-reacting bilirubin levels higher than 20 mg/dL (342 mol/L)
was 2 of 92 or approximately 2% (126).
In a study of full-term Turkish infants (154), those who had Coombs-positive ABO
incompatibility or Rh immunization had a greater risk (vs. infants without
hemolytic disease) of neurologic abnormalities and lower IQ scores when their
indirect-reacting bilirubin levels exceeded 20 mg/ dL (342 mol/L). These data
reinforce the belief that hemolysis is an important risk factor in bilirubindependent brain damage. The risk of neurologic abnormalities also was associated
with the duration of indirect-reacting hyperbilirubinemia higher than 20 mg/dL

(342 mol/L). In those exposed to these bilirubin levels for less than 6 hours, the
incidence of neurologic abnormalities was 2.3%. This increased to 18.7% if the
exposure lasted 6 to 11 hours and to 26% with 12 or more hours of exposure
(154). These data support earlier observations suggesting that the duration of
hyperbilirubinemia is related to the risk of long-term neurodevelopmental outcome
(155). It is difficult to define the risk of abnormal outcomes in infants with
hemolysis from causes such as red cell membrane defects and glucose-6phosphate dehydrogenase (G6PD) deficiency, although the risk of bilirubin
encephalopathy in G6PD deficiency appears to be similar to that of Rh hemolytic
disease (156).
Full-Term and Near-Term Infants Without

Hemolysis
Two issues in neonatal medicine that consistently generate controversy are the
relationship between hyperbilirubinemia and adverse developmental outcome in
nonhemolyzing newborns and the indications for treating these infants. These
issues are addressed in multiple studies and the reader is referred to extensive
reviews for details of the individual studies (124,139,140,144,152). When
carefully analyzed, the data tend to demonstrate that, in otherwise healthy
neonates without hemolytic disease, TSB levels that do not exceed approximately
25 mg/dL (428 mol/L) are very unlikely to place these infants at risk of adverse
neurodevelopmental consequences. Specifically, in many large studies there has
been no convincing demonstration of any adverse effect of such serum bilirubin
levels on IQs, definite neurologic abnormalities, or sensorineural hearing loss
(124,139,140,144,152).
No studies have looked specifically at infants who are 35 to 37 weeks of gestation,
although the data analyzed by Ip and associates includes infants of of 34 weeks
and older (139,140). In the very large CPP, the study population included all
infants with birth weights 2,500 g (145,152, 157). Presumably some of these
infants were in the 34- to 37-weeks' gestational age category. There are
insufficient followup data on infants who had TSB levels of 25 to 30 mg/dL (292 to
513 mol/L) to draw firm conclusions about this group. Nevertheless, a 21-month
followup of 26 Coombs-negative, healthy, term infants with TSB levels of 26.2 to
46.3 mg/dL (446 to 788 mol/L) revealed no neurodevelopmental abnormalities
and no hearing loss (158). Newman identified 11 infants out of 111,009 infants
born between 1995 and 1998 with TSB levels 30 mg/dL (513 mol/L); 5 were 35
to 37 weeks' gestation and 5 were 38 to 39 weeks' gestation. They were followed
for 1.5 to 5 years and none had neurologic or developmental problems (159).
When they combined both abnormal and suspicious neurologic examination
results, Newman and Klebanoff (160), in their analysis of the CPP, did
demonstrate a significant increase in abnormalities associated with increasing

bilirubin levels. The suspicious abnormalities included nonspecific gait
abnormalities, awkwardness, an equivocal Babinski reflex, abnormal cremasteric
reflex, abnormal abdominal reflex, failure of stereognosis, questionable hypotonia,
and gaze abnormalities. The most frequent abnormal findings were awkwardness
and abnormal cremasteric reflexes. When the abnormal and suspiciously abnormal
children were combined, the risk of abnormalities increased from 14.9% for those
whose TSB levels were less than 10.0 mg/dL (171 mol/L) to 22.4% for those
whose TSB levels exceeded 20 mg/dL (340 mol/L). Because 41,324 infants were
enrolled in this study, these differences were statistically highly significant, but
this finding should be kept in perspective. Even if the relationship between these
findings is causal, we have no evidence that the use of a bilirubin-lowering
intervention, such as phototherapy, at these low bilirubin levels would affect the
outcome. Finally, as Newman and Klebanoff (160) point out, even if bilirubin
levels had been prevented from exceeding 10 mg/dL (171 mol/L) in every infant,
the expected rate of abnormal or suspicious neurologic examination results would
only decline from 15.13% to 14.85%.
P.787
Less-reassuring information is provided by a study of a group of Israeli army
draftees (n = 1,948) in which their preinduction psychological and physical
examinations at age 17 years were matched with their newborn bilirubin levels.
Seidman and associates (161) found an association between the risk of an IQ
below 85 and a TSB level higher than 20 mg/dL (342 mol/L) in full-term boys
(but not with girls) with a negative Coombs test (p = 0.01). On the other hand, no
association was found between bilirubin levels and mean IQ score, the risk of
physical or neurologic abnormality, or hearing loss. In a subsequent analysis of a
similar population, Seidman and colleagues (162) found no increase in the risk of
an IQ less than 85 with TSB levels higher than 20 mg/dL (342 mol/L).
Two studies raised concerns regarding soft signs of neurologic dysfunction in
infants exposed to moderate levels of bilirubin (163,164). In a study performed in
the Netherlands, the investigators prospectively evaluated 20 jaundiced infants
whose TSB levels ranged from 13.6 to 26 mg/dL (233 to 444 mol/L) and
compared them with a control group of 20, healthy, nonjaundiced infants matched
for sex and gestational age (164). At age 1 year, 5 of 8 (63%) infants with TSB
levels between 19.6 and 26 mg/dL (335 and 444 mol/L) demonstrated minor
abnormalities in muscle tone and posture compared with 0 of 20 control infants
(p<0.001). In a German study, children at 7 years of age whose neonatal TSB
levels had exceeded 20 mg/dL (342 mol/L) scored significantly worse on a scale
designed to measure choreiform and athetoid movements. In that study 8 of 16
(50%) children in the hyperbilirubinemia group versus 3 of 18 (17%) in the
control group had abnormal scores (data not found in the original paper but kindly
provided by the authors) (141,163). The sample sizes in both the Dutch and
German studies were small, but the effect sizes were large. Nevertheless, the
outcome measurements are subjective and the blinding was not rigorous.
In a much earlier study, Johnson and Boggs followed 83 infants for 4 years and
found abnormal neurologic examinations in 14 of 68 (21%) children whose
indirect-reacting bilirubin levels were 15 mg/dL (257 mol/L) versus 0 of 15 in
those with TSB levels less than 15 mg/dL (155) (1 tail p = 0.047). Of the 14
children, 11 had minimal cerebral dysfunction and 3 had other abnormal signs,
including fine and gross motor delay, athetoid movements, and mild mental
retardation (it is not stated how many of the 3 infants had some or all of these
findings). In that study, however, 53% of the infants had hemolytic disease and
33% were premature, and there is no mention of whether or not the followup
evaluations were performed in a blinded fashion (155).
Comorbid Factors and Outcome

In term and near-term infants, there appears to be an increase in the likelihood of
neurologic damage if hyperbilirubinemia is accompanied by other risk factors
(43,139,140). The risk factors commonly listed include isoimmune hemolytic
disease, prematurity, G6PD deficiency, asphyxia, sepsis, acidosis, and
hypoalbuminemia (Table 35-12) (139,140).
In the CPP population, Naeye found that infants who were exposed to
chorioamnionitis prior to delivery had a higher incidence of low IQ scores and
neurologic abnormalities at age 7 years in relationship to increasing bilirubin levels
than did those infants who were not exposed to chorioamnionitis (165).
It is long been believed that infants with hemolytic disease are at a greater risk
for bilirubin-related brain damage than are those infants who do not have
hemolysis. This applies to Rh and ABO hemolytic disease, as well as to G6PD
deficiency. Notwithstanding the general agreement on this subject, the data
supporting this conclusion are not very strong, nor is there any documented
mechanism that explains why a serum bilirubin level of 20 mg/dL in an infant with
hemolytic disease should be more dangerous than the same level in an infant
without hemolysis. For example, there is no evidence of any difference in the
binding of bilirubin to albumin in infants with hemolytic disease as compared with
those who are not hemolyzing. Nevertheless, the consensus among experts is that
infants who have hemolytic disease should be treated more aggressively than
those who do not.
Bilirubin Binding Capacity, Kernicterus, and

Developmental Outcome
Because bilirubin that is free or loosely bound to albumin is more likely to cross
the blood-brain barrier (see Bilirubin Chemistry and Neurotoxicity and Blood-Brain
and Blood-Cerebrospinal Fluid Barriers above), our ability to predict the risk of
bilirubin encephalopathy might be improved by measurement of unbound bilirubin
or the reserve albumin binding capacity. A reduced albumin-binding capacity has
been associated with abnormal developmental outcome in some studies (155,166)
but not in others (167), although such associations have been found between free
bilirubin levels and abnormalities in the BAER (70,71,73,168). Elevated free
bilirubin concentrations have also been found in preterm infants with kernicterus
(44,72,77,169). Currently there are no bilirubin-binding tests in routine clinical
use in the United States, although a semiautomated peroxidase method has been
used in Japan (71,73) and a modification of this technique developed in the United
States (68). Nevertheless, there are no long-term studies of the developmental
outcome of infants in whom binding measurements have been obtained with this
technique.
Duration of Hyperbilirubinemia
A relationship has been described between neurologic and psychometric
abnormalities and the duration of exposure to TSB levels higher than 15 mg/dL
(154,155,166). Many
P.788
P.789
of the infants in these studies were either premature or had hemolytic disease. In
a Turkish study, exposure to TSB levels greater than 20 mg/dL (342 mol/L) for
less than 6 hours was associated with a 2.3% incidence of neurologic
abnormalities. This increased to 18.7% if exposure lasted 6 to 11 hours, and to
26% with 12 or more hours of exposure (154). In the large NICHHD collaborative
phototherapy trial, a 6-year followup of 224 control infants who did not receive
phototherapy and who had birth weights of less than 2,000 g showed no
association between IQ and duration of exposure to bilirubin (170).
TABLE 35-12 SUMMARY OF 88 CASE REPORTS OF TERM AND NEAR-TERM (GESTATIONAL AGE 34 WEEKS) INFANTS
WITH COMORBID FACTORS WHO HAD CLINICAL SIGNS OF ACUTE OR CHRONIC BILIRUBIN ENCEPHALPATHY OR
KERNICTERUS DIAGNOSED AT AUTOPSY
Diagnosis of Kernicterus
Comorbidity
ABO incompatibility
Rh incompatibility
Clinical
Acute Phase Acute Phase
Without
but Normal
Followup
Followup
12
27
19
31.68.2
(19.051.0)
Mean BW
SD
(Range), g
Gender
3,118680 6 females,
(2,270
4,313)
11 males 2
unknown,
32.17.1
3,063387 5 females, 4
(17.746.0)
(2,300 males, 24
3,969)
unknown
G6PD deficiency
10
0
1
2
13
31.88.5
3,353437 2 males, 11
(23.050.0)
(2,700 unknown
4,100)
Sepsis or infections
3
1
4
5
13
31.89.9
3,368586 7 females, 4
(14.547.8)
(2,580 males, 2
4,360)
unknown
Multiple conditions
3
0
4
3
10
29.116.1 2,913750 3 females, 4
(4.049.2)
(1,780 males, 3
3,686)
unknown
Total
25
3
48
12
88
31.69.0
3,155534 20 females,
(4.051.0)
(1,780 25 males, 42
4,360)
unknown
BW, birth weight; G6PD, glucose-6-phosphate dehydrogenase; TSB, total serum bilirubin.
a
Chronic
Sequelae
Mean Peak
TSBSD
(Range),
mg/dL
Autopsy Total(N)
33
Contain some missing data.
This infant had acute phase of kernicterus and chronic kernicterus sequelae and then died at the age of 19 months.
From Ip S, Chung M, Kulig J, et al. An evidence-based review of important issues concerning neonatal
hyperbilirubinemia. Pediatrics 2004;114:e130e153, with permission.
An 18-year followup of 55 boys with a history of neonatal hyperbilirubinemia (greater than 15 mg/
dL [257 mol/ L]) was performed in Norway at the time of military draft physical examinations
(171). Compared with the total cohort of Norwegian conscripts, there were no significant
differences revealed on physical examination or tests of vision, hearing, or IQ. However, seven
boys who had a history of positive Coombs tests and bilirubin in excess of 15 mg/ dL (257 mol/
L) for more than 5 days had significantly lower IQ scores than the national average.
Hearing Loss and Audiometric Evoked Responses

The BAER test is an accurate and noninvasive means of assessing the functional status of the
auditory nerve and the brainstem auditory pathway. Figure 35-12 shows the BAER tracing of a
normal full-term infant. The three positive waveforms labeled in the figure are those most easily
identified in the neonate (118). The latency for wave I represents the peripheral conduction time.
Latency of waves III and V and the interpeak latency of waves I to III, III to V, and I to V all
represent measurements of central conduction time. The interpeak latency I to V is referred to as
the brainstem conduction time. Reports also include amplitudes of the waveforms. These may
decrease or be lost in response to various insults.
Figure 35-12 A typical tracing of brainstem auditory evoked response has various components.
Wave I reflects the response of the peripheral auditory nerve; wave III reflects the superior
olive; waves IV to V reflect the inferior colliculus with peak and trough shown. Wave I peak to
waves IV to V trough (i.e., interpeak latency) reflects brainstem conduction time. (From Vohr BR.
New approaches to assessing the risks of hyperbilirubinemia. Clin Perinatol 1990;17:293-306,
with permission.)
Several studies have documented a relationship between TSB levels and the BAER (118,132,172),
and the acute changes seen in the BAER can be reversed by lowering the TSB level with
phototherapy or exchange transfusion (118). Abnormalities of the BAER are more closely related
to unbound bilirubin levels than to TSB level (70,71,73,168), but there are no studies relating
abnormalities in the BAER to long-term outcome.
Despite evidence for bilirubin damage to the auditory pathway in full-term and preterm infants
with classical kernicterus, there is virtually no evidence for a risk of hearing loss related to
hyperbilirubinemia in full-term infants who do not have hemolytic disease (152,161,170,173). In
a study of almost 17,000 children who received complete hearing and speech evaluations at 8
years of age, the incidence of sensorineural hearing loss in those who had TSB levels 20 mg/dL
(342 mol/L) or higher was 2.2%, and it was also 2.2% in those whose TSB levels were less than
20 mg/dL (342 mol/L) (152). In the NICHHD phototherapy study, the incidence of sensorineural
hearing loss in children followed to age 6 years was identical in the phototherapy and control
groups (1.8% vs. 1.9%) (167). Remarkably, in a followup of 36 children with the Crigler-Najjar
syndrome, none had evidence of sensorineural hearing loss (173).
Nevertheless, deficits in central hearing, speech, and language can occur in the absence of puretone hearing loss (174) and these may be manifestations of auditory neuropathy or dyssynchrony
(132). This recently described entity is defined as abnormal or absent BAER with normal inner ear
function as tested by cochlear microphonic responses or otoacoustic emissions (132) and has
been diagnosed in children with kernicterus (132).
Cry Analysis
An abnormal cry is a sign of neurologic distress and is associated with acute bilirubin
encephalopathy (40). Modest degrees of hyperbilirubinemia also affect the infant's cry (116,118).
Infant Behavior
Investigators have used the Brazelton Neonatal Behavioral Assessment Scale to evaluate the
effect of hyperbilirubinemia on infant behavior. Most studies show some effect, although several
are confounded by the use of phototherapy (118,140). Jaundiced infants score lower than controls
in habituation, orientation, motor performance, regulation of state, and autonomic stability (118).
Premature Infants and Low-Bilirubin Kernicterus

Kernicterus and Developmental Outcome
It is generally believed that premature infants are at greater risk of developing kernicterus or
bilirubin encephalopathy
P.790
than are full-term newborns exposed to similar bilirubin levels.
Watchko and Oski (150) provided a detailed historical review of kernicterus in preterm infants
from the 1950s, which has been updated (151). Reports during the years 1950 to 1965 suggested
that kernicterus or the clinical sequelae of hyperbilirubinemia were unlikely to develop if exchange
transfusions were used to maintain TSB levels below 18 to 22 mg/dL (308 to 376 mol/L) (150).
However, most of the infants in these studies were larger (1,250 to 2,500 g) and more mature
(28 to 36 weeks of gestation) than the extremely low-birth-weight infants currently seen in NICUs.
Between 1958 and 1972, a group of studies reported the occurrence of kernicterus at TSB levels
well below 20 mg/dL (342 mol/L) (150). In general, these infants were significantly more
premature and of much lower birth weight than those previously observed with kernicterus. Some
were exposed to sulfisoxazole, which previously was shown in a controlled trial to be a powerful
displacer of bilirubin from its binding to albumin (63,175). As a result of these findings, exchange
transfusion in preterm infants was recommended at TSB levels of less than 20 mg/dL (342 mol/
L). Publication of data from the CPP that suggested an association between impaired psychomotor
performance and TSB levels higher than 10 to 14 mg/dL (171 to 239 mol/L) in low-birth-weight
infants provided additional support for these recommendations (165,176).
In the NICHHD cooperative phototherapy study (1974 to 1976), infants were randomly assigned
to a control group that received no phototherapy or to a group that received phototherapy at
predetermined TSB levels. The criteria for exchange transfusion for all infants mandated exchange
transfusions at low levels of serum bilirubin (10 mg/dL [171 mol/L] in high-risk newborns with
birth weights less than 1,250 g) (177). Kernicterus was found in 4 of 76 autopsied infants whose
birth weights ranged from 760 to 1,270 g (47). Their peak TSB levels ranged from 6.5 to 14.2 mg/
dL (111 to 243 mol/L). All were asphyxiated or had hyaline membrane disease, and all had some
degree of periventricular-intraventricular hemorrhage (PIVH). Two had periventricular
leukomalacia (PVL) (47).
Surviving infants in the study were followed and evaluated at 6 years of age with the Wechsler
Verbal and Performance IQ test. No differences were found between the control and phototherapy
groups in the incidence of definite and suspect cerebral palsy, clumsy or abnormal movements,
hypotonia, or an IQ lower than 70. There were no differences between the two groups in growth,
speech, hearing loss, or evidence of hyperactivity (170).
Scheidt and colleagues (167) also published a 6-year followup of 224 control children from the
NICHHD study whose birth weights were lower than 2,000 g. None of these infants received
phototherapy, but bilirubin levels were maintained below specified levels by the use of exchange
transfusion. No relation was found between serum bilirubin levels and the incidence of cerebral
palsy, nor was there any association between maximal bilirubin level and IQ. IQ was not
associated with mean bilirubin level, time and duration of exposure to bilirubin, or measures of
bilirubin-albumin binding (167).
Two studies reviewed the risk factors previously suggested to predict the development of
kernicterus. They were unable to identify any risk factor or group of factors that was associated
with the development of kernicterus in the premature neonate, including birth weight less than
1,500 g, hypothermia, asphyxia, acidosis, hypoalbuminemia, sepsis, meningitis, drug therapy,
and TSB levels (178,179).
It is likely that there are some risk factors for the development of kernicterus that are unknown.
An excellent example of this possibility was the report from one NICU of an abrupt decrease in
kernicterus at autopsy in premature infants. The incidence of kernicterus fell from 31% to 0%
when the practice of flushing intravenous catheters with bacteriostatic saline that contained
benzyl alcohol was stopped (89). In an earlier study from the same NICU, the incidence of
kernicterus diagnosed postmortem among neonates of 25 to 32 weeks of gestation was a
remarkably high 25% (39). Benzyl alcohol is an agent that increases membrane fluidity and may
facilitate the passage of bilirubin into the brain (55). At the same institution, Watchko and
Claasen (160) found only three cases of kernicterus in 72 autopsies performed from 1984 through
1991 on newborns of less than 34 weeks of gestation who lived at least 48 hours. In the 69
newborns who did not have kernicterus, the peak TSB level ranged from 6.3 to 20.6 mg/dL (108
to 352 mol/L), and 56% had peak TSB values higher than those suggested for exchange
transfusion by the NICHHD phototherapy study guidelines (160). This sustained decrease in the
incidence of kernicterus confirms the experience in most nurseries that kernicterus in premature
newborns is now rare, although it has not disappeared completely.
Sugama and associates documented the surprising observation of the presence of hypotonia and
(in one infant) choreoathetosis together with the classical MRI findings of kernicterus in two
preterm infants at 31 and 34 weeks gestation (136). Neither of these infants was acutely ill and
their TSB levels were 13.1 mg/dL (224 mol/L) and 14.7 mg/dL (251 mol/L), respectively (136).
Govaert and associates reported MRI findings in five preterm and three term infants with
kernicterus. The TSB levels in the term infants ranged from 37.0 to 44.6 mg/dL (632 to 763
mol/ L), but in the preterm infants (25 to 29 weeks' gestation) peak TSB levels ranged from 8.7
to 11.9 mg/dL (148 to 204 mol/L). Serum albumin levels in these preterm infants were strikingly
low (1.4 to 2.1 g/dL).

Several recent studies have failed to document an association between maximal TSB levels and
developmental outcomes in very-low-birth-weight infants (180,181,182,183). van de Bor and
colleagues (184) found a relation between maximal TSB concentrations in the neonatal period and
cerebral palsy (not of the type characteristically found with kernicterus) at a corrected age of 2
years. No relation was found between maximal TSB concentrations and hearing defects. However,
in a followup of the same population at 5 years, no significant difference was found in mean
maximal TSB concentrations between children with and without
P.791
handicaps (180) although there was an association between TSB levels and handicap in children
who had suffered a grade I intracranial hemorrhage. This effect was not seen in the more severe
hemorrhages, but the number of infants with severe hemorrhages was small.
Although these data are reassuring, few of the infants in the reported studies had markedly
elevated bilirubin levels (180,181,182,183). Most recently, however, peak TSB levels were
associated with an increased risk of death, hearing loss and neurodevelopmental impairment in
extremely low-birth-weight (ELBW) infants born between 1994 and 1997 (185). Although these
associations were statistically significant, the effect sizes were very small. The data in all of these
observational studies must be treated with caution because of the multifactorial nature of the
causes of adverse neurologic sequelae, the possibility of compounding effects of variables on
outcomes, and the difficulty inherent in fully controlling for these variables in such studies.
Some studies suggest an association between hyperbilirubinemia and cystic PVL in low-birthweight infants (186,187,188), but others have not found this association (183). Despite the
associations described (all from multiple significance testing with the resultant possibility of
spurious conclusions), it is unlikely that hyperbilirubinemia is causally related to cystic PVL. PVL is
primarily an ischemic lesion, most likely caused by hypoperfusion of the periventricular white
matter. Bilirubin normally is not deposited in the periventricular region and primarily is toxic to
neurons and not the glial elements that predominate in the periventricular white matter.
Significant associations between peak TSB levels and hearing loss have been documented in lowbirth-weight (LBW) infants (189,190,191) and confirmed recently in a population of ELBW infants
(185).
Because of the known role of bilirubin as a powerful antioxidant and the fact that more than 90%
of ELBW infants receive phototherapy, the NICHHD neonatal research network has initiated a
prospective randomized trial to compare aggressive with conservative phototherapy in a
population of ELBW infants (192) (see section on Treatment). The results of this study should
provide important information about the association between low TSB levels and developmental
outcomes in ELBW infants, as well as the risks and benefits of phototherapy in this highly
vulnerable population.
Thus, although kernicterus is currently a very rare event in premature infants hospitalized in
NICUs, it is still occurring. Data from the NICHHD trial that is in progress at the time of this
writing should be of immense help in guiding our management of the jaundiced ELBW infant (192).
Until further information is available, there seems good reason to believe that the very-low-birthweight (VLBW) infant's brain is more susceptible to damage from a number of sources and, given
that preterm infants have lower serum albumin levels and less-effective albumin binding and are
much more likely to be sick than are full-term infants, it makes sense to take a more aggressive
approach to maintaining low bilirubin levels in this population (137,185).
Hyperbilirubinemia and Pulmonary Hemorrhage

Studies from the late 1940s and early 1950s suggested an association between pulmonary
hemorrhage and kernicterus (90,193,194). All of the infants described died of severe
erythroblastosis fetalis. We know that these infants were profoundly anemic, had marked
hypoalbuminemia, thrombocytopenia, and disturbances of coagulation, and occasionally were
hydropic. It is not surprising that some of these infants developed hemorrhagic pulmonary
edema. In an autopsy series of low-birth-weight infants with kernicterus, pulmonary hemorrhage
was not found more frequently in the kernicteric infants when compared with those who did not
have kernicterus (63). In animal studies, the infusion of bilirubin led to circulatory collapse and
anoxia and an increased incidence of hemorrhage, including pulmonary hemorrhages (100). On
the other hand, kernicteric Gunn rats showed a significantly lower prevalence of pulmonary
hemorrhage when compared with nonkernicteric animals, although the kernicteric group had
significantly higher bilirubin levels (195).
The Clinical Approach to the Jaundiced Newborn

Epidemiology of Neonatal Hyperbilirubinemia
An important first step in the diagnosis and manaWgement of any jaundiced newborn is an
understanding of the factors that normally affect neonatal bilirubin levels (Table 35-13). Some of
these factors have been identified only in large epidemiologic studies and their clinical relevance is
questionable, but there are some (designated by asterisks in Table 35-13) that repeatedly have
an important influence on TSB levels.
Genetic, Ethnic, and Familial Influences

East Asian and Native American infants have mean maximal TSB concentrations that are
significantly higher than those of white infants (196,197,198,199,200 and 201). Increased
bilirubin production appears to be one factor contributing to the hyperbilirubinemia in these
infants (200). In a Hispanic (primarily Mexican) population, 31% of infants had TSB levels greater
than 15 mg/dL (202) compared with 3% to 10% of infants in other populations (203,204). Black
infants in the United States and Great Britain have lower TSB levels than white infants
(197,198,205,206). Neonatal jaundice runs in families (207,208). In a study of 3,301 infants,
Khoury and associates (207) found that if a previous sibling had a TSB level higher than 12 mg/dL
(205 mol/L) or higher than 15 mg/dL (257 mol/L), the risk of similar TSB levels in subsequent
siblings was 3.1 and 12.5 times greater, respectively, than in siblings of infants who did not have
that degree of jaundice.
TABLE 35-13 EPIDEMIOLOGY OF NEONATAL JAUNDICE
Associated Factors
Race
Genetic or familial
Maternal
Effect on Neonatal Serum Bilirubin Levels

Increase
Decrease
No Effect
East Asian*
Native American
Greek
Hispanic (Mexican)*
Previous sibling with
jaundice*
Primipara (?)
Maternal age 25 years
Diabetes (if infant
macrosomic)
Hypertension
Oral contraceptive use at
time of conception
First-trimester bleeding
Decreased plasma zinc
level
African American*
Smoking
Drugs administered
Oxytocin
Phenobarbital
to mother
-Adrenergic
agents
Diazepam
Meperidine
Epidural anesthesia
Reserpine
Promethazine
Aspirin
Chloral hydrate
Heroin
Labor and delivery
Infant
Premature rupture of
Phenytoin
Fetal distress
membranes
Antipyrine
Alcohol
Low Apgar scores
Forceps delivery
Vacuum extraction*
Breech delivery
Decreasing gestation*
Male gender*
Gestation41 weeks*
Delayed cord clamping

Elevated cord blood
bilirubin level
Jaundice observed before
discharge
Predischarge TSB level in
higher risk zones (Fig.
35.20)
Cephalhematoma or
Formula feeding*
bruising*
Delayed meconium passage
Breast-feeding*
Drugs administered
to infant
Other
Caloric deprivation*
Larger weight loss after
birth*
Low serum zinc and
magnesium
Chloral hydrate
Pancuronium
Altitude
Short hospital stay after
birth*
* Most relevant clinical factors.
P.792
Maternal Factors
Smoking
Some studies suggest that infants of mothers who smoke during pregnancy have lower serum
bilirubin levels than infants of nonsmokers (197,209), but others have not found this (210,211).
These data are confounded by the fact that women who smoke are much less likely to breastfeed, and the likelihood of breast-feeding is inversely related to the number of cigarettes smoked
per day (212).
Diabetes
Macrosomic infants of insulin-dependent diabetic mothers are more likely to become jaundiced
than are control
P.793
infants (213). This most likely is the result of an increase in bilirubin production, which is directly
related to the degree of macrosomia in these infants (214). These infants have high
erythropoietin levels and evidence of increased erythropoiesis, so that ineffective erythropoiesis
and polycythemia probably are responsible for the increased bilirubin production (215,216). In
addition, diabetic mothers have three times more -glucuronidase in their breast milk than
nondiabetic mothers (216). This enzyme enhances the enterohepatic reabsorption of bilirubin (see
Breast-Feeding and Jaundice below).
Events During Labor and Delivery

Induction and Augmentation of Labor by Oxytocin.
Multiple studies and several controlled trials have shown an association between the use of
oxytocin to induce or augment labor and an increased incidence of neonatal hyperbilirubinemia,
although the mechanism for this is unclear (217,218).
Anesthesia and Analgesia.

Several studies associate epidural anesthesia, specifically, bupivacaine, with neonatal jaundice
(211,219,220). These agents readily cross the placenta and produce measurable levels in the new
born (221).
Other Drugs.
Tocolytics did not affect neonatal carboxyhemoglobin levels or the need for phototherapy
(222,223).
The administration of narcotic agents, barbiturates, aspirin, chloral hydrate, reserpine, and
phenytoin sodium to mothers was associated with lower TSB concentrations in their infants,
whereas the use of diazepam increased TSB levels by less than 1 mg/dL (224). Antipyrine
administered to the mother before delivery decreased TSB levels, and infants of heroin-addicted
mothers have lower TSB levels (225). Phenobarbital, if given in sufficient doses to the mother,
significantly lowers TSB levels during the first week (217,226).
Delivery Mode.
Vaginally delivered term newborns have higher TSB levels than those delivered by cesarean
section (227), although this was not found in a controlled trial involving low-birth-weight infants
(228). When compared with forceps delivery, the use of vacuum extraction does not increase the
number of babies who require phototherapy, although more clinical jaundice is seen with vacuum
extraction (229,230).
Placental Transfusion and Hyperviscosity.

Although a high hematocrit often is considered a risk factor for neonatal jaundice, controlled trials
of an intervention for infants with symptomatic hyperviscosity using partial exchange transfusions
showed no differences in the incidence of hyperbilirubinemia in the treated and control groups
(231,232,233). In one study, infants were held 30 cm below the introitus after delivery. If cord
clamping was delayed, the mean TSB level at age 72 hours was 7.7 mg/dL (132 mol/ L)
compared with 3.2 mg/dL (55 mol/L) in the early clamped group (234).
Cord Blood Bilirubin Levels.

More than 50 years ago, Davidson and associates (235) found an association between bilirubin
levels in cord blood and later neonatal bilirubin concentrations. This observation has been
confirmed in infants with and without hemolytic disease (236,237,238).
Neonatal Factors
Birth Weight and Gestation
Low birth weight and decreasing gestational age are strongly correlated with an increased risk of
hyperbilirubinemia (197,206,211,239,240 and 241). So-called near-term infants of 35 to 38
weeks' gestation are at significantly greater risk of hyperbilirubinemia than are full-term infants
(206,211,239,241). Compared with infants at 40 weeks' gestation, infants of 36 to 38 weeks of
gestation are 7 to 8 times, and those who are less than 36 weeks are 13 times, more likely to be
readmitted to the hospital with severe hyperbilirubinemia (241).
Gender
As a group, male infants consistently have higher bilirubin levels than females (206,211,240,241
and 242).
Caloric Intake and Weight Loss

Decreased caloric intake is associated with an increase in serum bilirubin in animals and humans
(32). The primary mechanism responsible for this appears to be an increase in the enterohepatic
circulation of bilirubin (32,33). A significant association also exists between hyperbilirubinemia
and weight loss in the first few days after birth (211, 240,241 and 242).
Type of Diet
Infants fed a casein-hydrolysate formula had significantly lower TSB levels from days 10 through
18 than did those infants fed standard casein or whey-predominant formulas (243). The
cumulative stool output of the infants fed the casein-hydrolysate was lower than that of the
infants fed the other formulas, suggesting that factors other than stool output and its effect on
the enterohepatic circulation must explain these observations.
Breast-Feeding and Jaundice

Multiple studies over the last 25 years have found a strong association between breast-feeding
and an increased incidence of neonatal hyperbilirubinemia (Figs. 35-13 and 35-14) (244).
Although occasional studies have not found this (245,246), a pooled analysis of 12 studies of
more than 8,000 newborns showed that breast-fed infants were three times more likely to
develop TSB levels of 12 mg/dL (205 mol) or higher and six times more likely to develop levels
of 15 mg/dL (257 mol) or higher than formula-fed infants (Fig. 35-14) (247). Ninety percent or
more of
P.794
infants readmitted to hospital in the first 2 weeks of life because of severe hyperbilirubinemia are
fully or partially breast-fed (239,241,248,249). In a northern California population, exclusively
breast-fed infants were six times more likely than formula-fed infants to develop a TSB greater
than 25 mg/dL (428 mol/L) (206). Of 61 term and near-term infants with kernicterus, 59 were
breast-fed (124) (the 2 formula-fed infants were G6PD deficient).
Figure 35-13 Distribution of maximum serum bilirubin concentrations in 1,260 breast-fed and
1,026 formula-fed white newborns with birth weights >2,500 g. TSB levels were measured in
every infant on the second or third hospital day and repeated if the concentration exceeded 12.9
mg/dL (221 mol/L). (From Maisels MJ, Gifford K. Normal serum bilirubin levels in the newborn
and the effect of breast feeding. Pediatrics 1986;78: 837-843, with permission.)
There is some debate about the studies listed above and the data illustrated in Fig. 35-13 (250).
Some studies show no differences in TSB levels between formula-fed and breast-fed infants in the
first few days after birth. Bertini and associates (246) studied infants in their well baby nursery.
Formula supplementation was given to breast-fed infants if a weight loss was considered to be
excessive (4% after 24 hours, 8% after 48 hours, or 10% after 72 hours). The investigators
found a positive correlation between TSB levels greater than 12.9 mg/dL (221 mol/L), weight
loss after birth, and breast-feeding requiring supplementation with formula. Breast-feeding, per se
was not associated with hyperbilirubinemia. They concluded that infants who are successfully
breast-fed and therefore lose little weight, are not more likely to be jaundiced than formula-fed
infants, whereas those who required formula supplementation because of excess weight loss were
more likely to be jaundiced. This supports an important role for caloric intake in the development
of jaundice and has led some experts to categorize jaundice associated with breast-feeding in the
first days after birth as starvation jaundice or breast-nonfeeding jaundice (250). The
implication is that if breast-fed infants were nursed effectively from birth, they would not be more
jaundiced than formula-fed infants and there is evidence to support this view (246,251,252).
Figure 35-14 Pooled analysis of 12 studies showing the percent of newborns with serum
bilirubin levels 12mg/dL (205 mol/L) in breast-fed and formula-fed newborns and, in 6 of the
12 studies, the percent of newborns with serum bilirubin levels 15 mg/dL (256 mol/L) (From
Schneider AP. Breast milk jaundice in the newborn. A real entity. JAMA 1986;255:3270-3274,
with permission.)
P.795
Jaundice associated with breast-feeding in the first 2 to 4 days of age has been called the breastfeeding jaundice syndrome or breast-feeding-associated jaundice, and that which appears later
(onset at 4 to 7 days of age and prolonged jaundice) has been called the breast milk jaundice
syndrome (253). There is considerable overlap between these two entities, and evidence to
support two distinct syndromes is meager, although calorie deprivation appears to play a major
role in the early type of jaundice. In addition to having higher TSB levels in the first 3 to 5 days
(see Fig. 35-13) (254), as a group, breast-fed infants have TSB levels that are higher than
formula-fed infants for at least 3 to 6 weeks (243,255,256). These are the same infants who have
high bilirubin levels in the first week of life, and it is hard to believe that those who are still
jaundiced at age 2 to 3 weeks represent a distinct group.
Prolonged indirect-reacting hyperbilirubinemia (beyond 2 to 3 weeks) occurs in 20% to 30% of all
breast-feeding infants and, in some infants, may persist for up to 3 months (257). TSB levels
were measured at age 28 to 33 days in 282 healthy, breast-fed Turkish infants that were 37
weeks' gestation or older. The TSB was greater than 5 mg/dL (85 mol/L) in 20.2% and greater
than 10 mg/dL (171 mol/L) in 6% of the infants. (258) Recent evidence suggests that in some
infants mutations of the UGT1A1 gene (Gilbert syndrome) play a role in the pathogenesis of this
hyperbilirubinemia (259,260).
Pathogenesis of Jaundice Associated with Breast-Feeding.

Factors that play a role in the pathophysiology of jaundice associated with breast-feeding are
reviewed in detail elsewhere (28,244,253) and are listed in Table 35-14. The primary mechanism
for jaundice in the breast-fed infant appears to be an increase in the enterohepatic circulation of
bilirubin. Some authors suggest an additional role for inhibitory substances in breast milk, but the
data regarding their contribution are conflicting (244).
Intestinal Reabsorption of Bilirubin.Intestinal reabsorption of bilirubin (the enterohepatic
circulation) appears to be the most important mechanism responsible for the jaundice associated
with breast-feeding (244). Breast-fed infants take in fewer calories than formula-fed infants in the
first days after birth and a relationship between decreased caloric intake and an increase in the
enterohepatic circulation of bilirubin has been shown (32,33). Breast-fed infants produce lowerweight individual stools, their cumulative stool output (by weight) is lower (243), and their stools
contain less bilirubin than those of formula-fed infants (28,261). An increase in stool excretion in
the first 21 days is associated with lower TSB levels and in the first 3 weeks, infants fed human
milk pass significantly less stool than do infants who are fed casein-predominant formulas (243).
Infants fed casein-hydrolysate formulas pass less stool, cumulatively, than those given whey- or
casein-predominant formulas (243).
TABLE 35-14 PATHOGENESIS OF JAUNDICE ASSOCIATED WITH BREAST-FEEDING

Decreased caloric intake
Less cumulative stool output and stools contain less bilirubin (compared with formulafed infants)
Increased intestinal fat absorption
Less formation of urobilin in gastrointestinal tract
Increased activity of -glucuronidase in breast milk
Mutations of the UGT1A1 gene (Gilbert syndrome)prolonged breast milk jaundice
The relationship between fecal bilirubin excretion and TSB levels may be related to the fecal
excretion of unabsorbed fat (262). Unconjugated bilirubin apparently associates with unabsorbed
fat in the intestinal lumen. When Gunn rats were fed orlistat, a substance that inhibits lipase, they
excreted more fat in their stools and their TSB levels were significantly lower (31). This suggests
that a substance that increases fecal excretion of fat will decrease the enterohepatic absorption of
unconjugated bilirubin and facilitate bilirubin excretion in the gut. Breast-fed infants have higher
fat absorption than formula-fed infants (possibly related to the presence of bile salt-stimulated
lipase in human milk [263]). It is possible that hyperbilirubinemia could be prevented or mitigated
by the administration of orlistat to newborns (31,262). All of these findings support a major role
for the enterohepatic circulation in the jaundice associated with breast-feeding.
Urobilinogen Formation.
In adults, bilirubin in the gut is reduced rapidly by the action of colonic bacteria to urobilinogen.
At birth, the fetal gut is sterile, and although there is an increase in the bacterial content of the
gut after delivery, the neonatal intestinal flora do not convert conjugated bilirubin to urobilin. This
leaves bilirubin in the bowel and allows it to be deconjugated and thus available for reabsorption.
Formula-fed infants excrete urobilin in their stools earlier than breast-fed infants do, perhaps as a
consequence of the effect of formula feeding on the intestinal flora (264). Thus, the effect of
breast milk on intestinal flora, by slowing the formation of urobilin, further enhances the
possibility of intestinal reabsorption of bilirubin.
-Glucuronidase.
-Glucuronidase is an enzyme that cleaves the ester linkage of bilirubin glucuronide, producing
unconjugated bilirubin, which can then be reabsorbed through the gut. Significant concentrations
of -glucuronidase are found in the neonatal intestine, and its activity is higher in human milk
than in infant formulas (244).
Gourley and Arend (265) found a positive relation between TSB levels and breast milk glucuronidase activity in the first 3 to 4 days after birth, but other researchers have not been able
to confirm these findings (257,266).
P.796
Meconium Passage
Because the enterohepatic circulation of bilirubin is an important contributor to neonatal
hyperbilirubinemia, increasing the rate of bilirubin evacuation from the bowel should decrease the
incidence of neonatal jaundice. Two randomized studies showed that the early passage of
meconium (stimulated by a rectal thermometer or a suppository) reduced peak TSB levels by
about 1 mg/dL (17mol) when compared with control groups (267,268).
Phenolic Detergents
The use of phenolic detergents to disinfect incubators and other nursery surfaces was associated
with an epidemic of neonatal hyperbilirubinemia in two hospitals (269,270). These detergents
should not be used in the nursery.
Altitude
Infants born 3,100 m above sea level are four times more likely to have a bilirubin level above 12
mg/dL (205 mol/L) than are those born at sea level (271). Both short- and long-term exposure
to high altitudes increases TSB levels in adults. The possible mechanisms for these observations
include an increase in bilirubin load because of high hematocrits and impaired conjugation and
excretion of bilirubin (272,273,274).
Drugs Administered to the Infant

The use of pancuronium and chloral hydrate in the neonate are associated with an increased risk
of hyperbilirubinemia (275,276,277). Chloral hydrate is metabolized to trichloroa- cetic acrid and
the toxic trichloroethanol, both of which accumulate in the tissues of compromised infants. The
administration of chloral hydrate is associated with both indirect-reacting and direct-reacting
hyperbilirubinemia (276).
Free-Radical Production
Bilirubin appears to have an important physiologic function as an antioxidant and may play a role
in the prevention of oxidative membrane damage in vivo (see Physiologic Role of Bilirubin below)
(278). Infants with circulatory failure, sepsis, aspiration syndromes, and asphyxiaconditions
believed to enhance free-radical productionhad a significantly lower daily rise in mean TSB
levels than control infants (279). These finding are consistent with the hypothesis that bilirubin is
a free-radical scavenger and is consumed as an antioxidant.
Jaundice in the Healthy Newborn

Cord Bilirubin Levels
Mean bilirubin levels in cord blood range from 1.4 to 1.9 mg/dL (24 to 32 mol/L) (235,280,281),
and elevated cord bilirubin levels are associated with an increased risk of hyperbilirubinemia
(235,237,280,281).
Normal Serum Bilirubin Levels and the Natural History of

Neonatal Jaundice
It is difficult to agree on what represents a normal bilirubin level in the term and near-term
infant. TSB levels vary considerably, depending on the racial composition of the population, the
incidence of breast-feeding, and other genetic and epidemiologic factors (see Table 35-13 and Fig.
35-15). An additional important factor is the large variation found in the laboratory measurements
of serum bilirubin, a problem that has been recognized for four decades and currently is
unresolved (282,283,284) (see Laboratory Measurements of Bilirubin below).
Reference to Fig. 35-15 gives some idea of the range of mean bilirubin values found in different
populations and the natural history of neonatal jaundice. Since the advent of exchange
transfusion and, more recently, phototherapy, it has been impossible to obtain a true picture of
the natural history of neonatal bilirubinemia because we treat some infants with rising TSB levels
in the first 72 to 96 hours so that what we see is, ultimately, a damped picture. Recognizing
these limitations, however, some recent studies have helped to clarify the picture (285,286).
An important change in the United States population has been the dramatic increase in breastfeeding at hospital discharge, from 30% in the 1960s to 65% in 2001 (287). In some hospitals,
85% or more of mothers are nursing their
P.797
infants on discharge from the hospital. Data from the CPP conducted from 1955 to 1961 (when
30% or fewer mothers breast-fed their infants) and more recent studies (254) found that
approximately 95% of all infants had a TSB concentration that did not exceed 12.9 mg/dL (215
mol/L), and this (95th percentile) became the accepted upper limit of physiologic
jaundice (205).
Figure 35-15 Mean total daily bilirubin concentrations in normal full-term and near-term
infants. [black down-pointing triangle], Fifty healthy Japanese newborn infants, 37 to 42 weeks
of gestation, all breast-fed. Excludes Rh and ABO incompatibility (292). ^, One hundred seventysix term breast-fed Canadian infants. Excludes Rh hemolytic disease, but includes nine ABO
incompatible infants with positive Coombs tests. Seventeen infants received phototherapy. ^,
One hundred sixty-four Canadian term formula-fed infants, seven ABO incompatible with positive
Coombs tests, and three who received phototherapy (280). , One thousand eighty-seven term
Israeli infants, 78% fully or partially breast-fed (Siedman D, personal communication, 1998) ^,
Fifty-six Nigerian term appropriate for gestational age infants. Excludes ABO or Rh
incompatibility and G6PD deficiency. Infants were largely breast-fed (291). (), Twenty-nine
full-term American infants, all formula-fed, approximately 50% African American and 50% white
(30).
More recent data suggest that these values no longer define normal TSB levels in the newborn
population (203,288,289). We now see more jaundiced babies, and the TSB levels found in the
normal population are significantly higher than previously reported. Three recent studies provide
consistent information regarding the upper limits of TSB levels found in the normal population. In
a study of 2,840 infants, all of whom had at least one TSB level measured after discharge from
hospital, the 95th percentile was a level of 17.5 mg/dL (300 mol/L) (289). This population was
43% white, 41% black, and 4% Asian; 59% of infants were fully or partially breast-fed. In 11
Kaiser Permanente Northern California Hospitals, the 95th percentile was a TSB level of 17.5 mg/
dL (298 mol/L) (203). In a multicenter study of infants 36 weeks or older in nurseries in the
United States, Hong Kong, Japan, and Israel, 2 SD above the mean for the peak TSB levels at 96
6.5 hours was 17 mg/dL (291 mol/L), and the 95th percentile was 15.5 mg/dL (265 mol/L)
(290). The consistency of these data suggests that we can now accept that the upper limit of
normal in diverse populations is a TSB level of about 17 to 18 mg/dL (291 to 308 mol/L). This
implies that a 4- to 5-day-old breast-fed infant whose TSB level is 15 to 16 mg/dL (291 mol/L)
does not require any laboratory investigation to find out why the infant is jaundiced, although
followup is necessary to ensure that the bilirubin levels do not become excessive (289). Data from
studies of predominantly breast-fed infants suggest that the normal mean peak TSB level is
approximately 8 to 9 mg/dL (137 to 154 mol/L) (240,280,290,291). In the international,
multicenter study, the mean TSB level at 96 6.5 hours was 9.3 mg/dL (290).
In the Japanese population (292) and in other populations of predominantly breast-fed infants, it
is clear that the TSB levels are substantially higher, reach their peak later, and remain elevated
for much longer than in formula-fed infants. No significant decline in TSB is seen in any of these
populations until after the fifth day (see Fig. 35-15).

Figure 35-16 shows idealized smoothed curves based on the data from a number of studies that
provide a guide to the expected course of bilirubin levels in a primarily breast-fed (60% to 70%)
western population (203,235,240, 242,280,289,290). Recognizing all of the limitations to these
data already discussed, Fig. 35-16 should be useful for plotting the course of jaundice in an infant
and using the velocity of the increase in TSB to make decisions about evaluation, follow-up, and
potential intervention.
Bhutani and associates developed a nomogram that defines predischarge risk zones for the
subsequent development of hyperbilirubinemia (289) (see Preventing Extreme Hyperbilirubinemia
and Kernicterus below). Because of sampling bias, however, this nomogram probably does not
describe the natural history of bilirubinemia in the newborn (293), although it is a very useful tool
for predicting the risk of subsequent hyperbilirubinemia (289) and is recommended by the AAP for
this purpose.
Figure 35-16 Smoothed curves from studies in diverse populations illustrating the expected
velocity of total serum bilirubin (TSB) levels and approximate values for the 50th and 95th
percentiles. Data for cord blood values come from the studies of Davidson and associates (235)
and Saigal and associates (280), values in the first 12 hours are from Frishberg and associates
(242), and subsequent values are from Bhutani (289), Seidman (personal communication,
1998), Maisels and associates (290), and Wood and associates (240). Data for the 95th
percentile are primarily obtained from the data of Bhutani and associates (289), but also from
the studies of Newman and associates (203), and Maisels and associates (290). These data
represent values that might be expected in a western, predominantly breast-fed (60% to 70%)
population. In view of the significant variations in different populations (see Fig. 35-15), as well
as the variations found in laboratory measurement (282,283), the values provided should be
used only as rough guidelines. Nevertheless, this graph can be useful in plotting the course of
neonatal jaundice, because it will demonstrate when the velocity of the TSB increase deviates
significantly from the curves shown. Note that the values must be plotted according to the
infant's age in hours, not days. Infants who have values that exceed the 95th percentile deserve
an evaluation to determine a potential cause for the jaundice, and they require careful
surveillance and followup to prevent the development of extreme hyperbilirubinemia. Infants who
have none of the epidemiologic risk factors for hyperbilirubinemia (see Table 35-13) but who
have TSB values that approach the upper percentiles, also should receive closer scrutiny.
Conversely, those whose values fall well below the 50th percentile probably require minimal
surveillance and followup for jaundice (289).
Pathologic Causes of Jaundice

Indirect-reacting Hyperbilirubinemia
Table 35-15 lists the causes of pathologic indirect-reacting hyperbilirubinemia in the neonate and
Table 35-16 lists the causes of direct-reacting hyperbilirubinemia (cholestatic jaundice). For
further discussion of neonatal cholestasis, please see Chapter 40.
TABLE 35-15 CAUSES OF INDIRECT HYPERBILIRUBINEMIA IN NEWBORNS
Increased production or bilirubin load on the liver

Hemolytic Disease
Immune mediated
Rh alloimmunization, ABO and other blood group incompatibilities
Heritable
Red cell membrane defects
Hereditary spherocytosis, elliptocytosis, pyropoikilocytosis, stomatocytosis
Red cell enzyme deficiencies
a
Glucose-6-phosphate dehydrogenase deficiency,

pyruvate kinase deficiency, and other erythrocyte enzyme deficiencies
Hemoglobinopathies
-Thalassemia, -thalassemia
Unstable hemoglobins
Congenital Heinz body hemolytic anemia
Other causes of increased production
ab
Sepsisa, ,
Disseminated intravascular coagulation
Extravasation of blood, hematomas, and pulmonary, abdominal, cerebral, or other
occult hemorrhage
Polycythemia
Macrosomic infants of diabetic mothers
Breast-milk jaundice
a
Pyloric stenosis
Small- or large bowel obstruction or ileus

Decreased Clearance
Prematurity
Glucose-6-phosphate dehydrogenase deficiency
Inborn errors of metabolism
Crigler-Najjar syndrome, types I and II
Gilbert syndrome
Galactosemia
Tyrosinemia
Hypermethioninemia
Metabolic
Hypothyroidism
Hypopituitarism
Decreased clearance also part of pathogenesis.
Elevation of direct-reading bilirubin also occurs.

P.798
Increased Bilirubin Load: Hemolytic Disease

Immune-Mediated Hemolytic Disease
Rh Erythroblastosis Fetalis.
Pathogenesis. Hemolytic causes of hyperbilirubinemia are discussed fully in Chapter 46. Most
cases of Rh immunization are a result of the Rh (D) antigen although alloimmunization can occur
to other fetal red cell surface antigens including the other antigens of the Rh blood-group system
(c, C, e, E, cc, and Ce) and those belonging to the Kell, Duffy, Kidd, and MNS systems
(294,295,296). Rh alloimmunization occurs when fetal red blood cells from an Rh (D)-positive
fetus cross the placenta into the circulation of an Rh-negative mother and less than 0.1 mL of
fetal red cells is sufficient to result in sensitization. Although Kleihauer-Betke measurements
indicate that fetal-maternal hemorrhage occurs in 75% of women, more sensitive molecular
detection techniques show that fetal red cells enter the maternal circulation in all pregnancies
(296). If the Rh-positive fetus is ABO
P.799
compatible with its mother, the likelihood of Rh immunization is 16%, but it is only 1.5% to 2% if
they are ABO incompatible. This is because the fetal ABO-incompatible red blood cells are rapidly
destroyed in the maternal circulation, diminishing the opportunity of the Rh antigen to induce an
immune response (294). Once a primary immune response to Rh antigen has been mounted,
however, ABO incompatibility between the mother and fetus conveys no protection against a
secondary immune response (294). The risk of alloimmunization following induced abortion is 4%
to 5%, and it is 2% following spontaneous abortion. Other invasive procedures, such as
amniocentesis, chorionic villus sampling, and fetal blood sampling, all increase the risk of fetal
maternal hemorrhage and, therefore, alloimmunization (296).
TABLE 35-16 CAUSES OF NEONATAL CHOLESTASIS
Bile duct abnormalities

Biliary atresia
Choledochal cyst
Caroli disease
Inspissated secretion
Gallstones
Spontaneous perforation of bile ducts
Neonatal sclerosing cholangitis
Infections
Systemic
Septicemia
Urinary tract infection

Hepatitic
TORCH (toxoplasmosis, other infections, robella, cytomegalovirus, and herpes
simplex)
Echovirus, adenovirus, coxsackievirus
Human herpes virus-6, varicella-zoster
HIV, hepatitis B
Inherited and metabolic disorders
1-Antitrypsin deficiency
Alagille syndrome
Galactosemia
Cystic fibrosis
Niemann-Pick type C
Progressive familial intrahepatic cholestasis
Gaucher disease
Wolman disease
Tyrosinemia
Zellweger syndrome
Carbohydrate-deficient glycoprotein syndrome
Dubin-Johnson syndromedeficiency in MRP2 (CMOAT [Canalicular multispecific
organic anion transporter])
Rotor syndrome
Bile acid synthetic disorders
Aagenaes syndrome
Endocrine disorders
Hypopituitarism
Hypothyroidism
Hypoadrenalism
Chromosomal disorder
Trisomy 21, 13, 18
Turner syndrome
Toxic
Parenteral nutrition
Chloral hydrate
Fetal alcohol syndrome
Vascular disorders
Budd-Chiari syndrome
Perinatal asphyxia
Multiple hemangiomata
Congestive heart failure
Miscellaneous
Familial hemophagocytic lymphohistiocytosis
ARC syndrome (arthrogryposis, renal tubular dysfunction, and cholestasis)
From McKiernan P. Neonatal cholestasis. Semin Neonatol 2002;7:53-165, with permission.
The initial response to the foreign antigen in the maternal circulation is for the maternal immune
system to produce immunoglobulin (Ig) M antibodies that do not cross the placenta. This
response is followed by production of IgG antibodies, which then cross the placental barrier. The
secondary immune response to repeat exposure to the Rh antigen produces anti-D IgG
antibodies. This response can be induced with as little as 0.03 ml of D-positive red blood cells
(297). The degree of Rh sensitization is related to the dose of antigen exposure and, therefore, to
the volume of transplacental hemorrhage (294).
Clinical Course.
Approximately 50% of affected infants do not require treatment; they are mildly anemic at birth
and never develop severe hyperbilirubinemia. Approximately 25% to 30% will require intervention
with phototherapy and/or exchange transfusion and approximately 20% to 25% are so severely
affected that they develop hydrops in utero (296). About half of this last group become hydropic
before 34 weeks' gestation and require direct intravascular fetal transfusion (294). A fetal
hematocrit of less than 30% is generally considered an indication for intrauterine transfusion,
which is performed, as required, until 34 to 35 weeks' gestation with delivery planned close to
term (296).
Prevention of Rh Sensitization.
Rh sensitization can almost always be prevented by the administration of Rh immunoglobulin to
Rh-negative women at 28 weeks of gestation and again within 72 hours of delivery of an Rhpositive infant (295). In the United States, the dose is 300 g, but in many other countries it is
100 to 125 g. If the Kleihauer-Betke test or the fetal red cell assay indicates that there is a
transplacental hemorrhage of more than 30 mL of fetal blood (which occurs in 1 in 400
pregnancies) then the dose of Rh (D) immunoglobulin must be at least 10 g/mL of fetal blood in
the maternal circulation (296). Rh immunoglobulin must also be given after abortion or
threatened abortion and after amniocentesis or chorionic villus sampling or any other invasive
intrauterine procedure. These interventions have dramatically reduced the incidence of
erythroblastosis fetalis caused by Rh (D) sensitization, which now has an estimated incidence of
about 1 per 1,000 live births (298).
Some laboratories have replaced the Kleihauer-Betke assay with a fetal red blood cell assay
(Fetalscreen, Ortho-Clinical Diagnostics, Raritan, NJ) (299). In this assay, an antihemoglobin F
antibody is added to the mother's blood to tag hemoglobin F molecules in fetal red cells. Flow
cytometry quantifies the number of fetal red cells (of a total of 50,000 maternal cells) so tagged.
If less than 0.1% of the maternal cells are tagged it is considered a negative result. Positive
results can be quantified to provide the volume of fetal blood in the mother's circulation and the
appropriate dose of Rh (D) immunoglobulin to be given.
In mothers who are already sensitized, the administration of intravenous immunoglobulin (IVIg)
in early pregnancy has had some benefit in cases of severe fetal alloimmunization (300). The
mechanism of action of IVIg appears to involve blockage of Fc receptors on macrophages in the
fetal reticuloendothelial system. In one study, fetal survival was 36% higher when high-dose IVIg
treatment preceded intrauterine transfusion than it was with transfusion alone (300).
Hydrops Fetalis.
The pathogenesis of hydrops fetalis, with its attendant edema and serous effusions, is not clear. It
commonly occurs when the fetal hemoglobin drops below 6 to 7 g/dL. The rapid production of
severe anemia in fetal sheep produced hydrops associated with an increased central venous
pressure and placental edema, whereas the same degree of anemia produced over a longer period
did not result in hydrops, placental edema, or an increased central venous pressure (301). In Rh
isoimmunization, fetal edema may result from the extensive erythropoiesis that takes place in the
fetal liver. This can disrupt the portal circulation and impair albumin synthesis (302,303). Fetuses
with severe hydrops also have elevated concentrations of atrial natriuretic factor (304). Hypoxia
produces myocardial dysfunction with increased umbilical venous pressure that leads to the
release of atrial natriuretic factor (305). Severely affected infants die of progressive
cardiorespiratory failure, in which asphyxia and hyaline membrane disease play a major role.
In one hydropic fetus with erythroblastosis fetalis, pulse-Doppler studies of left and right
ventricular outputs were obtained over time. Despite severe anemia, cardiac outputs were normal
and remained normal after in utero percutaneous intravascular transfusions, which reversed the
hydrops. These measurements of normal cardiac output in utero suggest that high-output failure
caused by anemia is not the mechanism for hydrops in these infants and supports the hypothesis
that portal hypertension and disruption of normal liver function from extramedullary
hematopoiesis is the primary mechanism for the development of hydrops in isoimmune hemolytic
disease of the fetus (306).
Treatment of Infants with Rh Hemolytic Disease and

Hydrops Fetalis.
See sections on Treatment, Phototherapy, and Exchange Transfusion, below.
Late Anemia.
A well-known late complication of intrauterine transfusion is the development of anemia in
P.800
the first months of life. This is commonly seen after about age 2 weeks and is characterized by a
persistently low reticulocyte count. Although originally considered to be a hyporegenerative
anemia (294,307), there is now some evidence that a decrease in erythropoietin production may
not be the primary cause of this late anemia. Recent studies suggest that the anemia is most
likely caused by persistence of anti-D antibodies and destruction of red blood cell precursors in
the marrow or of reticulocytes in the peripheral circulation (308). In a study of 30 neonates with
erythroblastosis, 18 of whom had received intrauterine transfusions and 12 who had not (308),
blood samples were analyzed for hemoglobin, erythropoietin, and reticulocytes. As the
hemoglobin declined between 10 and 80 postnatal days, there was a matching rise in
erythropoietin, which occurred both in infants who had received intrauterine transfusions and
those who had not. Thus, intrauterine transfusions, themselves, may not play such an important
role in suppressing erythropoiesis. Furthermore, reticulocyte counts remained persistently low
despite rising erythropoietin and falling hemoglobin levels. One possible explanation for this is the
effect of circulating anti-D antibodies on the erythrocyte precursors in the bone marrow and on
the reticulocytes in the peripheral circulation (308).
ABO Hemolytic Disease

Approximately 45% of Americans of western European descent have type O blood, and a similar
percentage are type A. Types B and AB make up the balance. African Americans are 50% type O,
29% type A, 17% type B, and 4% AB (309). Hemolytic disease related to ABO incompatibility is
generally limited to group A or B infants born to group O mothers, and it tends to run in families.
One study found an 88% risk of recurrence of ABO hemolytic disease in ABO incompatible
newborns born to parents whose first-born child was similarly affected (310).
In a prospective study of 4,996 consecutive live born infants (311) (Table 35-17), cord blood was
analyzed for blood type, hematocrit, and results of the direct antiglobulin test (DAT or Coombs
test) and the indirect Coombs test (311). The DAT detects antibodies attached to the red cell,
whereas the indirect Coombs test detects IgG antibody in the serum. Only 0.29% of type A, B, or
AB infants who were incompatible with their type A or B mothers had a positive DAT result,
whereas 32% of type A or B infants born to type O mothers had positive DATs. A positive DAT
was the best predictor of an elevated bilirubin level, but only 20% of infants with a positive DAT
developed TSB levels of 12.8 mg/dL (224 mol/L) (311). This large prospective study confirms
what was found in other smaller studies: although about one-third of group A or B infants born to
group O mothers have anti-A or anti-B antibodies attached to their red cells, only 1 in 5 of those
with a positive DAT have a modest degree of hyperbilirubinemia.
TABLE 35-17 BILIRUBIN LEVELS IN ABO-INCOMPATIBLE INFANTS ACCORDING TO THE
COOMBS TEST
Peak Serum Bilirubin 12.8 mg/dL
Coombs Test Results
Direct antiglobulin (Coombs) test positive

Indirect Coombs test positive
Both tests negative
(224 mol/L)
46/225 (20.4%)
29/309 (9.4%)
38/488 (7.8%)
From Ozolek J, Watchko J, Mimouni F. Prevalence and lack of clinical significance of blood
group incompatibility in mothers with blood type A or B. J Pediatr 1994;125:8791, with
permission.
Consequently, although ABO-incompatible DAT-positive infants are about twice as likely as their
compatible peers to have moderate hyperbilirubinemia, severe jaundice in these infants is
uncommon (311,312,313,314,315) and ABO hemolytic disease is a relatively rare cause of severe
hyperbilirubinemia (Table 35-18).
Diagnosing ABO Hemolytic Disease.

There is a wide spectrum of severity in ABO hemolytic disease. Nevertheless, this diagnosis
generally should not be made unless there is a positive DAT and clinical jaundice within the first
12 to 24 hours. Reticulocytosis and the presence of microspherocytes on the smear help to
confirm the diagnosis (Table 35-19).
ABO Incompatibility, Hyperbilirubinemia, and a Negative

Direct Antiglobulin Test.
Although the epidemiologic data suggest otherwise, most clinicians have seen the occasional ABOincompatible infant in whom there is an early rising TSB level, yet the DAT is negative. In the
past, we have often considered these to be cases of ABO hemolytic disease and have attributed
the negative DAT to the technical vicissitudes of Coombs testing in the laboratory. Recent data,
however, raise questions about alternative mechanisms.
One obvious possibility is that there is another cause for the hemolysis. Using measurements of
end-tidal carbon
P.801
monoxide concentration (ETCOc), a direct measurement of heme catabolism, Herschel and
associates (316) identified four DAT-negative ABO-incompatible neonates who had elevated
ETCOc levels. Further investigation revealed that two of these infants had G6PD deficiency and
one had elliptocytosis. Can ABO incompatibility with a negative DAT nevertheless contribute to
hyperbilirubinemia? Kaplan and associates (317) found that 43% of DAT-negative, ABOincompatible infants who were homozygous for the variant UGT promoter associated with Gilbert
syndrome, had a TSB level 15 mg/dL (256 mol/L) versus none of the ABO-incompatible DATnegative infants who were homozygous normal (for the variant promoter) (Fig. 35-17). There was
no difference between ABO-incompatible and ABO-compatible DAT-negative newborns, as long as
the ABO-incompatible neonates did not have Gilbert syndrome (see Inherited Unconjugated
Hyperbilirubinemia). These observations confirm, for the first time, that if another icterogenic
factor is present, then ABO-incompatible newborns are at risk for hyperbilirubinemia even if they
are DAT negative (317).
TABLE 35-18 DISCHARGE DIAGNOSIS IN 306 INFANTS ADMITTED WITH SEVERE
HYPERBILIRUBINEMIA
Diagnosis
Number
Percentage
290
94.8
1.0
11
3.6
Anti-E hemolytic disease
0.3
Galactosemia
Sepsis
1
0
0.3
0
Hyperbilirubinemia of unknown cause or breast-milk jaundice

Cephalhematoma or bruising
ABO hemolytic disease
Infants were readmitted after discharge as newborns. Mean age at admission was 5
days (range, 2 to 17 days), and mean bilirubin level was 18.5 2.8 mg/dL (range, 12.7
to 29.1 mg/dL).
a
Mother was type O, infant was type A or B, direct Coombs test was positive.
From Maisels MJ, Kring E. Risk of sepsis in newborns with severe hyperbilirubinemia.
TABLE 35-19 CRITERIA FOR DIAGNOSING ABO HEMOLYTIC DISEASE AS THE CAUSE OF
NEONATAL HYPERBILIRUBINEMIA
Mother group O, infant group A or B

and
Positive DAT
Jaundice appearing within 1224 hours
Microspherocytes on blood smear
Negative DAT but homozygous for Gilbert's syndrome (317)
Should a blood type and DAT be performed on the cord blood of all infants of group O mothers?
In these days of cost containment, this is a commonly asked question. A recent survey found that
58% of hospital blood banks in the United States were routinely performing Coombs tests and
blood typing on newborn cord bloods (318). Approximately 36% of hospitals tested all cord bloods
routinely, and 35% tested those of type O or Rh-negative mothers, even though the data suggest
that such routine screening is not warranted (311,318). Furthermore, even when such testing is
done, there is evidence that it is often ignored by the responsible pediatrician (318,319). The AAP
notes that routine cord blood screening for infants of group O, Rh-positive mothers, is an option,
but is not required provided there is appropriate surveillance, and risk assessment before
discharge and follow up (76) so that significantly jaundiced infants are not missed.
Heritable Causes of Hemolysis

Red Cell Membrane Defects
The red cell membrane defects that may produce hemolysis and hyperbilirubinemia in the
newborn include hereditary spherocytosis, elliptocytosis, pyropoikilocytosis, pyknocytosis and
stomatocytosis syndromes (320,321). Detailed descriptions of their clinical presentation and
management can be found elsewhere (320,321,322). The diagnosis of these disorders can be
difficult, because newborns commonly exhibit a marked variation in red cell membrane size and
shape (322).
In 75% of hereditary spherocytosis patients, inheritance is autosomal dominant so that a positive
family history that includes anemia, jaundice, gallstones, and splenectomy can often be elicited.
As in G6PD deficiency (see Glucose-6-Phosphate Dehydrogenase Deficiency below), the presence
of jaundice severe enough to require phototherapy in newborns with hereditary spherocytosis is
strongly related to an interaction with the Gilbert syndrome allele (323). Severe anemia and
hydrops fetalis has occurred
P.802
in infants with hereditary spherocytosis associated with defective genes for band 3 or spectrin
protein (321). The diagnosis can be made using the incubated osmotic fragility test which is a
reliable diagnostic tool in newborns when coupled with fetal red cell controls (321,322).
Figure 35-17 Incidence of hyperbilirubinemia defined as TSB 15 mg/dL (256 mol/L) in ABOincompatible and ABO-compatible (control) infants according to the UGT promoter genotype.
ABO-incompatible DAT-negative infants who were also homozygous for the variant UGT promoter
(Gilbert syndrome) had a significantly higher incidence of hyperbilirubinemia than did ABOincompatible DAT-negative infants who were homozygous normal for the UGT promoter. The
former subgroup also had a significantly greater incidence of hyperbilirubinemia than any of the
three UGT promoter genotype subgroups in the control (ABO-compatible) infants. (From Kaplan
M, Hammerman C, Renbaum P, et al. Gilbert's syndrome and hyperbilirubinaemia in ABOincompatible neonates. Lancet 2000;356: 652-653, with permission.)
Red Cell Enzyme Deficiencies

Glucose-6-Phosphate Dehydrogenase Deficiency
Epidemiology.
G6PD deficiency is the most common, clinically significant, red cell enzyme defect and affects as
many as 4,500,000 newborns each year (324). Although known for its prevalence in the
populations of the Mediterranean, the Middle East, the Arabian Peninsula, South East Asia, and
Africa, immigration and intermarriage have transformed G6PD deficiency into a global problem
(156). Nevertheless, most pediatricians in the United States do not think of G6PD deficiency when
confronted with a jaundiced infant (325). They should, particularly in black infants. Although
African American newborns, as a group, have lower TSB levels than white newborns (197,198,
205,206), G6PD deficiency is found in 11% to 13% of African American newborns. This means
that some 32,000 to 39,000 black male G6PD-deficient hemizygote newborns will be born
annually in the United States (326), and kernicterus has occurred in some of these infants
(124,327,328 and 329). In a recent report, G6PD deficiency was considered to be the cause of
hyperbilirubinemia in 19 of 61 (31.5%) infants who developed kernicterus (124).
Clinical Course.
Because G6PD deficiency in African American infants is a result of the less-severe Gd A- mutation
(326), most of these infants do not develop severe hyperbilirubinemia (326). If subjected to some
oxidative stress, however, they can develop acute hemolysis with a sudden increase in the TSB
(324). Although previous studies suggested otherwise (326), recent data show that, as a group,
black G6PD-deficient infants are significantly more likely than controls to develop
hyperbilirubinemia and three times more likely than control infants to require phototherapy (330).
Hemolysis and hyperbilirubinemia in G6PD-deficient neonates can be triggered by exposure to a
number of agents (156,324,327). These include naphthalene (found in moth balls), agents for
umbilical cord antisepsis, breast milk of a mother who has eaten fava beans (331), and perhaps
exposure to a variety of household chemicals (324). Neonatal infection is also a well recognized
trigger (156, 332). In most cases, however, no specific triggering agent or condition can be
identified and an acute hemolytic event is the exception, rather than the rule (156,333). Most
G6PD-deficient neonates have a more gradual onset of hyperbilirubinemia and there is evidence
that this hyperbilirubinemia has its origins in utero (334).
Pathogenesis of Hyperbilirubinemia in G6PD Deficiency.
The G6PD gene (Gd) is located on the X chromosome and hemizygous Gd- males have the full
enzyme deficiency and can be identified by screening tests (324). But female heterozygotes have
a wide range of enzyme activity and will frequently be missed by screening tests (335) even
though they are also at risk for hyperbilirubinemia (335,336).
In most cases of hyperbilirubinemia in G6PD-deficient neonates, there is no overt evidence of
hemolysis such as anemia and reticulocytosis (337,338), although in some populations the usual
indices of hemolysis are found (339). Conversely, blood COHb and ETCOc concentrations are
consistently elevated in G6PD-deficient infants (339,340,341). Nevertheless, Kaplan and
associates found no difference in COHb levels between hyperbilirubinemic (TSB greater than
15mg/dL [256 mol/L]) and nonhyperbilirubinemic G6PD-deficient infants (341). These
investigators also found that although COHb values were higher in G6PD-deficient neonates than
in normal neonates, there was no correlation between TSB levels and COHb values in the G6PDdeficient group (342). All of these observations suggest that while an increase in heme turnover is
clearly present in G6PD-deficient neonates, with the exception of those who suffer an acute
hemolytic event, hemolysis alone cannot be implicated as the primary mechanism responsible for
hyperbilirubinemia.
HPLC measurements of conjugated bilirubin fractions provide an index of the hepatic conjugating
capacity. Lower serum conjugated bilirubin fractions relative to serum total bilirubin
concentrations indicate diminished conjugating capacity (343). In G6PD-deficient neonates who
developed TSB levels greater than 15 mg/dL (256 mol/L), serum total, mono- and di-conjugated
bilirubin fractions were significantly lower than in nonhyperbilirubinemic G6PD-deficient infants
(344), suggesting that impaired conjugation plays a role in the pathogenesis of hyperbilirubinemia.
Finally, a remarkable interaction between G6PD deficiency and Gilbert syndrome was
demonstrated by Kaplan and associates (345). In this study of Israeli infants, neither the
presence of the variant UGT promoter (for Gilbert syndrome) by itself, nor G6PD deficiency alone
had a significant effect on the incidence of hyperbilirubinemia (TSB greater than 15 mg/dL [256
mol/L]), but there was a significant increase in hyperbilirubinemia in G6PD-deficient infants who
also had the variant UGT promoter (Fig. 35-18). The incidence of hyperbilirubinemia in G6PDdeficient neonates increased from 9.7% in normal homozygotes to 31.6% in the variant UGT
promoter heterozygotes to 50% in homozygotes for the variant promoter. No significant effect of
the variant UGT promoter was seen in the G6PD-normal infants (Fig. 35-18). Thus neither G6PD
deficiency alone, nor the abnormal UGT promoter alone (Gilbert syndrome), caused an increased
incidence of hyperbilirubinemia; both factors were needed to produce a significant increase in TSB
levels. It is interesting, however, that in Italian G6PD-deficient neonates, homozygosity for the
variant 7/7 promoter did not increase the risk of hyperbilirubinemia (346). Routine screening for
G6PD deficiency is currently not performed in the United States, although such screening would
be appropriate in the black
P.803
population (326). Table 35-20 summarizes the relationships and mechanisms involved in G6PD
deficiency and hyperbilirubinemia.
Figure 35-18 Incidence of hyperbilirubinemia, defined as TSB 15.0 mg/dL (256 mol/L), for
G6PD-deficient and control neonates, stratified for the three promoter genotypes of the gene
encoding the bilirubin conjugating enzyme UGT1A1. (From Kaplan M, Renbaum P, Levi-Lahad E,
et al. Gilbert syndrome and glucose-6-phosphate dehydrogenase deficiency: A dose-dependent
genetic interaction crucial to neonatal hyperbilirubinemia. Proc Natl Acad Sci U S A
1997;94:12128-12132, with permission.)
Treatment
The risk of kernicterus in G6PD-deficient infants with TSB levels above 20 mg/dL (342) mol/L)
appears to be comparable to that associated with Rh disease. Thus, in the presence of G6PD
deficiency, more aggressive treatment of these infants is needed (339,347,348).
Pyruvate Kinase Deficiency

This autosomal recessive disorder is less prevalent than G6PD deficiency and, in contrast to G6PD
deficiency, typically presents with jaundice, anemia, and reticulocytosis (322). Neonatal
hyperbilirubinemia is common and kernicterus has been reported (321). This disorder should be
considered in a hyperbilirubinemic neonate with a picture of nonspherocytic, DAT-negative
hemolytic anemia (322).
Other Disorders of the Embden-Meyerhof Pathway

Conditions such as hexokinase deficiency, glucose phosphate isomerase deficiency, and
phosphofructokinase deficiency occasionally are associated with hemolysis and neonatal
hyperbilirubinemia (321). Jaundice and anemia have also been described in isolated cases of 2,3bisphosphoglycerate mutase deficiency and phosphoglycerate kinase deficiency (321).
TABLE 35-20 RESULTS OF STUDIES ON THE PATHOGENESIS OF HYPERBILIRUBINEMIA
IN G6PD-DEFICIENT INFANTS
Increased red cell turnover (hemolysis) (339341) although infants with TSB >15 mg/dL
(256 mol/L) did not have higher COHb levels than those with TSB <15 mg/dL (341)
Decreased conjugating capacity in infants with TSB >15 mg/dL (344)
No increase in TSB >15 mg/dL (vs. G6PD-normal infants) unless deficient infants are also
heterozygotes or homozygotes for variant UGT promoter for Gilbert syndrome (345)
(found in Israeli but not Italian neonates) (346)
Hemoglobinopathies
These conditions generally do not present in the newborn period. Fetal hemoglobin (hemoglobin
[Hb] F) is composed of alpha () and gamma ()2 chains. Homozygous thalassemia (complete
absence of chain synthesis) results in profound hemolysis, anemia, hydrops fetalis, and almost
always stillbirth or death in the immediate neonatal period (322). As there are no chains in
hemoglobin F, -thalassemia does not manifest itself in neonates. Sickle cell disease is
asymptomatic in the neonate (349) because of the inhibitory effect of Hb F on Hb S
polymerization and cellular sickling (350). The expression of sickle cell disease is thus masked
until Hb S levels increase to more than 75% at about 6 months (350).
Extravascular Blood
Cephalhematomas, bruising, intracranial or pulmonary hemorrhage, or any occult bleeding may
lead to an elevated TSB level from breakdown of the extravascular erythrocytes (206,351,352
and 353). (The catabolism of 1 g of hemoglobin yields 35 mg of bilirubin.) In two reports, severe
hyperbilirubinemia followed delayed absorption of intraperitoneal blood in infants who had
received intraperitoneal fetal transfusions before birth (351,353). In both reported cases, despite
multiple exchange transfusions, hyperbilirubinemia was not controlled until peritoneal lavage was
performed. Massive adrenal hemorrhage has also caused severe hyperbilirubinemia (352).
In the VLBW infant, the presence of PIVH is associated with an increase in serum bilirubin levels
in some studies (354,355) but not in others (356). Amato and colleagues (356) studied 88 infants
with birth weights less than 1,500 g. Phototherapy was initiated only when serum bilirubin levels
exceeded 12 mg/dL (205 mol/L). The incidence of serum bilirubin levels greater than 12 mg/dL
was
P.804
39% in the PIVH group and 46.8% in the infants without PIVH. There was no difference in the
duration of phototherapy in the two groups.
Polycythemia
The catabolism of 1 g of hemoglobin produces 35 mg of bilirubin and it is often assumed that a
high hematocrit is a risk factor for neonatal jaundice, because an increase in the erythrocyte mass
should increase the bilirubin load presented to the liver. Nevertheless, mean bilirubin levels and
the incidence of hyperbilirubinemia were similar in polycythemic infants randomly assigned to
receive either partial exchange transfusions or symptomatic treatment (see Epidemiology of
Neonatal Jaundice) (231,232 and 233). In one study, however, infants were held 30 cm below the
introitus after delivery. If cord clamping was delayed, the mean TSB level at 72 hours was 7.7 mg/
dL (132 mol/L) compared with 3.2 mg/dL (55 mol/L) in the early clamped group (234).
Increased Enterohepatic Circulation

(See Physiologic Mechanisms of Neonatal Jaundice, Epidemiology of Neonatal Hyperbilirubinemia,
and Breast-feeding and Jaundice for the contribution of the enterohepatic circulation to neonatal
jaundice.) Intestinal obstruction or a delay in bowel transit time increases the enterohepatic
circulation by decreasing caloric intake and allowing more time for bilirubin deconjugation and
reabsorption. Jaundice is common in infants with small-bowel obstruction and occurs in infants
with pyloric stenosis (357,358,359). Correction of the obstruction produces a prompt decline in
bilirubin levels.
The pathogenesis of jaundice associated with pyloric stenosis has been debated for many years
(358) and it was suggested that caloric deprivation, as well as a decrease in UGT activity, might
play an important role (358). In Gilbert syndrome, caloric deprivation produces an increase in TSB
levels and a low activity of UGT has been seen in jaundiced infants with pyloric stenosis
(358,360). In a report of three formula-fed infants with hypertrophic pyloric stenosis and jaundice
(359), two were homozygous for the UGT variant promoter for Gilbert syndrome and one infant
was heterozygous for the promoter. These observations confirm a critical role for Gilbert
syndrome in the pathogenesis of jaundice associated with pyloric stenosis (see Gilbert Syndrome
below).
TABLE 35-21 INBORN ERRORS OF HEPATIC BILIRUBIN UGT EXPRESSION
Characteristic
Crigler-Najjar Type II
Crigler-Najjar Type 1
(Arias syndrome)
Gilbert Syndrome
Inheritance
Autosomal recessive
Autosomal recessive or Autosomal dominant

dominant
or recessive
UGT1 activity
Genetics
Absent
Nonsense or stop
<10%
Missense mutation
Hyperbilirubinemia
mutation
>20 mg/dL
Kernicterus
High risk
515 mg/dL
a
Low risk
50%
Variant promoter
35 mg/dL
No apparent risk
Marked hyperbilirubinemia can occur in some cases of Arias syndrome, which may place
the infant at high risk for kernicterus.
From Watchko JF. Indirect hyperbilirubinemia in the neonate. In: Maisels MJ, Watchko JF,
eds. Neonatal jaundice. London, UK: Harwood Academic Publishers, 2000:5166, with
permission.
Infants of Diabetic Mothers

Only macrosomic infants of mothers with insulin-dependent diabetes are at increased risk of
hyperbilirubinemia (213). This is probably the result of increased bilirubin production
(214,215,216) (see Epidemiology of Neonatal Hyperbilirubinemia: Maternal Factors).
Decreased Bilirubin Clearance
Inherited Unconjugated HyperbilirubinemiaInborn

Errors of Bilirubin Uridine Diphosphoglucuronate
Glucuronosyltransferase Activity
The structure and function of the UGT1A1 gene and the genomics of hyperbilirubinemia have been
reviewed in detail (22,23). (See Fig. 35-7 and Structure and Function of the Uridine
Diphosphoglucuronate Glucuronosyltransferase 1A1 Gene above) Because a single form of
bilirubin UGT accounts for almost all of the bilirubin glucuronidation activity in the human liver,
inherited defects of a single enzyme will cause jaundice. Three types of inherited UGT deficiency
are recognized.
Crigler Najjar Syndromes.

Crigler-Najjar syndromes types I and II (CN-1, CN-2) are caused by one or more mutations in any
of the five exons of the UGT1A1 gene (22,23), as well as mutations in the noncoding, intronic
region of the gene (see Fig. 35-7 and Table 35-21) (22). More than 30 different genetic mutations
have been identified in CN-1 syndrome and the gene frequency for CN-1 is estimated to be
1:1000 (23). Infants with CN-1 have almost
P.805
complete absence of bilirubin UGT activity. They develop severe hyperbilirubinemia in the first 2
to 3 days of life and often require exchange transfusion in the first week. Subsequently, intensive
home phototherapy controls bilirubin levels to some extent, but as these children get older,
increasing skin thickness and pigmentation and a decrease in the surface-area-to-body-mass ratio
render phototherapy less effective. A tanning bed phototherapy configuration is necessary to
obtain adequate irradiance and surface area exposure (see Phototherapy below) (361). Brain
damage can occur at any time, including adulthood, and plasmapheresis has been used to reduce
bilirubin concentrations during acute exacerbations of hyperbilirubinemia (362,363). One 16-yearold boy with CN-1 had 72 plasma exchanges over a period of 28 months before undergoing
orthotopic liver transplantation (364). Liver transplantation is currently the only available
definitive therapy, and serum bilirubin concentrations decline dramatically within hours of the
procedure (365).
Another option is the use of human hepatocyte transplantation (366,367). Because hepatic
architecture and function, except for bilirubin UGT activity, are normal in CN-1, transplantation of
isolated liver cells is an attractive option. The infusion of hepatocytes (obtained from a donor
liver) into the portal vein of a 10-year-old girl with CN-1 decreased TSB levels by 50% (366).
Hepatic bilirubin UGT activity increased from 0.4% to 5.5% of mean normal enzyme activity, and
more than 30% of the patient's bile pigments were now bilirubin glucuronides. The ultimate
treatment of the Crigler-Najjar syndrome, however, lies in the development of effective gene
therapy (368), although this has yet to be attempted in a human.
The administration of tin-protoporphyrin and tin-mesoporphyrin to children with CN-1 syndrome
(369, 370) reduced TSB levels and the need for phototherapy. Administration of oral calcium
phosphate significantly reduced TSB levels in patients with CN-1 who were receiving phototherapy
(371).
The diagnosis of Crigler-Najjar syndrome is made by HPLC analysis of serum and duodenal bile,
liver biopsy tissue enzyme assay, evaluating the response to phenobarbital (372,373), and by
molecular analysis of the UGT1A1 gene (374). In CN-1 disease, phenobarbital has little or no
effect on TSB concentrations, whereas in children with CN-2 disease, TSB levels can decrease by
30% to 80% during phenobarbital treatment (375). Phenobarbital acts via a phenobarbital-
responsive enhancer module that stimulates the UGT1A1 gene to induce production of
glucuronosyltransferase (376). There is now a world registry for the CN-1 syndrome that is a
unique source of information about this rare disease (377).
Suresh and Lucey (173) conducted a questionnaire survey of 42 patients ranging in age from 2
months to 21 years who had CN-1. Home phototherapy for 10 to 16 hours, principally at night,
was the mainstay of postneonatal therapy. Additional therapies included oral agar, antioxidants,
bilirubin oxidase, clofibrate, and cholestyramine, and liver transplantation was performed in 15 of
the children. All patients grew normally; in 77% the neurodevelopmental status was normal.
Those in school were doing well despite having had TSB levels of 15 to 29 mg/dL (257 to 496
mol/L) for many years. Although it often is stated that sensorineural hearing loss is the most
common form of bilirubin toxicity to the central nervous system, not one of 36 children evaluated
had a sensorineural hearing loss, suggesting that bilirubin may not be as ototoxic as commonly
believed. In these children it is important to avoid exacerbations of hyperbilirubinemia and to
manage intercurrent infections promptly. Albumin infusions and plasmapheresis are effective in
dealing with acute exacerbations of jaundice (173,372).
Infants with CN-2 disease (also known as Arias syndrome) generally have less-severe
hyperbilirubinemia, although it can occur, and kernicterus has been reported in some infants.
There is considerable overlap between CN-1 and CN-2 syndromes. Both infants and adults with
CN-2 syndrome respond readily to phenobarbital therapy, with a sharp decline in serum bilirubin
levels within 7 to 10 days. This response can be used to differentiate between the two syndromes
(372).
Gilbert Syndrome.
People with Gilbert syndrome have a mild, benign, chronic or recurrent unconjugated
hyperbilirubinemia with no evidence of liver disease or overt hemolysis. There is now evidence,
however, that these individuals also have an increase in heme turnover (378) (see below). Gilbert
syndrome is common, affecting approximately 6% to 9% of the general population, and both
autosomal dominant and recessive patterns of inheritance have been suggested. Typically, the
indirect-reacting hyperbilirubinemia is not recognized until after puberty and manifests itself
during fasting or intercurrent illness.
The genetic basis for this disorder has been clarified (22,23,379,380) and involves mutations of
the UGT1A1 gene promoter (see Fig. 35-7). In whites with Gilbert syndrome there is commonly a
variant promoter for the gene encoding UGT1A1. (This is not the case in East Asian populations.)
This promoter contains a two base-pair addition (TA) in the TATAA element that gives rise to
seven (TA)7 TAA(7/7) rather than the more usual six (TA)6 TAA(6/6) repeats in affected subjects
(379). There is an inverse relationship between the number of repeats and the activity of the
promoter: as the number of TA repeats increases, UGT activity decreases (379,381). Subjects
with Gilbert syndrome are homozygous for the variant promoter, providing a unique genetic
marker for this disorder. Hetero-zygotes have one allele each of the wild-type and variant
promoters (6/7) (379). The gene frequency for the 7/7 motif is 0.3 so that 9% of the general
population are homozygous and 42% are heterozygous (382). Thus about half of the white
population carries a Gilbert promoter on at least one allele.
Although most commonly diagnosed in young adulthood, it is now clear that Gilbert syndrome
plays a role in the pathogenesis of neonatal jaundice (22,23). Several investigators have shown
that neonates who are homozygous for the
P.806
variant 7/7 UGT gene promoter have a more rapid rise in their TSB levels (383) and higher TSB
levels at age 96 hours (Fig. 35-19) (384). Of neonates with TSB concentrations greater than 13
mg/dL (222 mol/L) 26.8% were homozygous for the variant 7/7 promoter versus 12.2% of
those whose TSB levels were 13 mg/dL (385). In a population of Scottish, primarily breast-fed
newborns with TSB levels of greater than 5.8 mg/dL (100 mol/L) after 14 days of life, 31% were
homozygous for the 7/7 Gilbert syndrome promoter genotype compared with only 6% of a control
group with acute jaundice (259). Of 17 breast-fed Japanese infants with prolonged jaundice, 16
had at least 1 mutation of the UGT1A1 gene, primarily of the G7IR type (260).
Figure 35-19 Newborn mean serum bilirubin levels as a function of postnatal age and presence
of Gilbert promoter abnormality. Homozygous normal UGT genotype (6/6); heterozygous variant
UGT genotype (6/7); homozygous variant UGT genotype (7/7). (Redrawn from Watchko JF.
Indirect hyperbilirubinemia in the neonate. In: Maisels MJ, Watchko JF, eds. Neonatal jaundice.
London, UK: Harwood Academic Publishers, 2000:51-66, with permission, from data of RoyChowdhury N, Deocharan B, Bejjanki HR, et al. Presence of the genetic marker for Gilbert
syndrome is associated with increased level and duration of neonatal jaundice. Acta Paediatr
2002;91:100-101.)
Thus, Gilbert syndrome plays an ubiquitous role in the pathogenesis of neonatal

hyperbilirubinemia. The combination of the Gilbert genotype with other icterogenic factors such as
breast-feeding (259,260), G6PD deficiency (345), ABO incompatibility (317), and pyloric stenosis
(359) dramatically increases a newborn's risk of hyperbilirubinemia.
Mutations of the Uridine Diphosphoglucuronate

Glucuronosyl-transferase
1A1 Gene Coding Area Associated with Gilbert Syndrome. Unlike white populations with Gilbert
syndrome, the TATAA promoter variations are rare in East Asian populations (22). In these
populations, Gilbert syndrome appears to result from missense mutations in the coding area of
the UGT1A1 gene. The most common of these is a G A transition at nucleotide 211, which
causes arginine to replace glycine at position 71 of the corresponding protein product (386,387).
This G71R variant is prevalent in Japanese, Korean, and Chinese populations (387) and other
mutations have also been reported in association with Gilbert syndrome in these populations
(388,389). The presence of these mutations could account for the higher TSB levels found in
Japanese and other East Asian newborns.
Heterozygosity for Uridine Diphosphoglucuronate

Glucuronosyl-transferase 1A1 Noncoding and Coding Area
Mutations.
Because approximately 50% of the population carry a Gilbert type promoter on at least one allele,
it is not surprising that some individuals may be heterozygous for the variant promoter gene and
also heterozygous for a coding region mutation of the gene (22,23). Coinheritance of the Gilbert
promoter and a coding region mutation of the gene can also lead to jaundice (23). This is true not
only for patients who are homozygous for the Gilbert genotype, but also for compound
heterozygotes of a Gilbert type promoter and a structural region mutation of the UGT1A1 (382). A
striking example of this situation was recently described in twin girls who developed marked
hyperbilirubinemia and kernicterus and were found to be compound heterozygotes for the Gilberttype UGT promoter and a coding region mutation of the UGT1A1 gene (382).
Based on the gene frequencies for the Gilbert promoter and structural mutations of the UGT1A1
gene, at least 1 in 3,300 infants will be compound heterozygotes for Gilbert and UGT1A1 coding
region mutations and at risk for significant hyperbilirubinemia (23,380). Watchko notes that the 1
in 3,300 likelihood of this compound heterozygosity is similar to the frequency of TSB levels
greater than 30 mg/dL (greater than 513 mol/L) (203,288).
We do not know if Gilbert syndrome plays an important role in the pathogenesis of extreme
hyperbilirubinemia (TSB levels greater than 30 mg/dL [513 mol/L]), but the fact that poor
feeding and weight loss (a state resembling fasting) (124,390), as well as very low direct-reacting
bilirubin fractions, are seen in some of these infants suggests that this is a possibility worthy of
investigation.
Other Inborn Errors of Metabolism

Galactosemia.
Galactosemia is a rare disease (incidence, 1 in 35,000 to 60,000 live births), and jaundice may be
one of the presenting features (391); but infants with significant hyperbilirubinemia caused by
galactosemia almost all have some other manifestations of the disease, most often poor feeding,
vomiting, excessive weight loss, irritability, and lethargy. Hyperbilirubinemia during the first week
of life is almost exclusively unconjugated, and
P.807
the conjugated fraction tends to rise during the second week, probably reflecting liver damage.
The presence of a positive family history, lethargy, poor feeding, or other signs of illness merit
additional diagnostic evaluation, including testing the urine for reducing substances using Clinitest
(Miles Inc., Diagnostic Division, Elkhart, IN). Escherichia coli sepsis is the most devastating
complication in the newborn (391).
Tyrosinemia and Hypermethioninemia.

The relation between these inborn errors of metabolism and jaundice is primarily a result of the
presence of neonatal liver disease, which initially may manifest as indirect-reacting
hyperbilirubinemia but which generally is accompanied by some evidence of cholestasis (i.e.,
direct-reacting hyperbilirubinemia).
Hypothyroidism.
Prolonged indirect-reacting hyperbilirubinemia is one of the clinical features of congenital
hypothyroidism (392,393,394 and 395), a condition that must be ruled out in any infant who has
indirect-reacting hyperbilirubinemia beyond 2 to 3 weeks of age. Although widespread availability
of screening programs for congenital hypothyroidism should allow early identification of this
problem as a possible cause of jaundice, screening programs do not detect every infant, and
errors are more likely to occur with early discharge of infants in whom the thyroxine (T4) level
may still be spuriously elevated.
The pathogenesis of hyperbilirubinemia associated with hypothyroidism is not clear, and
administration of triiodothyronine to full-term and preterm infants does not lower peak serum
bilirubin levels (396,397). In one infant with prolonged jaundice, UGT activity in a liver biopsy
sample was unmeasurable (393), but the jaundice resolved following thyroxine administration.
Conversely, when rats underwent thyroidectomy UGT activity increased and they also developed
cholestasis (398).
Drugs
The use of pancuronium and chloral hydrate is associated with higher bilirubin levels in sick
preterm infants (275,276 and 277), and chloral hydrate is associated with an increased risk of
direct-reacting hyperbilirubinuria (276).
Breast-Milk Jaundice
See Epidemiology of Neonatal Hyperbilirubinemia, Breast-Feeding and Jaundice, above.
Prolonged Indirect-reacting Hyperbilirubinemia

Prolonged indirect-reacting hyperbilirubinemia is defined as indirect-reacting bilirubinemia
persisting beyond 2 weeks of age in the full-term infant; Table 35-22 lists its causes. The
association of prolonged indirect-reacting hyperbilirubinemia with pyloric stenosis is well described
but the pathogenesis of this association has never been clear, although delayed gastric emptying
and enterohepatic circulation were thought to play a role. In a report of three infants with pyloric
stenosis and hyperbilirubinemia, two were homozygous, and one heterozygous for the variant
promoter for Gilbert syndrome (359), implying a key role for Gilbert syndrome in the
pathogenesis of the prolonged hyperbilirubinemia that sometimes occurs in pyloric stenosis.
TABLE 35-22 CAUSES OF PROLONGED INDIRECT HYPERBILIRUBINEMIA
Breast-milk jaundice
Hemolytic disease
Hypothyroidism
Pyloric stenosis
Crigler-Najjar syndrome
Gilbert syndrome
Extravascular blood
Mixed Forms of Jaundice

Sepsis and Urinary Tract Infection
Jaundice is one sign of bacterial sepsis and some reports suggest that unexplained
hyperbilirubinemia may be the only manifestation of sepsis in otherwise healthy-appearing
newborns (399,400,401). Should newborns with unexplained hyperbilirubinemia be subjected to
lumbar puncture and blood and urine cultures even if they appear otherwise well? I evaluated 306
newborns admitted within 21 days of birth with indirect-reacting hyperbilirubinemia (peak TSB
level 18.5 2.8 mg/dL; range: 12.7 to 29.1 [316 48 mol/L; range: 217 to 498]) (see Table 3518). No case of sepsis was diagnosed (249).
In a study of 160 asymptomatic jaundiced infants presenting to an emergency department in Los
Angeles, urine cultures (catheter specimens) were positive (greater than 10,000 colony-forming
units [cfu]/mL) in 12 (7.5%) of the infants (402) (0 of 44 circumcised boys, 9 of 94
uncircumcised boys, and 3 of 62 girls). A positive urine culture was more likely when jaundice was
first noted after 8 days of age and if there was an elevated direct-reacting bilirubin concentration
(402). The authors recommended that all asymptomatic, jaundiced infants presenting to an
emergency department (402) should have a urine culture, but this view has been questioned
(403).
Given the prevalence of jaundice in the newborn, the presence of indirect-reacting
hyperbilirubinemia as the only manifestation of bacteremia or incipient sepsis, must be a very
rare occurrence. Furthermore, the finding of a positive blood or urine culture in a newborn with
indirect-reacting hyperbilirubinemia does not prove that the infection is the cause of the jaundice.
However, infants who appear sick or have late-onset jaundice after the initial icterus has resolved,
or infants with direct-reacting hyperbilirubinemia or something else in the history, physical
examination, or laboratory investigations that is out of the ordinary, should be evaluated carefully
for possible sepsis or urinary tract infection.
Hypopituitarism
Prolonged jaundice that is predominantly cholestatic (elevated direct-reacting bilirubin) has been
described in
P.808
infants with congenital hypopituitarism (404,405), although in some the hyperbilirubinemia is
indirect. The pathogenesis of hyperbilirubinemia in this condition remains to be elucidated.
Other Causes
Congenital syphilis, the TORCH (toxoplasmosis, other agents, rubella, cytomegalovirus, and
herpes simplex) group of chronic intrauterine infections, and coxsackievirus B infection are the
other important causes of mixed jaundice. The clinical features and diagnoses of these conditions
are described in Chapter 48.
Newer Developments that Affect Our Approach to the

Jaundiced Newborn
In the last several years, three factors have emerged that have colored our approach to the evaluation and
management of neonatal jaundice. The first is a series of case reports indicating that kernicterus, thought to be
almost extinct, is still occurring (124,133,328,329,390,406). The second is the decreasing hospital stay for
newborn infants, and the third is an increase in the incidence of neonatal jaundice.
Kernicterus Still Occurs

Neonatologists still see occurrences of kernicterus and reports in the literature (124,133,328,329,390,406)
suggest a resurgence of this devastating condition. Neverthe-less, there are no appropriate data to support
this perception, because there is no uniform surveillance for the reporting of kernicterus over the last 3 to 4
decades, no agreed on case definition for kernicterus, and, most important, no denominators (i.e., population
base) for the case reports listed. Nevertheless, the fact that kernicterus, although rare, still occurs, demands our
attention because it is nearly always preventable by relatively simple interventions (76).
A preliminary analysis of the reported cases of kernicterus suggests the possible root causes (124,407) listed in
Table 35-23.
Early Discharge and the Risk of Jaundice

In addition to the global trend toward a shorter hospital stay for newborns, several studies (239,241,288) but
not all (206,408), have found that early discharge itself is associated with an increased risk of significant
hyperbilirubinemia and even kernicterus (124,328). Recognizing both the risk of unrecognized jaundice in
infants discharged early, as well as the fact that, in infants discharged before 72 hours, the TSB level is almost
always still rising, the AAP guidelines for followup of infants discharged before 72 hours are stringent (76) (see
section on Preventing Extreme Hyperbilirubinemia and Kernicterus).
It is not clear why infants who are discharged early are at greater risk of developing significant
hyperbilirubinemia. The vast majority of these hyperbilirubinemic infants are breast-fed, and it is possible that
mothers who have longer postpartum stays are more rested, have more time to receive advice and counsel
regarding breast-feeding, and are therefore able to nurse their babies more effectively. More frequent and
effective lactation, as well as improved caloric intake, decreases the likelihood of hyperbilirubinemia
(251,252,409). Early discharge also may have a negative effect on the ability of mothers to assimilate and
process the information that they receive regarding lactation and infant care (410). As a result, they may nurse
the infants less effectively.
TABLE 35-23 ROOT CAUSES OF REPORTED CASES OF KERNICTERUS
Early discharge with failure to ensure appropriate followup.

Failure to evaluate the measured TSB level based on the infant's age in hours.
Failure to recognize risk factors for severe hyperbilirubinemia.
Underestimating the severity of jaundice by clinical (i.e., visual) assessment.
Lack of concern regarding the neurotoxic potential of bilirubin.
Shorter hospital stays also have necessitated a readjustment in our thinking with regard to the meaning of
specific bilirubin levels. To date, this has proved to be a difficult adjustment for pediatricians who are
accustomed to using a specific TSB level (irrespective of the infant's age) as an indication for reassurance or
concern. The data of Bhutani and associates (289) graphically illustrate this point (Fig. 35-20). Clinicians
commonly refer to jaundice occurring on day 2 or day 3, but reference to the data of Bhutani and associates
(see Predischarge Measurement of the Bilirubin Level below) and Fig. 35-20 indicates just how misleading this
thought process can be. A TSB level of 8 mg/dL at 24.1 hours is above the 95th percentile and calls for
evaluation and close followup, whereas the same level at 47.9 hours is in the low-risk zone and probably
requires no further concernyet both of these values occur on day 2 (289). One point is worth emphasizing: If
newborns are discharged at less than age 36 hours, their bilirubin levels (with very rare exceptions) can only be
going in one directionup. The recognition that jaundice is now primarily an outpatient problem requires us to
develop a consistent approach to the monitoring and surveillance of these infants if we are to prevent the
development of extreme hyperbilirubinemia and kernicterus (76).
More Jaundiced Newborns

There is evidence that the incidence of neonatal jaundice is increasing. In the days of the major CPP (1959 to
1966), only 5% of infants had bilirubin levels 13 mg/dL (222 mol/L) and less than 1% had a peak TSB level
of20 mg/dL(342 mol/L) (205). However, in three more recent studies, the 95th percentile for a population of
term and
P.809
near-term infants ranged from 15.5 to 18 mg/dL (289,290, 411). In a study of 11 Kaiser Permanente northern
California hospitals, 2% of infants had peak TSB levels20 mg/dL (203). Factors that may be responsible for this
increase in jaundice include an increase in the number of infants breast-fed on discharge from the hospital (30%
in the 1960s, 65% to 70% in 2001) (412,413) and shorter hospital stays (239,241,288).
Figure 35-20 Nomogram for designation of risk in 2,840 well newborns 36 week's gestational age with birth
weight 2,000 g or 35 weeks' gestational age and a birth weight 2,500 g or more based on the hour-specific
serum bilirubin values. The serum bilirubin level was obtained before discharge, and the zone in which the
value fell predicted the likelihood of a subsequent bilirubin level exceeding the 95th percentile (high-risk zone)
as shown in Table 35-27. Note that because of sampling bias (293) this nomogram should not be used to
represent the natural history of neonatal hyperbilirubinemia. (From Bhutani VK, Johnson L, Sivieri EM.
Predictive ability of a predischarge hour-specific serum bilirubin for subsequent significant hyperbilirubinemia in
healthy-term and near-term newborns. Pediatrics 1999;103:6-14, with permission.)
Prevention, Identification, and Management of

Neonatal Hyperbilirubinemia
The Approach to a Jaundiced Infant
The Institute of Medicine (IOM) issued a report in 2001 that describes a chasm in health care
quality which needs to be crossed for patients to receive better care (414). The report calls for a
systems approach drawing on the rapid evolution of knowledge about complex adaptive
systems. The IOM report offers 10 simple rules to redesign and improve care (414) Palmer
and associates (415) studied the care of newborns with hyperbilirubinemia in two large managed
care organizations and describe how this care might be improved by using practical applications of
the ideas about complex adaptive systems.
The recently published guideline of the AAP provides an approach to the evaluation and treatment
of hyperbilirubinemia in the term and near-term newborn 35 weeks of gestation (76). Although
bilirubin encephalopathy or kernicterus should almost always be preventable, it is still occurring in
this group of infants. Table 35-24 lists the key elements of the recommendations provided in the
AAP guideline. The discussion that follows deals primarily with infants 35 weeks of gestation,
although the basic principles apply to the evaluation and management of hyperbilirubinemia in all
infants.
Primary Prevention
Prenatal Testing
Screening for Isoimmunization
All pregnant women should be tested for ABO and Rh (D) typing and undergo a serum screen for
unusual isoimmune antibodies (76). If such prenatal testing has not been performed, then a
direct Coombs test, a blood type, and an Rh (D) type on the infant's cord blood should be done;
this should always be done if the mother is Rh-negative. In addition to identification of potentially
Rh-sensitized infants, this testing is obligatory because it identifies Rh-negative mothers who
require anti-D gammaglobulin to prevent Rh (D) sensitization.
In infants of group O, Rh-positive mothers, the AAP recommends that routine testing for blood
type and Coombs test is optional provided there is appropriate surveillance
P.810
and risk assessment before discharge and followup (76) so that significantly jaundiced infants are
not missed. (See ABO Hemolytic Disease above.)
TABLE 35-24 KEY ELEMENTS OF THE AMERICAN ACADEMY OF PEDIATRICS GUIDELINE
ON MANAGEMENT OF HYPERBILIRUBINEMIA IN THE NEWBORN INFANT 35 WEEKS OF
GESTATION
1. Promote and support successful breast-feeding.

2. Establish nursery protocols for the identification and evaluation of hyperbilirubinemia.
3. Measure the TSB or TcB level on infants jaundiced in the first 24 hours.
4. Recognize that visual estimation of the degree of jaundice can lead to errors, particularly in
darkly pigmented infants.
5. Interpret all bilirubin levels according to the infant's age in hours.
6. Recognize that infants <38 weeks' gestation, particularly those who are breast-fed, are at
higher risk of developing hyperbilirubinemia and require closer surveillance and
monitoring.
7. Perform a systematic assessment on all infants prior to discharge for the risk of severe
hyperbilirubinemia.
8. Provide appropriate followup based on the time of discharge and the risk assessment.
9. Treat newborns, when indicated, with phototherapy or exchange transfusion.
TcB, transcutaneous bilirubin.

From Maisels MJ, Baltz RD, Bhutani V, et al. Management of hyperbilirubinemia in the
newborn infant 35 or more weeks of gestation. Pediatrics 2004;114:297316, with
permission.
Preventing Hyperbilirubinemia
To prevent severe hyperbilirubinemia, we need to (a) identify infants who are at risk for
developing hyperbilirubinemia, (b) follow them closely, and (c) treat them with phototherapy
when indicated.
Ensuring Successful Breast-feeding.

Because exclusively breast-feeding an infant is strongly associated with an increased risk of
hyperbilirubinemia (see Breast-Feeding and Jaundice above), the only primary preventive
intervention available to us is to ensure the adequacy and success of breast-feeding. In many, but
not all, cases of severe hyperbilirubinemia in breast-fed infants, poor caloric intake associated
with inadequate breast-feeding (and manifested by increased weight loss) appears to play an
important role (218,246,254,390). Exclusively breast-feeding an infant is also a well documented
risk factor for dehydration (416), considered by some to be a cause of hyperbilirubinemia,
although there is no plausible mechanism for this. It is much more likely that in breast-fed,
dehydrated infants, it is the caloric deprivation and its effect on the enterohepatic circulation of
bilirubin (see Table 35-14) that is primarily responsible for the severe hyperbilirubinemia (244).
Thus, the clinician's primary role in preventing hyperbilirubinemia is to help ensure that breastfeeding will be successful. The first step is to ask mothers to nurse their infants at least 8 to 12
times per day for the first several days (417), because increasing the frequency of nursing
significantly decreases the likelihood of subsequent hyperbilirubinemia (251,252,409). Evidence of
adequate intake in breast-fed infants also includes 4 to 6 thoroughly wet diapers within 24 hours
and the passage of 3 to 4 stools per day by the fourth day. By the third to fourth day, the stools
in adequately breast-fed infants should have changed from meconium to a mustard yellow, mushy
stool (418). Weight loss must also be monitored. Unsupplemented breast-fed infants experience
their maximum weight loss by day 3 and, on average, lose 6.1 2.5% (SD) of their birth weight
(243,246,251,261,419,420,421 and 422). Thus, approximately 5% to 10% of exclusively breastfed infants lose 10% or more of their birth weight by day 3, suggesting that adequacy of intake
should be evaluated and the infant monitored if weight loss is greater than 10% (423). These
assessments will also help to identify breast-fed infants who, because of inadequate intake, are at
risk for dehydration.

In many hospitals it is a common practice to provide supplemental feedings of water or dextrose
water to breast-fed infants in the mistaken belief that this will lower their TSB levels. As
illustrated in Fig. 35-21, this practice does not decrease TSB levels (419,420) but it does interfere
with the establishment of effective lactation (424).
Identifying the Jaundiced Newborn

Clinical Assessment
All infants should be monitored routinely for the development of jaundice and nurseries should
have established protocols for the assessment of jaundice. In newborns, jaundice is detected by
blanching the skin with digital pressure, thus revealing the underlying color of the skin and
subcutaneous tissue. It is important for this assessment to be done in a well-lit room or,
preferably, in daylight at a window.
The Cephalocaudal Progression of Jaundice

Dermal icterus is seen first in the face and then progresses in a caudal manner to the trunk and
extremities so that, for a given bilirubin level, the skin of the face will appear more yellow than
that of the foot. First observed more than 100 years ago and confirmed by several investigators
using visual observation as well as transcutaneous bilirubinometry, the cephalocaudal progression
of dermal icterus is a useful clinical tool but less reliable once the TSB level exceeds 12 mg/dL
(205 mol/L) (425,426,427,428,429,430,431 and 432).
Figure 35-21 Effect of supplementary feeding on serum bilirubin levels on day 6 in breast-fed
newborns. Nursing infants were randomly assigned to receive supplemental water (n=15),
dextrose water (n=17), or no supplement (n=17). Bilirubin levels on day 6 (mean standard
error of mean [SEM]) are shown. There were no significant differences between any of the
groups. Drawn from the data in Nicoll A, Ginsburg R, Tripp JH. Supplemen-tary feeding and
jaundice in newborns. Acta Pediatr Scand 1982;71:759-761.
P.811
Mechanism of the Cephalocaudal Progression

Knudsen suggests that the cephalocaudal color difference in newborns is best explained by
conformational change in the bilirubin-albumin complex (431). Following its formation, bilirubin is
bound tightly to albumin, and the initial binding process is extremely rapid (within 10
milliseconds). This is followed by a train of slow, relaxing changes in the conformation of the
bilirubin-albumin complex commencing within 1 to 30 seconds, final conformation being reached 8
minutes after the initial binding (433). This time course suggests that initially there is a lower
bilirubin-binding affinity to albumin (until the final stage of conformation has occurred) and thus
less-effective bilirubin-albumin binding in the blood immediately after it has left the
reticuloendothelial system. The affinity increases after the blood reaches the distal portions of the
body and the conformational changes in the bilirubin- albumin complex are completed. Knudsen
suggests that some of the yellow color of the skin is the result of precipitated bilirubin acid and in
the presence of reduced bilirubin-binding affinity to albumin, there is an increased precipitation of
bilirubin acid and thus an increase in the yellow color of the skin (431). This is more likely to
occur in the proximal parts of the body because of the conformational changes in the young
bilirubin-albumin complexes. Consistent with this hypothesis, Knudsen (431) found that there was
a significant and linear correlation among the cephalocaudal color difference, the plasma bilirubin
concentration, and the square of the hydrogen ion concentration. The cephalocaudal color
difference was inversely related to the reserve albumin concentration, thus suggesting that the
conformational change in the young bilirubin-albumin complex enhances the precipitation of
bilirubin acid in the skin of the proximal parts of the body. If this is the mechanism, then objective
measurement of the yellow color of the skin using transcutaneous bilirubinometry could be a
better predictor of potential bilirubin encephalopathy than a serum bilirubin measurement.
How Good Is the Visual Assessment of Jaundice?

The ability of clinicians to diagnose jaundice varies widely (235,425,426,434,435), although
studies suggest that home health nurses (436) and some neonatologists (435) can effectively
evaluate newborns for the presence and severity of jaundice, and that newborns whose TSB levels
exceed 12 mg/dL (205 mol/L) will almost always be identified as jaundiced (235,425).
Nevertheless, visual estimation of actual bilirubin levels from the degree of jaundice can be
unreliable and can certainly lead to errors (235,426,434). Errors are also more likely to occur in
darkly pigmented infants. Consequently, a low threshold should be used for measuring the TSB.
There is also a problem with relying on visual diagnosis in the first 24 to 36 hours. The difference
between a TSB level of 5 mg/dL (85 mol/L) and 8 mg/dL (137 mol/L) cannot be perceived by
eye, yet this is the difference between a TSB level at the 50th percentile and one that is at the
95th percentile (see Figs. 35-16 and 35-20). The potential errors associated with visual diagnosis
have led some experts to recommend that all newborns should have a laboratory measurement of
TSB prior to discharge (124) (see Predischarge Measurement of the Bilirubin Level). An
alternative is the use of the newer noninvasive (transcutaneous) devices that can provide a valid
estimate of TSB levels less than 15 mg/dL (257 mol/L) (140,204,437,438).
Noninvasive Measurements of Bilirubin

The Ingram Icterometer (Cascade Health Care Products, Salem, OR) is a piece of transparent
plastic on which is painted 5 transverse strips of graded yellow hue. The instrument is pressed
against the nose, the yellow color of the blanched skin is matched with the appropriate yellow
stripe, and a jaundice score is assessed. This device is simple and inexpensive and, remarkably,
as a screening tool, it performs as effectively as much more sophisticated instruments (439). It
has also been used effectively by nurses and even parents in the home (425,440).
P.812
The Minolta Air Shields jaundice meter (Air-Shields, Hatboro, PA) was the first electronic device
marketed for transcutaneous bilirubin measurement. With this instrument and its successors
(models JM-102 and -103), pressure is applied to a photoprobe and a strobe light is generated by
a xenon tube. The light passes through a fiberoptic element, penetrating the blanched skin and
entering the subcutaneous tissue. The reflected light returns through a second fiberoptic bundle to
the spectrophotometric module where the intensity of the yellow color, corrected for hemoglobin,
is measured and displayed. In the original jaundice meter this number was displayed as an
arbitrary unitthe transcutaneous bilirubin (TcB) index. The newer model, JM-103, displays a
bilirubin value that corresponds closely to the measured TSB level. The JM-103 jaundice meter
uses two wavelengths and a dual optical path system (204,438). The principle of operation
includes the formation of two beams, one of which reaches only the shallow areas of the
subcutaneous tissue, while the other penetrates the deeper layers. The differences between
optical densities are detected by blue and green photocells. The measurement of bilirubin
accumulated primarily in the deeper subcutaneous tissues should decrease the influence of other
skin pigments such as melanin and hemoglobin (438).
Another transcutaneous device, the BiliChek (Respironics, Marietta, GA) employs multiple
wavelengths and measures TcB by using the entire spectrum of visible light (380 to 760 nm)
reflected by the skin. The absorption of hemoglobin, melanin, and the effect of dermal thickness
are isolated mathematically, and absorption of light by bilirubin in the capillary bed and
subcutaneous tissues is isolated by spectral subtraction.
To date, studies with the JM-103 and the BiliChek have shown a close correlation between TcB
and TSB measurements in mixed racial populations (204,437,441,442). When compared with
HPLC measurements of TSB, the BiliChek was as good or better than the TSB measured in the
clinical laboratory (441) although the accuracy of the BiliChek in a Hispanic (Mexican) population
has been questioned (202). In this population, the BiliChek tended to under estimate the TSB,
particularly when the TSB was greater than 10 mg/dL (171 mol/L) (202). The JM-103 is less
precise in black infants than in white infants, although the predominant tendency in black infants
is for the TcB value to overestimate TSB levels (204) so that dangerous errors are unlikely to
occur.
TcB measurements become less precise with decreasing gestation and are of questionable utility
below 30 weeks' gestation (443). Because phototherapy bleaches the skin, both visual
assessments of jaundice and TcB measurements in infants undergoing phototherapy are not
reliable (444). It might be possible, however, to use existing TcB instruments in infants who are
undergoing phototherapy if an area of skin used for TcB measurements is protected from light
(444). Data are limited regarding the use of TcB measurements following phototherapy. In one
study, TcB measurements 18 to 24 hours after discontinuation of phototherapy correlated well
with the TSB and the correlation improved further after an additional 24 hours (444).
Because TcB measurements are so easy to perform, repeated measurements can be obtained
over the course of an infant's stay in the nursery. Plotting serial measurements will provide a
good indication of whether TcB levels are rising and crossing percentiles and must therefore be
followed more closely (see Fig. 35-20) Although the TcB measurement provides a good estimate
of the TSB level, as with any laboratory measurement, TcB levels should not be considered in
isolation, nor should critical decisions be based on a single measurement.
Both the BiliChek and JM-103 have an acceptable level of diagnostic accuracy, particularly when
used as a screening device to place infants in a risk category for followup, or as an indication for
obtaining a TSB (445). In some populations, TcB measurements may not be able to identify the
precise bilirubin level with an acceptable degree of accuracy (the wrong job for the tool) (445).
Currently, it is probably inadvisable to rely entirely on TcB measurements as a substitute for TSB
measurements but they can certainly be helpful in many ways. When used as a screening tool TcB
measurements can help us to answer the questions Should I worry about this infant? and
Should I obtain a TSB on this infant? (445). For both of these purposes, the physician can set a
value for a TcB measurement (based on the infant's age in hours and other risk factors) above
which a TSB level will always be obtained. In our nursery, the nurses currently measure the TcB
on all infants and automatically obtain a TSB if the TcB is above the 75th percentile (289). In our
study with the JM-103, the chance of a TcB measurement underestimating the TSB level by 3 mg/
dL (51.3 mol/L) or more was only 0.6%, so an alternative approach is to ask, If the real TSB
value is TcB plus 3 mg/dL (51 mol/L), is there a reasonable chance that this will change my
management? If the answer is yes, a TSB should be measured. This allows the clinician to take
into account other risk factors, including the gestation. In this manner, no infant with significant
hyperbilirubinemia should be missed and many infants and their families will be spared the
trauma, cost, and inconvenience of having a laboratory measurement of serum bilirubin. The
ability to measure the TcB in the office or other outpatient settings, including the home,
noninvasively and instantaneously, should prove of inestimable value in the monitoring and
management of the jaundiced newborn infant and should largely avoid the potential errors
associated with clinical estimation of bilirubin levels.
Laboratory Measurements of Bilirubin

Total Serum Bilirubin
This subject has been reviewed in detail (284,446,447) and addressed, most
recently by Lo and associates (282). Although this is one of the most commonly
performed
P.813
laboratory measurements in the newborn, the measurement of TSB concentration
remains remarkably inaccurate. Repeated surveys over the last 3 decades have
disclosed a high level of interlaboratory variation in the measurements of both
total and direct-reacting serum bilirubin concentrations in neonatal sera. In a
recent study, a sample with a known TSB concentration of 14.8 mg/dL (243 mol/
L) was analyzed in 14 different university hospital laboratories. The mean ( SD)
measured TSB was 15.2 2.5 mg/dL (coefficient of variation 16.4%) and the
range was 12.1 to 18.5 mg/dL (207 to 316 mol/L) (283). This remarkably wide
range of values in skilled hands suggests that considerable caution is necessary
before generalizing bilirubin levels from one institution to the wider universe of
newborns. Within-laboratory variation generally is considered to be lower, but in
the study of Vreman and associates (283), over time, the coefficient of variation
for repeated measurements of the same sample was as high as 17.2% in one
laboratory (283).
The most recent analysis of laboratory neonatal TSB measurements was
conducted by Lo and associates (282) who examined data submitted by
laboratories that participate in the College of American Pathologists (CAP)
neonatal bilirubin and chemistry surveys. They compared the data from
laboratories throughout the United States with the reference method for
measuring TSB (448). They found an acceptable coefficient of variation ranging
from 1.9% to 4.5%, in the four most commonly used laboratory instruments.
When compared with the reference method, the mean values from these
instruments differed from the reference method by 21.6% to 10.9%. The
authors attributed these discrepancies to the use of different methods used
(primarily with the Vitros instrument, Ortho-Clinical Diagnostics, Raritan, NJ) in
the neonatal bilirubin versus the chemistry surveys of the CAP, as well as the
presence of a nonhuman protein base and ditaurobilirubin in the survey specimens
(115). Lo and associates concluded that it is impossible to evaluate the accuracy
of TSB measurements from the CAP surveys because the standard specimens
consist of bovine serum containing a mixture of unconjugated bilirubin and
ditaurobilirubin. They recommend that survey specimens should consist of human
serum enriched with unconjugated bilirubin. The CAP has adopted this
recommendation and is now using human-serum-based specimens in their
neonatal bilirubin and chemistry surveys (BT Doumas, personal communication,
October, 2003).
These variations between laboratories might explain the frequent occurrence in
clinical practice of an infant being admitted for treatment of hyperbilirubinemia
because an outside laboratory found a high TSB level, but when the test is
repeated in the hospital laboratory, the TSB level is 5 or 6 mg/dL (85 to 103 mol/
L) lower. Of course, there is no way of knowing which value is correct. A 16.4%
coefficient of variation between laboratories means that if the true serum bilirubin
value is 20 mg/dL (342 mol/L), the 95% confidence limits of a repeat
measurement at another laboratory could fall anywhere between 14.4 and 26.6
mg/dL (246 to 455 mol/L). Because our followup, surveillance, and intervention
in jaundiced infants are based on TSB values, spurious underestimation of the TSB
concentration might lead to withholding of necessary therapy, and overestimation
will produce unnecessary clinical intervention.
Direct-Reacting and Conjugated Bilirubin

Similar concerns exist with regard to measurements of direct-reacting or
conjugated bilirubin where laboratory measurements are considered, at best,
only an approximation to the true value (449). In one study, measurements of
direct-reacting bilirubin using the Dupont ACA (automatic clinical analyzer) or
Coulter Dacos analyzer produced direct-reacting bilirubin levels in a population of
full-term newborns at one hospital that were twice as high as those measured in a
similar population at another institution, where a Kodak Ektachem 700 (now
known as the Vitros) method was used (450). The Vitros method measures
conjugated bilirubin, whereas the other methods measure direct-reacting bilirubin.
Although often used synonymously, direct-reacting bilirubin is not the same as
conjugated bilirubin. Direct-reacting bilirubin refers to the bilirubin that reacts
directly with diazotized sulfanilic acid (i.e., without the addition of an accelerating
agent), whereas conjugated bilirubin refers to bilirubin that has been made water
soluble by binding with glucuronic acid in the liver.
Site of Blood SamplingCapillary Versus Venous

Blood Samples
Data regarding the differences in TSB levels when measured in capillary or venous
samples are conflicting (451,452). It is useful to recall, however, that virtually all
of the published data regarding the relationship of TSB levels to kernicterus or
developmental outcome are based on capillary TSB levels. Thus, for the purposes
of clinical decision making, capillary blood samples are the gold standard and
there is no reason to delay initiation of treatment to obtain a venous blood sample
to confirm an elevated capillary TSB level.
Sampling Technique
Because of the well-known effects of light on bilirubin, laboratory manuals
recommend that blood samples be protected from light until the serum is
analyzed. Using serum samples with bilirubin levels of 16.0, 11.8, and 7.9 mg/dL
(273, 202, and 135 mol/L), Sykes and colleagues (453) found that, under the
usual clinical conditions, there was no measurable effect of ambient light on
serum bilirubin levels for at least 8 hours.
When Is Laboratory Measurement of Bilirubin

Indicated?
A TcB and/or TSB level should be done on every infant who is jaundiced in the
first 24 hours after birth (407). In a
P.814
cohort of 105,384 newborns of 2,000 g birth weight and at least 36 weeks'
gestational age, jaundice was noted in the medical record in only 2.8% of infants
within 18 hours and in 6.7% within 24 hours (454). Compared with those who
were not jaundiced on the first day, those jaundiced within 24 hours were much
more likely to receive phototherapy (18.9% vs. 1.7%) and were three times more
likely to develop a bilirubin level of 25 mg/dL (428 mol/L) or higher (454). Thus,
the presence of jaundice in the first 24 hours represents a significant risk factor
for the subsequent development of severe hyperbilirubinemia and, as
recommended by the AAP (76), a TSB should be measured in any infant who is
jaundiced in the first 24 hours.
TABLE 35-25 LABORATORY EVALUATION OF THE JAUNDICED INFANT
Indications
Jaundice in first 24 hours

Jaundice appears excessive for
infant's age
Assessments
Measure TSB
Measure TSB
Infant receiving phototherapy or

TSB rising rapidly (i.e., crossing
percentiles [see Fig. 35-20]) and
unexplained by history and
physical examination
Blood type and Coomb test, if not

obtained with cord blood
Complete blood count and smear
Measure direct-reacting or conjugated
bilirubin
It is an option to do reticulocyte count,
G6PD, ETCOc, if available
Repeat TSB in 424 hours, depending
TSB approaching exchange levels
on infant's age and TSB level

Do reticulocyte count, G6PD, albumin,
or not responding to phototherapy and ETCOc, if available

Elevated direct (or conjugated)
bilirubin level
Do urinalysis and urine culture;

evaluate for sepsis if indicated by
history and physical examination
Jaundice present at or beyond age Total and direct-reacting (or

3 weeks, or sick infant
conjugated) bilirubin level; if directreacting bilirubin elevated, evaluate for
causes of cholestasis
Check results of newborn thyroid and
galactosemia screen; evaluate infant
for signs or symptoms of
hypothyroidism
ETCOc, end-tidal carbon monoxide corrected for ambient carbon
monoxide; G6PD, glucose-6-phosphate dehydrogenase; TSB, total serum
bilirubin concentration.
From Maisels MJ, Baltz RD, Bhutani V, et al. Management of
hyperbilirubinemia in the newborn infant 35 or more weeks of gestation.
The need for and timing of a repeat TSB or TcB depends on the zone in which the
TSB falls (see Fig. 35-20), the age of the infant, other clinical risk factors, and the
subsequent evolution of the bilirubin level (Table 35-25). A TcB and/or TSB level
should also be done if, at any time, the jaundice appears excessive for the infant's
age. Recognizing the risk of error in the visual estimation of bilirubin levels
(235,425,426,434), particularly in darkly pigmented infants, and depending on
the clinical circumstances, a TSB or TcB should be measured if there is any doubt
about the degree of jaundice (76).
Seeking a Cause for the Jaundice

Table 35-25 lists the indications for laboratory evaluation of jaundiced newborns.
In some infants, the cause of hyperbilirubinemia is apparent from the history and
physical examination. For example, jaundice in a severely bruised infant generally
needs no further explanation. In addition, the usual laboratory tests (hematocrit,
complete blood count, reticulocyte count, and smear) are neither specific nor
sensitive and rarely identify a specific cause for the hyperbilirubinemia (198,254)
even in infants who are readmitted to hospital with TSB levels of 18 to 20 mg/dL
(308 to 340 mol/L) or higher (see Table 35-18) (241, 249). One study, however,
did find a good correlation between measurements of ETCOc and reticulocyte
counts (corrected for the hematocrit) in infants with positive DATs (455). The
cause of jaundice should be sought in an infant receiving phototherapy or whose
TSB level is rising rapidly, that is, crossing percentiles (see Fig. 35-20), and is not
explained by the history and physical examination. Table 35-25 lists the
appropriate laboratory tests.
Interpreting the Bilirubin Level

All TSB levels must be interpreted in terms of the infant's age in hours (see Figs.
35-16 and 35-20). An infant with a TSB that exceeds the 95th percentile, or in
whom the rate of rise of the TSB crosses percentiles (see Fig. 35-20), or exceeds
0.2 mg/dL per hour deserves further evaluation and followup.
Elevated Direct-Reacting or Conjugated Bilirubin

Levels
If the TSB is mg/dL (85 mol/L) or lower a direct-reacting or conjugated
bilirubin of greater than 1.0 mg/dL (17.1 mol/L) is considered elevated. For TSB
values greater than 5 mg/dL (85 mol/L) a direct-reacting or conjugated bilirubin
of more than 20% of the TSB is considered abnormal. If the laboratory measures
conjugated bilirubin using the Vitros system, any value greater than 1 mg/dL
P.815
(17.1 mol/L) is considered abnormal. An infant who has an elevation of directreacting or conjugated bilirubin should have a urinalysis and a urine culture
performed because this could be an early sign of urinary tract infection (see Table
35-25). A sepsis evaluation should also be considered. Late onset of jaundice
(after the fourth or fifth day) has also been associated with urinary tract infection
(402) (see also Pathologic Causes of Jaundice: Sepsis above).
Sick Infants and Infants Jaundiced Beyond Age 3

Weeks
These infants should have a measurement of total and direct-reacting or
conjugated bilirubin to identify cholestasis (see Table 35-25). If the direct-reacting

or conjugated bilirubin level is elevated then additional evaluation is needed for
the causes of cholestasis (see Table 35-16 and Chapter 40). This approach is
essential for the early identification of infants with biliary atresia. If these infants
are to benefit from the operation of portoenterostomy, surgery should be
performed as soon as possible and, ideally, before age 60 days (456). Infants who
are jaundiced beyond age 3 weeks should also have the results of the newborn
thyroid and galactosemia screen checked. Table 35-22 lists the causes of
prolonged indirect-reacting hyperbilirubinemia.
Preventing Extreme Hyperbilirubinemia

and Kernicterus
Most of the cases of kernicterus reported in the last decade did not occur in
infants who had ABO or Rh hemolytic disease, but in apparently healthy term and
near-term newborns with extremely high bilirubin levels (usually well above 30
mg/dL [513 mol/L]) (124,390,406), in infants with G6PD deficiency
(124,133,328,329), and in very sick newborns with low bilirubin levels (43). If we
can prevent extreme hyperbilirubinemia we should be able to prevent almost all
cases of kernicterus, but jaundice is very common and extreme hyperbilirubinemia
(30 mg/dL [513 mol/L] or higher) is rare, occurring in only about 1 in 10,000
infants (203). To ensure that we do not miss these rare infants, however, we need
to follow and measure serum bilirubin levels in many infants, and treat some
infants with phototherapy who will never develop severe hyperbilirubinemia
(139,140). Ip and associates have calculated that 5 to 10 infants with TSB levels
between 15 and 20 mg/dL (257 to 342 mol/L) receive phototherapy to prevent 1
infant from reaching a TSB of 20 mg/dL (342 mol/L) (139,140).
Hospital Policies and Jaundiced Newborns

According to recommendation 2.2 of the AAP Clinical Practice Guideline (76),
clinicians should ensure that all infants are routinely monitored for the
development of jaundice, and nurseries should have established protocols for the
assessment of jaundice. Jaundice should be assessed whenever the infant's vital
signs are measured, but no less than every 8 to 12 hours. Recommendation
2.2.1 states that protocols for the assessment of jaundice should include the
circumstances in which nursing staff can obtain a TcB or order a TSB
measurement. Nursing staff have always assumed the responsibility for
monitoring infants for jaundice but, for the first time, hospitals are now asked to
develop protocols under which nurses can obtain a TcB or TSB without a physician
order. Thus a nurse who notices jaundice in an infant who is younger than 24
hours old, will obtain a TcB or TSB and notify the physician of the result. If a TcB
measurement is used as a screening tool, TSB measurements can be obtained by
nurses whenever the TcB measurement is above a certain bilirubin percentile (see
Fig. 35-20 and Noninvasive Measurements of Bilirubin above).
Assessing the Risk of Severe Hyperbilirubinemia

Recommendation 5.1 of the AAP Clinical Practice Guideline (76) states: Before
discharge, every newborn should be assessed for the risk of developing severe
hyperbilirubinemia, and all nurseries should establish protocols for assessing this
risk. Such assessment is particularly important in infants who are discharged
before age 72 hours. The guideline continues, The AAP recommends 2 clinical
options used individually or in combination for the systematic assessment of risk:
predischarge measurement of the bilirubin level using TSB or TcB and/or
assessment of clinical risk factors (76). This must be followed by appropriate
evaluation, monitoring, surveillance, and followup to ensure that severe
hyperbilirubinemia is identified early and treated appropriately.
Clinical Risk Factors

Table 35-26 lists the risk factors for the development of hyperbilirubinemia.
Almost all of these factors can be identified readily without recourse to the
laboratory but, because these risk factors are common and the risk of severe
hyperbilirubinemia is small, individually, these factors are of limited use as
predictors of severe hyperbilirubinemia. Nevertheless, if no risk factors are
present, the risk of severe hyperbilirubinemia is extremely low, and the more risk
factors that are present, the greater the risk of severe hyperbilirubinemia (206).
Some factors, such as breast-feeding, gestation age below 38 weeks, and
significant jaundice (need for phototherapy) in a previous sibling, seem to be
particularly important. It is remarkable that almost every recently described case
of kernicterus occurred in a breast-fed infant, even when the infant had
underlying G6PD deficiency (124,133,328,390).
Decreasing Gestation: The Near-term Newborn

Decreasing gestation has been repeatedly identified as a very important
contributor to hyperbilirubinemia (197,206,211, 239,240,241). We found that
infants of 36 weeks or less gestation
P.816
are 13 times more likely than infants at 40 weeks' gestation to be readmitted for
severe jaundice (241). Infants at 38 weeks of gestation are considered term
babies yet they are 4 times more likely to develop a TSB. 25 mg/dL (428 mol/L)
than those at 40 weeks (206). Although these near-term infants, at 35 to 38
weeks of gestation, are cared for in well-baby nurseries, they are much more
likely to nurse ineffectively, receive fewer calories, and have a greater weight loss
than their truly term counterparts. When combined with less effective hepatic
clearance because of prematurity, it is not surprising that they often become more
jaundiced.
TABLE 35-26 RISK FACTORS FOR DEVELOPMENT OF SEVERE

HYPERBILIRUBINEMIA IN INFANTS 35 WEEKS' GESTATION (IN
APPROXIMATE ORDER OF IMPORTANCE)
Major risk factors

Predischarge TSB or TcB level in the high risk zone (see Fig. 35-20)
(289,437)
Jaundice observed in the first 24 hours (454)
Blood group incompatibility with positive direct antiglobulin test, other
known hemolytic disease (e.g., G6PD deficiency), elevated ETCOc
Gestational age 35 to 36 weeks (206,241)
Previous sibling received phototherapy (211,241)
Cephalhematoma or significant bruising (206)
Exclusive breast-feeding, particularly if nursing is not going well and
weight loss is excessive (206,241)
East Asian race
Minor risk factors

Predischarge TSB or TcB in the high-intermediate risk zone (see Fig.
35.20) (289,437)
Gestational age 37 to 38 weeks (206,241)
Jaundice observed before discharge (241)
Previous sibling with jaundice (211,241)
Macrosomic infant of a diabetic mother (216,582)
Maternal age 25 years (206)
Male gender (206,241)
Decreased risk (these factors are associated with decreased risk of
significant jaundice; listed in
order of decreasing importance)
TSB or TcB in the low risk zone (see Fig. 35-20) (289,437)
Gestational age 41 weeks (206)
Exclusive bottle feeding (206,241)
a
Black race (206)

a
Race as defined by mother's description.

From Maisels MJ, Baltz RD, Bhutani V, et al. Management of
hyperbilirubinemia in the newborn infant 35 or more weeks of gestation.
Predischarge Measurement of the Bilirubin Level
We know that infants who are clinically jaundiced in the first few days are much
more likely to later develop significant hyperbilirubinemia (239,241). Bhutani and
associates (289) measured TSB concentrations in 13,003 infants prior to their
discharge from the hospital. In 2,840 infants additional TSB levels were measured
at least once in the 5 to 6 days following discharge. Infants with ABO
incompatibility and positive Coombs tests were excluded, as were Rh-sensitized
infants. The investigators plotted the TSB levels against the infant's age in hours
and created a nomogram with percentiles that defined a high risk (above 95th
percentile), a low risk (values less than 40th percentile), and an intermediate risk
(40th to 95th percentile) zone (289) (see Fig. 35-20) Using this nomogram, these
investigators found that of infants whose TSB levels fell in the high-risk zone,
39.5% subsequently had values above the 95th percentile. If the TSB level fell in
the low-risk zone no baby subsequently developed a TSB above the 95th
percentile. Table 35-27 shows these results. Note, however, that Fig. 35-20 does
not describe the natural history of bilirubinemia in the newborn. Because only
2,840 (21.9%) of the 13,003 infants had subsequent TSB levels measured, there
is a sampling bias (toward more jaundiced babies) particularly after 48 to 72
hours, so that the lower zones are spuriously elevated (293).
These data do show that obtaining a TSB level prior to discharge is a very useful
way of predicting the risk (or absence of risk) of subsequent significant
hyperbilirubinemia. The predictive value of a predischarge TSB has been
confirmed in several other studies (285,457,458,459), and it defines a group of
infants who, at least as far as hyperbilirubinemia is concerned, may not require an
early followup. This information is very useful in situations where early followup is
difficult or impossible. Predischarge TSB levels also will alert pediatricians to those
infants who, because their TSB levels fall in the high-risk zone, require much more
careful surveillance and followup until there is clinical or laboratory evidence
supporting a declining bilirubin level.
TABLE 35-27 RISK ZONE AS A PREDICTOR OF HYPERBILIRUBINEMIA
TSB Before Discharge
High-risk zone (>95th

percentile)
High-intermediate-risk
zone
Low-intermediate-risk
zone
Low-risk zone
Newborns
(Total = 2,840),
n (%)
Newborns Who
Subsequently
Developed a TSB Level
>95th Percentile, n (%)
172 (6.0)
68 (39.5)
356 (12.5)
46 (12.9)
556 (19.6)
12 (2.26)
1756 (61.8)
From Bhutani VK, Johnson L, Sivieri EM. Predictive ability of a predischarge

hour-specific serum bilirubin for subsequent significant hyperbilirubinemia
in healthy-term and near-term newborns. Pediatrics 1999;103:6-14, with
permission.
P.817
Measuring Bilirubin Production

When heme is catabolized, CO is produced in equimolar quantities with bilirubin,
and measurement of blood COHb concentration, CO production, or CO excretion
provides a measurement of bilirubin production (27,460). The development of a
simple noninvasive method for measuring ETCOc, provides a technique for
quantifying hemolysis (460) and can identify those infants with high or low rates
of bilirubin production. Although the measurement of ETCOc can identify the
presence of hemolysis and is useful as a diagnostic test (316,330,461), it has not
proved useful as a routine method of screening infants to predict the likelihood of
subsequent hyperbilirubinemia (285).
Followup
Identifying the risk of subsequent severe hyperbilirubinemia before the newborn
leaves the nursery is of little value unless appropriate follow-up is provided.
Recommendation 6.1.1 in the AAP Guideline (76) states that: All infants should
be examined by a qualified health care professional in the first few days following
discharge to assess infant well-being and the presence or absence of jaundice.
The timing and location of this assessment will be determined by the length of
stay in the nursery, presence or absence of risk factors for hyperbilirubinemia (see
Table 35-26 and Fig. 35-20) and the risk of other neonatal problems. This
examination can take place in an office, clinic or home and can be done by a
physician, physician assistant, or nurse. Although written and oral information
should be provided to all parents about newborn jaundice, it is unfair and
unreasonable to rely only on parents to identify a significantly jaundiced infant.
Table 35-28 provides the AAP recommendations for the timing of followup. If, for
geographic, climatic, socioeconomic, or other reasons, appropriate followup
cannot be ensured and the predischarge assessment suggests the presence of a
significant risk for severe hyperbilirubinemia, it may be necessary to delay
discharge either until appropriate followup can be planned, or the period of
greatest risk has passed (72 to 96 hours) (76).
At the time of the followup visit, the infant should be assessed for weight gain,
adequacy of intake, the pattern of voiding and stooling, and the presence or
absence of jaundice. Unless a transcutaneous instrument is available or, at the

time of discharge, a followup TSB has already been scheduled, clinical judgment is
used to determine the need for a bilirubin measurement. Because of the potential
for error in the visual estimation of bilirubin levels, however, a TSB or TcB should
be measured if there is any doubt about the degree of jaundice. This is
particularly important in darkly pigmented infants where the visual estimation of
bilirubin levels can lead to significant error.
Information for Parents

An essential component of insuring the safety of a newborn is to make sure that
parents are informed (among other things) about newborn jaundice. The AAP (76)
recommends
P.818
that all hospitals should provide written and oral information for parents at the
time of discharge. This should include an explanation of jaundice, the need to
monitor infants for jaundice and advice on how this should be done. The AAP has
produced a parent information document entitled Jaundice FAQs (Appendix F)
which is available at http://www.aap.org/family/jaundicefaw.htm.
TABLE 35-28 AAP RECOMMENDATIONS FOR TIMING OF FOLLOW UP
VISITS AFTER NEWBORN DISCHARGE
Followup should be provided as follows:
Infant Discharged
Should Be Seen by Age
Before age 24 hours

72 hours
Between 24 and 47.9 hours
96 hours
Between 48 and 72 hours
120 hours
For some newborns discharged before 48 hours, two followup visits may be
required; the first between 24 and 72 hours, and the second between 72
and 120 hours. Newborns discharged at 48 to 72 hours should be seen for
a followup visit at 72 to 120 hours. Clinical judgment should be used in
determining followup. Earlier or more frequent followup should be provided
for those who have risk factors for hyperbilirubinemia (see Table 35-26),
while those discharged with few or no risk factors can be seen after longer
intervals.
From Maisels MJ, Baltz RD, Bhutani V, et al. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation.
TABLE 35-29 GUIDELINES FOR THE USE OF PHOTOTHERAPY AND

EXCHANGE TRANSFUSION IN LOW-BIRTH-WEIGHT INFANTS BASED ON
BIRTH WEIGHT
a
Total bilirubin level (mg/dL [mol/L] )

b
Birth Weights (g)
Phototherapy
1,500
5-8 (85-140)
13-16 (220-275)
1,500-1,999
8-12 (140-200)
16-18 (275-300)
2,000-2,499
11-14 (190-240)
18-20 (300-340)
Note that these guidelines reflect ranges used in neonatal intensive care
units. They cannot take into account all possible situations. Lower bilirubin
concentrations should be used for infants who are sick (e.g., sepsis,
acidosis, hypoalbuminemia) or who have hemolytic disease.
a
Consider initiating therapy at these levels. Range allows discretion based
on clinical conditions or other circumstances. Note that bilirubin levels refer

to total serum bilirubin concentrations. Direct-reacting or conjugated
bilirubin levels should not be subtracted from the total.
b
Used at these levels and in therapeutic doses, phototherapy should, with

few exceptions, eliminate the need for exchange transfusions.
Levels for exchange transfusion assume that bilirubin continues to rise or

remains at these levels despite intensive phototherapy.
Treatment
Hyperbilirubinemia can be treated in three ways: (a) exchange transfusion removes bilirubin mechanically; (b)
phototherapy converts bilirubin to products that can bypass the liver's conjugating system and be excreted in
the bile or in the urine without further metabolism; and (c) pharmacologic agents that interfere with heme
degradation and bilirubin production, accelerate the normal metabolic pathways for bilirubin clearance, or inhibit
the enterohepatic circulation of bilirubin. Phototherapy is the most common treatment in use for
hyperbilirubinemia; exchange transfusions generally are reserved for phototherapy failures. The bilirubin level at
which intervention is necessary is still a contentious issue.
TABLE 35-30 GUIDELINES FOR USE OF PHOTOTHERAPY AND EXCHANGE TRANSFUSION IN PRETERM
INFANTS BASED ON GESTATIONAL AGE
Total bilirubin level (mg/dL [mol/L])
a
Gestational Age (Weeks)
Phototherapy
Sick
36
32
28
24
14.6 (250)
8.8 (150)
5.8 (100)
4.7 (80)
17.5 (300)
14.6 (250)
11.7 (200)
8.8 (150)
Well
20.5
17.5
14.6
11.7
(350)
(300)
(250)
(200)
Rhesus disease, perinatal asphyxia, hypoxia, acidosis, hypercapnia.

From Ives NK. Neonatal jaundice. In: Rennie JM, Roberton NRC, eds. Textbook of neonatology. New
York:Churchill Livingston, 1999:715732, with permission.
Principles Underlying the Recommendations

Recommendations and guidelines for the use of phototherapy and exchange transfusion in term, near-term and
preterm infants are provided in Tables 35-29 to 35-32 and Figs. 35-22 and 35-23. It would be ideal if the
guidelines for the implementation of phototherapy and exchange transfusion relied on evidence-based estimates
of when the benefit of these interventions exceeded their risks and costs (76). Ideally, these estimates should
come from randomized trials or high quality, systematic observational studies but such studies are rare
(139,140). Thus, treatment guidelines must rely on relatively uncertain estimates of risks and benefits
P.819
and the recognition that using a single TSB level to predict long-term behavioral and developmental outcomes is
not reliable and will lead to conflicting results (139,140).
TABLE 35-31 GUIDELINES ACCORDING TO BIRTH WEIGHT FOR EXCHANGE TRANSFUSION IN LOWBIRTH-WEIGHT INFANTS BASED ON TOTAL BILIRUBIN (mg/dL) AND BILIRUBIN-TO-ALBUMIN
RATIO (mg/g) (WHICHEVER COMES FIRST)
Standard risk
Total bilirubin
B:A ratio
<1,250 g
1,2501,499 g
1,5001,999 g
2,0002,499 g
13
15
17
18
5.2
6.0
6.8
7.2
10
4.0
13
5.2
15
6.0
17
6.8
High risk
Total bilirubin
B:A ratio
Risk factors: Apgar <3 at 5 minutes; PaO2 <40 mm Hg 2 h; pH 7.15 1 h; birth weight <1000 g,
hemolysis; clinical or central nervous system deterioration; total protein 4 g/dL or albumin 2.5 g/dL.
B:A ratio, bilirubin-to-albumin ratio.
From Ahlfors CE. Criteria for exchange transfusion in jaundiced newborns. Pediatrics 1994;93:488
494, with permission.
The background to treatment decisions for hyperbilirubinemia has been provided (see section on Bilirubin
Toxicity.) The basic principles underlining the recommendations given in Tables 35-29, 35-30, 35-31 and 35-32
and Figs. 35-22 and 35-23 are as follows:
Phototherapy is a highly effective method for preventing and treating hyperbilirubinemia (see
Phototherapy below). The main demonstrated value of phototherapy is that it reduces the need for
exchange transfusion (177,217, 462,463). Ip and associates have calculated that about 5 to 10 term and
near-term infants with TSB levels between 15 and 20 mg/dL (257 to 342 mol/L) will need to receive
phototherapy in order to prevent the TSB in 1 infant from reaching 20 mg/dL (the number needed to
treat) (139,140). This means that about 8 to 9 of every 10 infants with these TSB levels will not reach 20
mg/dL (342 mol/L), even if they are not treated. As phototherapy has proven to be a generally safe
procedure, however, the (unnecessary) treatment of many infants is considered appropriate in order to
prevent some infants from reaching TSB levels that are considered potentially dangerous (see
Phototherapy below for a discussion of its safety and complications).
The recommended TSB levels for exchange transfusion are intended to keep TSB levels below those at
which kernicterus has been reported (124,139,140), although it is recognized that rare cases of
kernicterus have occurred in apparently healthy infants and in sick infants at unexpectedly low TSB levels
(43,124,139, 140). With rare exceptions, exchange transfusion is recommended only after intensive
phototherapy has failed to keep the TSB level below the exchange transfusion level (76).
As discussed above (see section on Bilirubin Toxicity) infants (particularly those with the risk factors listed
in Tables 35-29,35-30 and 35-31 and Figs. 35-22 and 35-23) (43,45,154) and preterm infants
(45,150,152,464) are at a greater risk for developing kernicterus at lower bilirubin levels than are well
term infants, although
P.820
some studies have not confirmed all of these associations (152,178,179). On the other hand, there is no
doubt that infants of lower gestation are at a greater risk of developing high TSB levels (203,206).
Because one of the primary goals of treatment is to prevent further increases in the TSB levels, treatment
is recommended at lower TSB levels at younger ages.
TABLE 35-32 GUIDELINES FOR INITIATING PHOTOTHERAPY AND EXCHANGE TRANSFUSIONS IN

NICHHD NEONATAL RESEARCH NETWORK TRIAL (B. MORRIS, PERSONAL COMMUNICATION, 2002)
Birth Weight
(g)
Aggressive Management
Exchange
Phototherapy Begins
Transfusion
501750
ASAP after enrollment
7511000
ASAP after enrollment
13.0 mg/dL
Conservative Management
Exchange
Phototherapy Begins
Transfusion
8.0 mg/dL
13.0 mg/dL
15.0 mg/dL
10.0 mg/dL
15.0 mg/dL
Enrollment is expected within 1236 hours after birth, preferably between 12 and 24 hours.
ASAP, as soon as possible.
From Maisels MJ, Watchko JF. Treatment of jaundice in low birthweight infants. Arch Dis Child Fetal
Neonatol Ed 2003;88:F459F463, with permission.
Figure 35-22 AAP guidelines for phototherapy in hospitalized infants of 35 or more weeks' gestation (76).
Note: These guidelines are based on limited evidence and the levels shown are approximations. The guidelines
refer to the use of intensive phototherapy which should be used when the TSB exceeds the line indicated for
each category. Infants are designated as higher risk because of the potential negative effects of the
conditions listed on albumin binding of bilirubin (77,96), the blood-brain barrier (60), and the susceptibility of
the brain cells to damage by bilirubin (55).
Intensive phototherapy implies irradiance in the blue-green spectrum (wavelengths of approximately 430 to
490 nm) of at least 30 W/cm2 per nm (measured at the infant's skin directly below the center of the
phototherapy unit) and delivered to as much of the infant's surface area as possible. Note that irradiance
measured below the center of the light source is much greater than that measured at the periphery.
Measurements should be made with a radiometer specified by the manufacturer of the phototherapy system.
If total serum bilirubin levels approach or exceed the exchange transfusion line (Fig. 35-23) the sides of the
bassinet, incubator, or warmer should be lined with aluminum foil or white material (575). This will increase the
surface area of the infant exposed and increase the efficacy of phototherapy (472).
If the total serum bilirubin does not decrease or continues to rise in an infant who is receiving intensive
phototherapy, this strongly suggests the presence of hemolysis.
Infants who receive phototherapy and have an elevated direct-reacting or conjugated bilirubin level (cholestatic
jaundice) may develop the bronze baby syndrome. See section on phototherapy for the use of phototherapy in
these infants. (From Maisels MJ, Baltz RD, Bhutani V, et al. Management of hyperbilirubinemia in the newborn
infant 35 or more weeks of gestation. Pediatrics 2004;114:297-316, with permission.)
It is recognized that it is quite often difficult, and sometimes impossible, to rule out an underlying hemolytic
process. There is evidence, for example, that the majority of infants with elevated serum bilirubin levels in the
first 2 to 4 days of life have an increase in heme turnover, if not frank hemolysis (465). Although measurements
of ETCOc can identify the hemolyzing infant (316,336,455,461, 465), standard diagnostic tests for hemolysis,
such as the reticulocyte count, hematocrit, or examination of the peripheral smear, are neither sensitive nor
specific (218, 466). Nevertheless, a recent study did find a good correlation between measurements of ETCOc
and reticulocyte counts (corrected for the hematocrit) in infants with positive Coombs tests (455).
Most term and near-term infants readmitted with hyperbilirubinemia do not have documented hemolytic disease
(241,248,249,467). Of 61 cases of kernicterus, 10 (15%) were considered to have hemolysis and 19 of 61
(32%) were G6PD deficient (124). It is important to note that infants with G6PD deficiency are frequently not
anemic and often have none of the classical manifestations of hemolytic disease (326,327).
Figure 35-23 AAP guidelines for exchange transfusion in infants who are 35 or more weeks' gestation (76).
Note that these suggested levels are based on limited evidence, and the levels shown are approximations. See
section on exchange transfusion for risks and complications of exchange transfusion. During birth
hospitalization, exchange transfusion is recommended if the TSB rises to these levels despite intensive
phototherapy. For readmitted infants, if the TSB level is above the exchange level, repeat TSB measurement
every 2 to 3 hours and consider exchange if the TSB remains above the levels indicated after intensive
phototherapy for 6 hours.
P.821
Risks
When used properly, phototherapy is a very effective and safe method of lowering the serum bilirubin
concentration. Its use has drastically decreased the need for exchange transfusion (Figs. 35-24 and 35-25) and
it has probably contributed to the virtual disappearance of kernicterus in the low-birth-weight infant.
Conversely, in full-term and near-term newborns, phototherapy can lead to separation of the mother and infant,
increase parental concern, decrease the likelihood of successful breast-feeding, and adversely affect the
P.822
motherinfant relationship (356). At a time when pediatricians are trying to promote breast-feeding in the face of
considerable odds (e.g., early discharge from hospital, the rapid return of mothers to the workforce, commercial
marketing of formulas), attention should be given to an intervention that has a negative impact on the nursing
mother.
Figure 35-24 Number of infants in different populations who received exchange transfusions between 1957
and 1997. (1) In Athens, Greece, exchange transfusions were performed on 134 infants of 30,830 live births
(birth weight 2,500 g) between 1957 and 1961 and on 91 infants of 180,594 live births between 1980 and
1992 (577). (2) At the University of Melbourne, Melbourne, Australia, between 1971 and 1980, exchange
transfusions were performed in 114 infants of 41,057 live births and between 1981 and 1989, in 134 infants of
47,080 live births (all birth weights) (578). (3) Between 1988 and 1997 exchange transfusions were performed
in 8 infants of 55,128 live births at William Beaumont Hospital, Royal Oak, MI (all birth weights) (579).
The risks of exchange transfusion include risks from the transfused blood and from the procedure (see section
on Exchange Transfusion). Experience with exchange transfusion is decreasing (see Figs. 35-24 and 35-25) and
with new types of phototherapy (463) and other interventions in immune hemolytic disease, it is likely to
decrease even further (370,468,469). It is now common for a resident to complete a 3-year pediatric training
program without ever having performed an exchange transfusion. Under the circumstances, the mortality and
morbility for this procedure is likely to increase in the years ahead.
Figure 35-25 Number of infants in different populations with birth weight <1,500 g who received exchange
transfusions between 1974 and 1997. (1) A total of 215 newborns weighing <1,500 g in the NICHD Cooperative
Phototherapy Trial were assigned to the control group (did not receive phototherapy) (540). Seventy-seven of
215 infants (35.8%) received a total of 161 exchange transfusions. In the phototherapy group, 17 of 196
(8.7%) infants received exchange transfusions. These data are included to illustrate the frequency of exchange
transfusion before the introduction of phototherapy. (2) Of a total of 1,338 live births weighing <1,500 g in the
Netherlands in 1983, 37 infants (2.8%) required at least 1 exchange transfusion (580). (3) Of 833 live births
(weighing 500 to 1,500 g) in a 17-county region in North Carolina, 2 infants required an exchange transfusion
(0.24%) (182). (4) No exchange transfusions were performed in 1,213 live births weighing <1,500 g at the
William Beaumont Hospital, Royal Oak, MI between 1988 and 1997 (579).
P.823
Treating Full-Term and Near-Term Newborns

Most infants who are treated for neonatal jaundice have no identifiable pathology and no evidence of hemolytic
disease (see Table 35-18) (218,241,248,249). Approximately 90% are fully or partially breast-fed (249).
Guidelines for the treatment of full-term and near-term infants are given in Figs. 35-22 and 35-23. Table 35-33
is an example of a clinical pathway for the management of a newborn readmitted for phototherapy or exchange
transfusion.
Breast-Fed Infants
Of infants who develop bilirubin levels high enough to require phototherapy and who do not have evidence of
isoimmunization or other obvious hemolytic disease, 80% to 90% are fully or partially breast-fed (218,249).
Table 35-34 provides an approach to the prevention and treatment of jaundice associated with breast-feeding.
Observational studies show that increasing the frequency of breast-feeding during the first few days after birth
decreases TSB levels (251,252,409). We performed a controlled trial in which mothers were randomly assigned
to a frequent or demand breast-feeding schedule. We found no significant difference between the TSB levels
measured in the two groups at an average age of 55 hours (470). Given the natural history of jaundice in breastfed infants, however, it is certain that maximum TSB had not yet been achieved in these infants. However, the
observational data of Yamauchi and Yamanouchi (252) show a very strong inverse and linear relationship
between the frequency of nursing in the first 24 hours and the probability of hyperbilirubinemia on day 6.
TABLE 35-33 EXAMPLE OF A CLINICAL PATHWAY FOR MANAGEMENT OF THE NEWBORN INFANT
READMITTED FOR PHOTOTHERAPY OR EXCHANGE TRANSFUSION
Treatment
Use intensive phototherapy and/or exchange transfusion as indicated in Figures 35-22 and 35-23.
Laboratory tests
TSB and direct bilirubin level
Blood type (ABO, Rh)
Direct antibody (Coombs) test
Serum albumin
Complete blood cell count with differential and smear for red cell morphology
Reticulocyte count
ETCOc (if available)
Glucose-6-phosphate dehydrogenase if suggested by ethnic or geographic origin, or if poor response
to phototherapy
Urine for reducing substances
If history and/or presentation suggest sepsis, perform blood culture, urine culture, cerebral spinal
fluid for protein, glucose, cell count, and culture
Interventions
If TSB 25 mg/dL (428 mol/L) or 20 mg/dL (342 mol/L) in a sick infant or infant <38 weeks'
gestation, obtain a type and crossmatch, and request blood in case an exchange transfusion is
necessary
In infants with isoimmune hemolytic disease and TSB rising in spite of intensive phototherapy or
within 23 mg/dL (3451 mol/L) of exchange level (see Fig. 35-24), administer intravenous
immunoglobolin 0.51 g/kg over 2 hours and repeat in 12 hours if necessary
If infant's weight loss from birth is >12%, or if there is clinical or biochemical evidence of
dehydration, recommend formula or expressed breast milk; if oral intake is in question, give
intravenous fluids
For infants receiving intensive phototherapy
Breast-feed or bottle-feed (formula or expressed breast milk) every 23 hours

If TSB 25 mg/dL (428 mol/L) repeat TSB within 23 hours
If TSB 2025 mg/dL (342428 mol/L), repeat within 34 hours; if TSB <20 mg/dL (342 mol/L)
repeat in 46 hours; if TSB continues to fall, repeat in 812 hours
If TSB is not decreasing, or is moving closer to level for exchange transfusion, or TSB: albumin ratio
exceeds levels shown in Fig. 35-23 consider exchange transfusion; see Fig. 35-23 for exchange
transfusion recommendations
When TSB is below 1314 mg/dL (239 mol/L) discontinue phototherapy
Depending on the cause of the hyperbilirubinemia, it is an option to measure TSB 24 hours after
discharge to check for rebound
From Maisels MJ, Baltz RD, Bhutani V, et al. Management of hyperbilirubinemia in the newborn infant
35 or more weeks of gestation. Pediatrics 2004;114:297316, with permission.
In many hospitals it is a common practice to provide supplemental feedings of water or dextrose water to breastfed infants in the mistaken belief that this will lower their TSB levels. On the contrary, this practice does not
lower TSB levels and should be abandoned (see Fig. 35-21) (419,420). Furthermore, Kuhr and Paneth (424)
found that an increase in dextrose water intake in the first 3 days of life was significantly related to a decrease
in breast-milk intake on the fourth day.
When the TSB in a breast-fed infant reaches a level at which intervention is being considered, a number of
options exist (see Table 35-34). Two randomized controlled trials evaluated some of these interventions. When
TSB levels reached 15 mg/dL (257 mol/L), Amato and associates (471) compared the effect of interrupting
breast-feeding versus
P.824
phototherapy. There was no difference between the groups in the amount of time needed to reduce the bilirubin
to less than 12 mg/dL (205 mol/L). When the TSB level reached 17 mg/dL (291 mol/L), Martinez and
colleagues assigned 125 full-term breast-fed infants to one of four interventions (462); Table 35-35 shows the
results. Although discontinuing breast-feeding and using phototherapy was the most effective strategy, in
neither of these studies was intensive phototherapy used (472). Notwithstanding the options listed in Table 3534, any interruption of nursing is undesirable and it is my practice to recommend that breast-feeding be
continued while the infant is treated with intensive phototherapy unless very severe hyperbilirubinemia is
present (TSB levels in excess of 25 mg/dL [428 mol/L]) (see Phototherapy below).
TABLE 35-34 APPROACHES TO THE PREVENTION AND TREATMENT OF JAUNDICE ASSOCIATED WITH
BREAST-FEEDING
Prevention
1. Encourage frequent nursing (i.e., at least eight times per day)
2. Do not supplement with water or dextrose water
Treatment options
1.
2.
3.
4.
Increase frequency of nursing

Continue nursing, administer phototherapy
Temporarily interrupt nursing for 48 hours, substitute formula
If treating with phototherapy, supplement with expressed breast milk or formula if infant's intake is
inadequate, weight loss is excessive, or the infant is dehydrated
Low-Birth-Weight Infants
The management of hyperbilirubinemia in low-birth-weight infants has been reviewed (473) and Tables 3529,35-30 to 35-31 provide suggested guidelines from three different sources. Over the last 2 decades, there has
been a remarkable decrease in the incidence of kernicterus found at autopsy in infants who died in NICUs. Some
of this may be a result of the liberal use of phototherapy. Certainly, phototherapy has dramatically decreased
the necessity for exchange transfusion, which, in low-birth-weight infants, is now almost exclusively carried out
in the occasional infant with severe Rh hemolytic disease or extensive bruising (463).
TABLE 35-35 EFFECT OF INTERVENTIONS FOR JAUNDICE IN BREAST-FED INFANTS

Group
1
Discontinue Breast-
Discontinue Breastfeeding Substitute
Continue Breast-
Formula, Use
Phototherapy
feeding, Use
Phototherapy
38
36
5 (14%)
Continue Breast- feeding, Substitute

feeding
Formula
No.
Serum
bilirubin 20
25
26
6 (24%)
5 (19%)
1 (3%)
mg/dL (342
mol/L)
When the serum bilirubin reached 17 mg/dL (291 mol/L), infants were randomly assigned to one of
the four interventions above. The number (%) of infants whose bilirubin level subsequently reached or
exceeded 20 mg/dL (342 mol/L) is shown. Conventional (not intensive) phototherapy was used.
a
Significantly different versus group 1 (p = 0.013) and group 2 (p = 0.036). From Martinez JC,
Maisels MJ, Otheguy L, et al. Hyperbilirubinemia in the breast-fed newborn: a controlled trial of four
interventions. Pediatrics 1993;91:470473, with permission.
As shown in Tables 35-29,35-30 and 35-31, phototherapy generally is used according to a sliding scale: the
lower the birth weight, the lower the TSB level at which phototherapy is instituted. In view of the known
antioxidant properties of bilirubin, it is possible that maintaining very low TSB levels by the aggressive use of
phototherapy might have other, less desirable, consequences (181) (see section on Bilirubin Toxicity-Premature
Infants and Low Bilirubin Kernicterus and Physiologic Role of Bilirubin below). To address this and other issues
the, NICHHD Neonatal Research Network initiated a trial, the design of which is shown in Table 35-32 (192). The
results of this important study should provide guidance with regard to the risks and benefits of phototherapy in
the ELBW population.
P.825
Elevated Direct-Reacting or Conjugated Bilirubin Levels

There are no helpful data and, as a result, little guidance on how we should deal with the occasional infant who
has a high TSB as well as a significant elevation of direct-reacting bilirubin. Because direct-reacting bilirubin is
not toxic to the central nervous system, some practitioners have based their decisions for exchange transfusion
on the level of indirect-reacting serum bilirubin level (rather than the TSB) as the sole criterion for exchange
transfusion (76). The AAP makes a firm recommendation against this practice and notes that the direct-reacting
(or conjugated) bilirubin should not be subtracted from the total.
Kernicterus has been described in infants who have elevated TSB levels but in whom the indirect-reacting
bilirubin was well below 20 mg/dL (342 mol/L) (135, 474). One infant had the bronze baby syndrome and
another was a 54-hour-old male infant with erythroblastosis fetalis whose TSB level was 45.2 mg/dL (773 mol/
L), of which 31.6 mg/dL (540 mol/L) was direct reacting. Thus, the total indirect-reacting bilirubin in this infant
was only 13.6 mg/dL (135). I have seen the records of other, similar cases. Kernicterus in these infants may be
the result of competitive displacement (by direct-reacting bilirubin) of unconjugated bilirubin from its binding
site to albumin.
Ebbesen (475) found that infants with elevated direct-reacting bilirubin levels of 6.4 to 9.9 mg/dL (109 to 169
mol/L) and the bronze baby syndrome had a decrease in reserve albumin-binding capacity. Conversely, some
infants with extremely high, but predominantly direct-reacting, serum bilirubin levels have come to no harm.
Hemolytic Disease
As discussed previously, infants with hemolytic disease appear to be at a greater risk of developing bilirubin
encephalopathy than are nonhemolyzing infants with similar TSB levels. The reasons for this are not clear. In
the early studies of Rh disease, almost all infants were delivered prematurely (to prevent stillbirth); many were
asphyxiated and severely ill. It is unlikely that the risk of kernicterus in infants with Rh disease, treated in
today's intensive care environment and with similar TSB levels, would be nearly as great. Although it has been
suggested that infants with hemolytic disease may have a decrease in their bilirubin-binding capacity, when
measured, this has not been found to be the case (476). Similarly, we have no obvious explanation for the
increased risk of bilirubin encephalopathy in infants with G6PD deficiency.
Hydrops Fetalis
Hydropic infants generally suffer significant hypoxia in utero. Women who are to deliver such infants should be
managed exclusively in perinatal centers capable of the full range of obstetric and neonatal intensive care.
Hydropic infants, as well as infants who are severely anemic (hematocrit less than 35%) and asphyxiated,
require immediate treatment. Exchange transfusion of about 50 mL/kg of packed cells soon after birth raises the
hematocrit to approximately 40%. Phlebotomy should not be routinely performed on these infants because they
usually are normovolemic and may be hypovolemic (306, 477,478). No manipulations of blood volume should be
performed without appropriate measurements of central venous and arterial blood pressures. For accurate
monitoring of central venous pressure, however, the umbilical venous catheter must enter the inferior vena cava
by way of the ductus venosus. If the catheter is in a portal vein or the umbilical vein, the pressures so measured
are meaningless and preclude interpretation of the infant's circulatory status. In addition, before making
therapeutic decisions based on measurements of central venous pressure, the physician must also correct
acidosis, hypercarbia, hypoxia, and anemia. Serum glucose levels should be monitored carefully, because
hypoglycemia is common.
Phototherapy
For comprehensive reviews of this subject, the reader is referred to the monograph by Jahrig and
associates (479), and to a recent chapter by the author (480).
It helps to understand how phototherapy works if we consider that light is an infusion of discrete
photons of energy that correspond to the individual molecules of a drug in a conventional
medication. Absorption of these photons by bilirubin molecules in the skin leads to the therapeutic
effect in much the same way as binding of drug molecules to a receptor has a desired effect.
Tables 35-36 to 35-38 list the factors that influence the dose and efficacy of phototherapy and
define the radiometric quantities used in assessing the dose.
Terminology
Light Spectrum
The spectrum of light delivered by the phototherapy unit is determined by the type of light source
and any filters used. Because of the optical properties of bilirubin and skin, the most effective
lights are those with wavelengths that are predominantly in the blue-green spectrum (481).
TABLE 35-36 FACTORS THAT DETERMINE THE DOSE OF PHOTOTHERAPY
Spectrum of light emitted

Irradiance of light source
Design of phototherapy unit
Surface area of infant exposed to the light
Distance of infant from light source
TABLE 35-37 FACTORS THAT AFFECT THE DOSE AND EFFICACY OF PHOTOTHERAPY
Factor
Spectrum of light
emitted
Mechanism/Clinical
Implementation and
Relevance
Rationale
Blue-green spectrum
most effective; at
these wavelengths,
light penetrates skin
well and is absorbed
maximally by bilirubin
Special blue fluorescent

tubes or other light
sources that have most
output in the blue-green
spectrum and are most
effective in lowering TSB
Clinical Application
Use special blue

tubes or LED light
source with output
in blue-green
spectrum for
intensive PT
Spectral irradiance irradiance

Irradiance measured with
(irradiance in
rate of decline in TSB a radiometer as W/cm2/
certain wavelength
nm; standard PT units
band) delivered to
deliver 810 W/cm2/nm;
If special blue
surface of infant
W/cm2/nmtubes 1015
as possible to
increase
cm above infant
irradiance; Note:
intensive PT requires 30
fluorescent tubes
used, bring tubes
as close to infant
cannot do this
with halogen
lamps because
danger of burn;
special blue will
produce an
irradiance of at
least 35 W/cm2/nm
Spectral power
(average spectral
surface area
irradiance across
rate of decline in TSB infant to PT
exposed
surface area)
For intensive PT expose

Place lights above
maximum surface area of and fiberoptic pad
or special blue
fluorescent tubes
below infant; for
maximum
exposure line sides
of bassinet warmer
bed, or incubator
with aluminum foil
Cause of jaundice
PT likely to be less
effective if jaundice
caused by hemolysis
or if cholestasis
present (directreacting bilirubin)
TSB level at start

of PT
The higher the TSB,

the more rapid the
decline in TSB with PT
When hemolysis
present start PT at
lower TSB levels.
Use intensive PT.
Failure of PT
suggests hemolysis
is cause of
jaundice; if directreacting bilirubin,
watch for bronze
baby syndrome or
blistering
Use intensive PT
for higher TSB
levels; anticipate
more rapid
decrease in TSB
when TSB >20 mg/
dl (342 mol/l)
LED, light-emitting diode; PT, phototherapy; TSB, total serum bilirubin.

a
Available in Olympic BiliBassinet (Olympic Medical, Seattle, WA). From Maisels MJ, Baltz
RD, Bhutani V, et al. Management of hyperbilirubinemia in the newborn infant 35 or more

weeks of gestation. Pediatrics 2004;114:297316, with permission.
P.826
There is a common misconception that ultraviolet (UV) light is used for phototherapy. None of the
light systems described emit any significant amount of UV radiation, and the small amount of UV
light that is emitted by fluorescent tubes is in longer wavelengths (>320 nm) than those that
cause erythema (479). In any case, almost all UV light produced is absorbed by the glass wall of
the fluorescent tube and by the Plexiglas cover of the phototherapy unit.
Irradiance
Irradiance is the radiant power incident on a surface per unit area of the surface.
TABLE 35-38 RADIOMETRIC QUANTITIES USED
Quantity
Irradiance (radiant power

incident on a surface per unit
Dimensions
Usual Units of Measure
W/m2
W/cm2
Spectral irradiance (irradiance

in a certain wavelength band)
W/m2 per nm (or W/m2)
W/cm2 per nm
Spectral power (average
W/m
mW/nm
area of the surface)
spectral irradiance across a

surface area)
From Maisels MJ. Why use homeopathic doses of phototherapy?
Pediatrics1996;98:283287, with permission.
The irradiance in a specific wavelength band is called the spectral irradiance and is expressed as
W/cm2/nm (see Table 35-38). There is a direct relationship between the efficacy of
phototherapy and the irradiance used (Fig. 35-26) (482) and the irradiance is directly related to
the distance between the light and the infant (Fig. 35-27) (472).
The Dose-Response Relationship of Irradiance

Figure 35-26 shows that there is a direct relationship between the irradiance used and the rate at
which the serum bilirubin declines under phototherapy (482). The data in Fig. 35-26 suggest that
there is a saturation point
P.827
beyond which an increase in the irradiance produces no added efficacy. We do not know,
however, that a saturation point exists. As the conversion of bilirubin to excretable photoproducts
is partly irreversible and follows first-order kinetics, there may not be a saturation point so we do
not know the maximum effective dose of phototherapy.
Figure 35-26 Relationship between average spectral irradiance and decrease in serum bilirubin
concentration. Full-term infants with nonhemolytic hyperbilirubinemia were exposed to special
blue light (Phillips TL 52/20W) of different intensities. Spectral irradiance was measured as the
average of readings at the head, trunk, and knees. Drawn from the data of Tan (482). (From
Maisels MJ. Why use homeopathic doses of phototherapy? Pediatrics 1996;98:283-287, with
permission.)
Effect on Irradiance of the Light Spectrum and the

Distance between the Infant and the Light Source.
Figure 35-27 shows that the light intensity (measured as spectral irradiance) is inversely related
to the distance from the source. The relationship between intensity and distance is almost (but
not quite) linear, indicating that these data do not obey the law of inverse squares, which states
that the light intensity will decrease with the square of the distance (472). This law applies only to
a point source of light and the light source in phototherapy units has features of a cylindrical and
planar source, not a point source. Figure 35-27 also demonstrates the dramatic difference in
irradiance produced within the 425- to 475-nm band by different types of fluorescent tubes.
Figure 35-27 Effect of light source and distance from the light source to the infant on average
spectral irradiance. Measurements were made across the 425- to 475-nm band using a
commercial radiometer (Olympic Bilimeter Mark II). The phototherapy unit was fitted with eight
24-inch fluorescent tubes. () Special blue, General Electric 20-W F20T12/BB tube; () blue,
General Electric 20-w F20T12/B blue tube; ([black up-pointing triangle]) daylight blue, four
General Electric 20-W F20T12/B blue tubes and four Sylvania 20-W F20T1 2/D day light tubes;
(^) daylight, Sylvania 20-W F20T12/D daylight tubes. Curves were plotted using linear curve
fitting (True Epistat; Epistat Services, Richardson, TX). The best fit is described by the equation
y=AeBX. (From Maisels MJ. Why use homeopathic doses of phototherapy? Pediatrics
Spectral Power
This is the product of the skin surface irradiance and the spectral irradiance across this surface
area. Because irradiance and the surface area of the infant exposed to phototherapy are key
elements in determining the efficacy of phototherapy, the use of spectral power is the only
meaningful way to compare the dose of phototherapy received by infants under the different
phototherapy systems (472). Calculations of spectral power show why a much more effective
dose of phototherapy is delivered to the infant by using the appropriate fluorescent tubes than
can be delivered using a fiberoptic phototherapy system (472).
Mechanism of Action
Phototherapy detoxifies bilirubin by converting it to photoproducts that are less lipophilic than
bilirubin and can bypass the liver's conjugating system and be excreted without further
metabolism (10). It is not known exactly where the process of phototherapy takes place, but it
probably does not take place in the skin cells. It is more likely that it works on bilirubin bound to
albumin in the superficial capillaries or in the interstitial space (483).
Bilirubin Photochemistry
When bilirubin absorbs light, photochemical reactions occur. Although many such reactions have
been observed in vitro, only three have been shown to occur in vivo during phototherapy.
Configurational(Z E) Isomerization
Isomers are substances that have the same molecular formula but different physicochemical
properties. There are four possible configurational isomers of bilirubin (Fig. 35-28). In infants
receiving phototherapy, the stable 4Z,15Z isomer is converted predominately to the 4Z,15E
isomer (Figs. 35-28 and 35-29).(10) The formation of 4Z,15E bilirubin is spontaneously reversible
in the dark and occurs rapidly in bile. Thus the 4Z,15E bilirubin formed in the skin and excreted
by the liver is readily converted back to ordinary unconjugated bilirubin. The conversion of the
4Z,15Z isomer to the 4Z,15E isomer is also reversible by light (10).
Figure 35-28 Configurational and structural isomers of 4Z, 15Z bilirubin in infants undergoing
phototherapy.
P.828
When infants are exposed to phototherapy, photoisomerization occurs almost instantaneously, but
the clearance of the light-generated 4Z,15E isomer is very slow (T~15 hours). Thus, although
configurational isomerization is extremely rapid and accounts for the bulk of the photochemical
outcomes, it probably plays only a minor role in lowering the serum bilirubin concentration
because although it is formed fastest, it has nowhere to go (484).
Structural Isomerization
In this reaction (Fig. 35-30), intramolecular cyclization of bilirubin (an irreversible process) occurs
in the presence of light to form a substance known as lumirubin that can be excreted in bile
(without the need for conjugation) and in urine (but at a much lower rate than in bile) (484).
During phototherapy, the serum concentration of lumirubin is about 2% to 6% of the TSB, which
is much lower than the concentration of the configurational isomers that form about 20% of the
total bilirubin. But, because lumirubin is cleared from the serum much more rapidly than the
4Z,15E isomer, it is likely that lumirubin formation is mainly responsible for the phototherapyinduced decline in serum bilirubin in the human infant (484). It is quite possible, however, that
the contribution of the other isomers, 4E,15Z and 4E,15E bilirubin, also is important (485).
Figure 35-29 Z-E carbon-carbon double-bond configurational isomerization of bilirubin in

humans. (From McDonagh AF, Lightner DA. Like a shrivelled blood orangebilirubin, jaundice
and phototherapy. Pediatrics 1985;75:443-455, with permission.)
Figure 35-30 Intramolecular cyclization of bilirubin in the presence of light to form lumirubin.
(From McDonagh AF, Lightner DA. Like a shrivelled blood orangebilirubin, jaundice and
phototherapy. Pediatrics 1985;75:443-455, with permission.)
Photooxidation
Bilirubin can be photooxidized to water-soluble, colorless products that can be excreted in the
urine. This is a slow process and is probably only a minor contributor to the elimination of bilirubin
during phototherapy.
Figure 35-31 summarizes the general mechanisms of phototherapy in neonatal jaundice.
Clinical Use and Efficacy of Phototherapy

There are more than 50 published controlled trials of the clinical use of phototherapy (see Maisels
(217) for a description and analysis of these studies), confirming that phototherapy is effective in
preventing and treating hyperbilirubinemia and dramatically reduces the need for exchange
transfusion (217).
As described previously, there is a clear relationship between the dose of phototherapy and the
measured decrement in the serum bilirubin level (see Fig. 35-26) (482). Tables 35-36 and 35-37
list the factors that affect the dose and efficacy of phototherapy. The other important
P.829
factors that determine the efficacy of phototherapy are the cause of the hyperbilirubinemia
(hemolytic conditions respond less well) and the initial bilirubin level, the rate of decline being
proportional to the initial bilirubin concentration (159,486,487).
Figure 35-31 General mechanisms of phototherapy for neonatal jaundice. Chemical reactions
(solid arrows) and transport processes (broken arrows) are indicated. Pigments may be bound to
proteins in compartments other than blood. Some excretion of photoisomers, particularly
lumirubin, in urine also occurs. (From McDonagh AF, Lightner DA. Like a shrivelled blood
orangebilirubin, jaundice and phototherapy. Pediatrics 1985;75:443-455, with permission.)
Light Sources
Fluorescent Tubes
Daylight or cool-white fluorescent tubes provide adequate phototherapy where the objective is to
control a slowly rising serum bilirubin level in a preterm or term infant, but they are less effective
than special blue tubes (see Fig. 35-27). Special blue fluorescent tubes are most effective
because they provide light predominantly in the blue-green spectrum. At these wavelengths light
penetrates skin well and is absorbed maximally by bilirubin (472,484). The imprint F20-T12/BB
(General Electric, Westinghouse, Sylvania) or TL52/20W (Phillips, Eindhoven, The Netherlands) is
found on special blue tubes. Note that these are different from regular blue tubes (labeled F20T12/B) (484). Special blue tubes do impart a bluish tinge to the infant and this may obscure
cyanosis. In the term nursery this is of little concern, and if used in the NICU, monitoring by pulse
oximetry is all that is necessary. Turquoise fluorescent light (Osram, Denmark) has recently been
compared with special blue light in the treatment of preterm infants (488). The emission peak of
the turquoise lamps is at 490 nm with a bandwidth of 65 nm. Ebbesen and associates found that
these lamps were as effective as special blue lamps but required 23% less irradiance than special
blue lamps to produce the same effect on the bilirubin levels (488). They suggest that the use of
less irradiance could be associated with fewer side effects.
Light-Emitting Diodes
A new method of delivering high-intensity, narrow-band light has been described (468,489). The
use of high-intensity gallium nitride light-emitting diodes (LEDs) permits high irradiance in the
spectrum of choice (blue, blue-green, etc.) with minimal heat generation. The device is low
weight, low voltage, low power, and portable, and could be an effective means of providing
intensive phototherapy in the hospital or at home. Although only limited clinical trials have been
performed to date (489), an LED phototherapy unit is commercially available (Natus Inc., San
Carlos, CA).
Halogen Lamps
These lamps have the advantage of being more compact than fluorescent systems but, unlike
fluorescent lamps, they cannot be brought close to the infant (to increase the irradiance) without
incurring the risk of a burn. In addition, the surface area covered by most halogen lamps is
relatively small, and the spectral power will be less than that produced by a bank of fluorescent
lights.
Fiberoptic Systems
These systems consist of a light that is delivered from a tungsten-halogen bulb through a
fiberoptic cable and emitted from the sides and ends of the fibers inside a plastic pad (472,490).
These systems have the advantage of not requiring eye patches. They are less bulky than
conventional phototherapy equipment and provide a convenient way to deliver double
phototherapy when it is necessary to expose more of the infant's surface area (see following).
They also are useful for home phototherapy. Their major disadvantage is that they have very low
spectral power because of their small surface area. This is the result of the inverse relationship
between surface area and irradiance. For a given light source, enlarging the pad means that the
light must be distributed over a greater area, thus reducing the irradiance (when compared to a
small pad and the same light source). To achieve high levels of spectral irradiance, manufacturers
must compromise by reducing the size of the pad, which exposes a relatively small surface area of
the infant to the light (490). Conversely, because VLBW infants are so small, fiberoptic pads can
be used quite effectively in these infants (491,492,493) and fiberoptic phototherapy does not
attenuate the normal post prandial increase in intestinal blood flow (493) (see Blood Flow below).
Full-Term and Near-Term InfantsUsing Phototherapy

Effectively
Phototherapy initially was used in low-birth-weight and full-term infants primarily to prevent
slowly rising serum bilirubin levels from reaching levels that might require an exchange
transfusion. Today, phototherapy is often used in full-term and near-term infants who have left
the hospital and are readmitted on days 4 to 7 for treatment of TSB levels of 20 mg/dL (342 mol/
L) or more. These infants need a full therapeutic dose of phototherapy (now termed intensive
phototherapy) to get the bilirubin level down as soon as possible (472,487). Intensive
phototherapy implies the use of high levels of irradiance in the 425- to 490-nm band (usually 30
W/cm2 per nm) delivered to as much of the infant's surface area as possible. How this can be
achieved is described below. The light spectrum, the irradiance, and the infant's exposed surface
area are the key elements in determining the bilirubin response to phototherapy (see Tables 3536 and 35-37). Thus, to provide the most effective phototherapy for these full-term and near-
term infants, we use special blue fluorescent tubes and bring them as close to the infant as
possible. To do this, a full-term infant should be in a bassinet, not an incubator, because the top
of the incubator prevents the light from being brought sufficiently close to the infant. In a
bassinet it is possible to bring the fluorescent lights within about 10 cm of the infant and to
produce a spectral irradiance of more than 50 W/cm2 per nm (see Fig. 35-27). This does not
overheat naked full-term infants. Note, however, that halogen phototherapy lamps cannot be
positioned closer to the infant than recommended by the manufacturers without incurring the risk
of a burn.
P.830
To increase the exposed surface area fiberoptic pads should be placed below the infant. This type
of double phototherapy is approximately twice as effective as single phototherapy in low-birthweight infants and almost 50% better in full-term infants (494,495). Another way of increasing
the surface area of the infant exposed to light is to place a reflecting material (a white sheet or
aluminum foil) within or around the bassinet or incubator so that light is reflected onto the infant's
skin. Different systems of surrounding the infant with fluorescent lights have been described
(487,496,497). Using such a system, Hansen (487) reported declines in serum bilirubins of 10 to
11 mg/dL (170 to 185 mol/L) within 2 hours. On occasion, I have used two or even three
fiberoptic pads to cover almost the entire lower surface of the infant. Although data from Tan
(482) suggest that there is a saturation point beyond which an increase in the irradiance produces
no added efficacy, we do not know that a saturation point exists. Given that the conversion of
bilirubin to excretable photoproducts is partly irreversible and follows first-order kinetics, there
may not be a saturation point. Certainly with existing equipment there is probably no such thing
as an overdose of phototherapy.
Because phototherapy acts on bilirubin that is present in the extravascular space as well as in the
superficial capillaries, it has become a common practice to turn the infant at intervals from supine
to prone and back. Two randomized studies suggest that this is not a good idea. Yamauchi and
associates (498) found that turning term infants every 6 hours did not improve the efficacy of
phototherapy compared with those who are not turned. Shinwell and associates (499) turned
infants from prone to supine every 2.5 hours. They found that infants who were not turned had a
significantly more rapid decrease in their TSB levels in the first 24 hours of phototherapy than did
those who were turned (499).
Measuring the Dose of Phototherapy

Table 35-38 shows the radiometric quantities used in measuring the phototherapy dose. The
quantity most commonly reported in the literature is the spectral irradiance. In the nursery,
spectral irradiance can be measured using commercially available radiometers. These instruments
take a single measurement across a band of wavelengths, typically 425 to 475 nm or 400 to 480
nm. Unfortunately, there is no standardized method for reporting phototherapy dosages in the
clinical literature so it is difficult to compare published studies on the efficacy of phototherapy and
manufacturer's data for the irradiance produced by different systems. Measurements of irradiance
from the same system, using different radiometers, can also produce significantly different
results. The width of the phototherapy lamp's emissions spectrum (narrow vs. broad) will affect
the measured irradiance. Measurements under lights with a very focused emission spectrum (e.g.,
blue-light-emitting diode) will vary significantly from one radiometer to another as the response
spectra of the radiometers vary from manufacturer to manufacturer. Broader-spectrum lights
(fluorescent and halogen) have fewer variations among radiometers. Manufacturers of
phototherapy systems generally recommend the specific radiometer to be used in measuring the
dose of phototherapy when their system is used.
It is also important to recognize that the measured irradiance will vary widely depending on where
the measurement is taken. Irradiance measured below the center of the light source can be more
than double that measured at the periphery, and this drop off at the periphery will vary with
different phototherapy units. Ideally, irradiance should be measured at multiple sites under the
area illuminated by the unit and the measurements averaged. The International Electrotechnical
Commission (500) defines the effective surface area as the intended treatment surface that is
illuminated by the phototherapy light. The commission uses 60 30 cm as the standard-sized
surface.
Is it Necessary to Measure Phototherapy Doses Routinely?

Although it is not necessary to measure spectral irradiance before each use of phototherapy, it is
important to perform periodic checks of phototherapy units to make sure that an adequate
irradiance is being delivered.
Preterm Infants
Conventional phototherapy with daylight or cool-white fluorescent lights, halogen lamps, or
fiberoptic pads have all been effective in decreasing TSB levels and the number of exchange
transfusions performed in low-birth-weight infants (see Fig. 35-25) (217). For those with severe
bruising or hemolytic disease, however, intensive phototherapy (as described previously) should
be used.
Prophylactic Phototherapy in Extremely Low-Birth-Weight

Infants
Earlier data suggested that phototherapy, initiated prophylactically from birth in infants weighing
less than 1,250 g, was no more effective than phototherapy initiated when the TSB reached 5 mg/
dL (85mol/L) (501). In a recent study of infants weighing less than 1,000 g, those who received
prophylactic phototherapy from birth had significantly lower peak TSB levels than did those who
received phototherapy when their TSB levels reached 150 mol/L (8.8 mg/dL). There was no
difference between the groups in the duration of phototherapy (502).
Intermittent Versus Continuous Phototherapy

Clinical studies comparing intermittent with continuous phototherapy have produced conflicting
results (479). In practice, on-off cycles complicate nursing care and probably are more trouble
than they are worth. There is no doubt, however, that in the majority of circumstances
phototherapy does not need to be continuous. It can and certainly should be interrupted during
feeding or brief
P.831
parental visits. On the other hand, when bilirubin levels are very high, intensive phototherapy
should be administered continuously until a satisfactory decline in the TSB has occurred.
Hydration and Feeding

Breast-fed infants who are readmitted to the hospital with TSB levels commonly have excess
weight loss as a result of a combination of mild dehydration and poor caloric intake. In these
infants, it makes sense to provide supplemental calories and fluids using a milk-based formula,
because formulas inhibit the enterohepatic circulation of bilirubin and help to lower the bilirubin
level. Because lumirubin is excreted in the urine, maintaining adequate hydration and good urine
output also helps to improve the efficacy of phototherapy. Routine supplementation (with
dextrose water) of all infants receiving phototherapy is not indicated.
Biologic Effects and Complications

Even though phototherapy has been used in millions of infants for more than 30 years, reports of
significant toxicity are exceptionally rare. Human, animal, and in vitro studies suggest that the
products of photodecomposition have no neurotoxic effects (503,504).
Skin
Bilirubin is a photosensitizer and, in some circumstances, could act as a photodynamic agent in
the presence of light and produce damage. Severe blistering and photosensitivity during
phototherapy for jaundice have been described in infants with congenital erythropoietic porphyria
(505). The presence of congenital porphyria or a family history of porphyria are absolute
contraindications to the use of phototherapy. Rarely, bullous eruptions have occurred in infants
with hemolytic disease and transient porphyrinemia who received phototherapy (506,507). All of
these infants had significant cholestasis (elevated direct-reacting bilirubin levels), and plasma
proto- and coproporphyrin levels were elevated in the two infants in whom they were measured.
Other photosensitizers may also be hazardous in infants receiving phototherapy. A 32-weeks'
gestation neonate who received intravenous fluorescein for angiography was subsequently
exposed to phototherapy (508). This infant developed a partial-thickness burn, which was
probably related to the phototoxicity from the fluorescein that produces photosensitization by
generation of a superoxide anion when exposed to light at a wavelength of 418 nm in the visible
light range (508).
A partial-thickness burn occurred on the back of a 25-weeks' gestation 800-g infant receiving
phototherapy with a fiberoptic system (509). This infant required assisted ventilation and
inotropic support and died at age 85 hours. There was an extensive, erythematous, denuded area
of skin on the back, similar to a partial thickness burn and from which serous fluid oozed. The
Ohmeda Company issued a medical device safety alert on January 26, 1996, in which it reported
that four extremely premature infants (25 weeks' gestation) had developed purplish-red
necrotizing lesions during the use of the BiliBlanket phototherapy system. In all of these infants,
conditions were present that might reduce skin integrity such as birth trauma, hypotension, poor
perfusion of the skin or bacterial contamination of the incubator or bed. It is unlikely that these
were thermal burns although it is important to note that skin of these extremely premature
infants is remarkably fragile. Two neonates have been described who developed erythema (one
was blistering) following exposure to daylight fluorescent lamps without Plexiglas shielding (510).
Children with the Crigler-Najjar syndrome receiving phototherapy for 2 to 3 years often develop
pigmented lesions and tanning as well as skin atrophy.
The Bronze-Baby Syndrome

When infants with cholestatic jaundice are exposed to phototherapy they may develop a grayishbrown discoloration of the skin, serum, and urine known as the bronze baby syndrome (511). The
pathogenesis of this syndrome is unknown, but it may be related to an accumulation of
porphyrins and other metabolites in the plasma of infants who develop cholestasis (511,512).
Although it occurs exclusively in infants with cholestasis, not all infants with cholestatic jaundice
develop the syndrome. An animal model of the bronze baby syndrome has been produced by
ligating the common bile duct in adult Wistar rats and subjecting them to phototherapy (513).
This syndrome generally has had few deleterious consequences, although a full-term infant with
the bronze baby syndrome who died was shown to have kernicterus at autopsy (474). The
maximal TSB level in this infant was 18 mg/dL (308 mol/L) and the direct-reacting bilirubin was
4.1 mg/dL (70 mol/L). If there is a need for phototherapy, the presence of direct-reacting
hyperbilirubinemia should not be considered a contraindication to its use. Because the products of
phototherapy are excreted in the bile, the presence of cholestasis will decrease the efficacy of
phototherapy. Nevertheless, infants with direct-reacting hyperbilirubinemia often show some
response to phototherapy. In infants receiving phototherapy who develop the bronze baby
syndrome, exchange transfusion should be considered if the TSB is in the intensive phototherapy
range and phototherapy does not promptly lower the TSB. Because of the paucity of data, firm
recommendations cannot be made. Note, however, that the direct-reacting serum bilirubin should
not be subtracted from the TSB concentration in making decisions about exchange transfusions
(76) (see Fig. 35-23).
Eye Damage
Because light can be toxic to the retina, the eyes of infants receiving phototherapy should be
protected with appropriate eye patches (514). Note that displaced eye patches may
P.832
obstruct the nares and produce apnea and that shielding infant's eyes and depriving them of
visual stimuli could be harmful. The pattern of visual evoked potentials in preterm infants whose
eyes had been shielded until they reach 32-weeks postconceptional age where compared with
those in infants with unshielded eyes (515). At 41 and 50 weeks postconception, and at age 3
years, no differences between the groups where found in the visual evoked responses (515).
Commercially available eye shields, if properly applied, prevent more than 98% of light
transmission (516). Note, however, that in addition to the potential risk of irritation or even
corneal abrasion from eye patches, investigators have found an increase in bacterial pathogen
isolation and purulent conjunctivitis in infants whose eyes are patched (517).
Insensible Water Loss and Thermal Regulation

Acute changes in the thermal environment occur directly after the initiation of phototherapy
(518), and infants receiving phototherapy have an increase in peripheral blood flow that may
increase the insensible water loss. Term and preterm infants receiving phototherapy from
fluorescent tubes and studied naked in an incubator with an ambient relative humidly of 50% and
a servocontrolled environment, showed no increase in water loss from the skin (519). However, a
group of Israeli investigators studied 31 preterm infants (gestations 25 to 36 weeks) and
measured transepidermal water loss in 7 body areas before and during phototherapy (under
fluorescent tubes). They found a mean increase of 26.4% in transepidermal water loss during
phototherapy (520). Grnhagen and associates (521) also found an increase in transepidermal
water loss of approximately 20% during halogen spotlight phototherapy despite constant skin
temperature and relative humidity. These authors suggest that it is appropriate to increase
maintenance fluids for preterm infants by 0.35 mL/kg per hour while they are receiving halogen
light phototherapy. On the other hand, this may not be necessary if the infant's fluid status is
monitored regularly by means of urine output, urine specific gravity, serum electrolytes and
change in weight, and fluids adjusted accordingly.
Blood Flow
Under normal circumstances, enteral feedings induce a significant increase in blood flow velocity
in the superior mesenteric artery and conventional phototherapy will blunt this postprandial
mesenteric blood flow in term and preterm infants (493,522). In preterm infants, Pezzati and
associates (493) showed that fiberoptic phototherapy, unlike conventional phototherapy, did not
attenuate the post prandial increase in intestinal blood flow. They also found nonstatistically
significant increases in the passage of loose, watery, stools and abdominal distention in infants
receiving conventional versus fiberoptic phototherapy. They concluded that fiberoptic
phototherapy is preferable to conventional phototherapy for treating hyperbilirubinemia in
preterm infants.
Cell Damage
Phototherapy can produce DNA strand breaks in cell cultures and DNA strand breakage increases
when cells are irradiated in the presence of bilirubin (523). Because light penetrates the thin
scrotal skin and perhaps even reaches the ovaries, it has been suggested that shielding the
gonads with diapers maybe indicated during phototherapy (524). There is no human or animal
evidence, however, to support this practice, and the limited depth of penetration of light makes
the possibility of DNA damage to gonads quite unlikely.
Other Complications
The products of photodecomposition have no direct neurotoxic effects. Although phototherapy can
produce DNA strand breaks in cell cultures, there is no evidence that this occurs in humans. A
relationship has been described between the use of phototherapy and the risk of patent ductus
arteriosus in very-low-birth-weight infants (525, 526). The possible mechanisms for this effect are
not clear, but may be related to a mechanism similar to nitric oxide-induced vasorelaxation (526).
Intravenous Alimentation
Intravenous alimentation solutions should be protected from phototherapy lights. The exposure of
amino acid solutions to light in the blue spectrum produced a significant reduction in tryptophan
(527). In addition, when a multivitamin solution was added to the amino acids, a 40% reduction
in methionine and 22% reduction in histidine occurred (527).
Diarrhea
Infants receiving phototherapy have an increased incidence of diarrhea (528) and stools become
darker and have a greenish tinge (479). These changes are most likely related to the increased
excretion of unconjugated bilirubin into the gut.
Home Phototherapy
The economic and social pressures for early discharge of infants from hospital after delivery have
led to the widespread use of home phototherapy. Because most of the devices commonly used for
home phototherapy do not provide the same degree of irradiance or surface area exposure as
those available in the hospital, home phototherapy, of necessity, is used in the prophylactic rather
than in the therapeutic mode. Nevertheless, when used appropriately, home phototherapy poses
no obvious hazards to the infant and is certainly much cheaper than hospital treatment
(529,530,531,532,533). The use of fiberoptic systems and other compact devices has made it
easier to administer phototherapy at home.
P.833
Figure 35-22 lists the AAP recommendations regarding TSB levels at which home phototherapy is
appropriate. Home phototherapy avoids parent-child separation and there is evidence that
mothers of infants who receive phototherapy at home are less likely to stop breast-feeding during
the period of phototherapy and, if stopped, are more likely to resume breast-feeding than are
mothers whose infants are treated in the hospital (534). When compared with other interventions
used in the home such as apnea monitors, nasal oxygen, and ventilators, phototherapy must
certainly rank among the more benign.
Sunlight Exposure
In their original description of phototherapy, Cremer and associates (535) demonstrated that
exposure of newborns to sunlight would lower the serum bilirubin level. Although sunlight
provides sufficient irradiance in the 425- to 475-nm band to provide phototherapy, the practical
difficulties involved in safely exposing a naked newborn to the sun either inside or outside (and
avoiding sunburn) preclude the use of sunlight as a reliable therapeutic tool and it is therefore not
recommended.
Crigler-Najjar Syndrome
Other than liver transplantation, the only therapy available to children with the Crigler-Najjar
syndrome is phototherapy. This is achieved in most of these children by means of specially
designed (noncommercial) home phototherapy devices that provide adequate phototherapy to the
growing child and even the adolescent. As the child gets older, phototherapy becomes less
effective presumably as a consequence of thickening of the skin, an increase in skin pigmentation,
a decrease in surface area relative to body mass, and the need to provide phototherapy only
during sleep to allow the normal activities of childhood during the day. The most satisfactory
systems provide a tanning bed configuration. The child lies on a transparent surface directly
above special blue fluorescent tubes. Bubble wrap and plastic lilos have been used for this
purpose but, because of their low porosity, produce patient discomfort (361,536,537). Job and
associates have used a standard mesh or high transmission fabric stretched over an adjustable
tension frame (361). This is similar to a traditional hammock and permits adequate transmission
of blue light as well as patient comfort.
Watchko has comprehensively reviewed this subject (538) and Edwards and
associates (539) have reviewed the basic indications for, and contraindications to,
performing exchange transfusions. A few issues are discussed here. The
prevention of Rh hemolytic disease with Rh immunoglobulin and the more
effective use of phototherapy has led to a dramatic decline in the number of
exchange transfusions performed (463) (see Figs. 35-24 and 35-25). It is now
quite possible for a pediatric resident to complete a 3-year training program
without ever having performed an exchange transfusion or even witnessed one.
As fewer and fewer of these procedures are done, it is quite likely that the risks of
complications will increase.
Exchange Transfusion Risks

Table 35-39 lists potential complications of exchange transfusion and the overall
risks of exchange transfusion were reviewed by Ip and associates (139).
Reporting the morbidity and mortality associated with exchange transfusion is
difficult because both are significantly dependent on the clinical state of the infant
before the exchange transfusion. Morbidity and mortality are much lower in term
infants with idiopathic hyperbilirubinemia than in sick preterm infants who might
be critically ill at the time of the exchange transfusion. In addition, definitions of
exchange transfusion-related mortality are not consistent between studies and it
is difficult to determine from the literature whether the exchange transfusion
procedure itself, or other morbidities were responsible for the deaths of these
infants. In addition, many of the published studies refer to infants born before
1970 and there are few U.S. studies of infants born after 1990. Contemporary
information
P.834
on the mortality of exchange transfusions is difficult to obtain, given the
infrequency of the procedure. In full-term and near-term infants who are
relatively well, the risk of death is low (540,541,542). The overall mortality is
approximately 3 in 1,000 procedures (540,541,543). In the NICHD cooperative
phototherapy study (540), morbidity (apnea, bradycardia, cyanosis, vasospasm,
thrombosis) was observed in 22 of 328 (6.7%) of exchange transfusions
performed. Of the 22 adverse events, however, 6 were mild episodes of
bradycardia associated with calcium infusion. Excluding those infants, as well as
two who experienced transient arterial spasms, the incidence of significant
morbidity associated with the procedure was 5%.
TABLE 35-39 POTENTIAL COMPLICATIONS OF EXCHANGE TRANSFUSION
Cardiovascular
Arrhythmias
Cardiac arrest
Volume overload
Embolization with air or clots
Thrombosis
Vasospasm
Hematologic
Sickling (donor blood)

Thrombocytopenia
Bleeding (overheparinization of donor blood)
Graft-versus-host disease
Mechanical of thermal injury to donor cells
Gastrointestinal
Biochemical
Bowel perforation
Hyperkalemia
Hypernatremia
Hypocalcemia
Hypomagnesemia
Infectious
Acidosis
Hypoglycemia
Bacteremia
Virus infection (hepatitis, cytomegalovirus)
Malaria
Miscellaneous
Hypothermia
Perforation of umbilical vein
Drug loss
Apnea
From Watchko JF. Exchange transfusion in the management of neonatal
hyperbilirubinemia. In: Maisels MJ, Watchko JF, eds. Neonatal jaundice.
London, UK: Harwood Academic, 2000:169176, with permission.
Jackson (541) reported a 15-year experience (1980 to 1995) of exchange
transfusion in 106 infants at the Children's Hospital and University of Washington
Medical Center in Seattle. Eighty-one were healthy, and there were no deaths in
these infants, although one child developed severe necrotizing enterocolitis
requiring surgery. There were 25 sick infants, 3 (12%) of whom had serious
complications from the exchange transfusion and 2 (8%) died; three additional
deaths were considered possibly due to the exchange transfusion. Thus, the
total number of deaths in sick infants that were possibly caused by the exchange
was 5 of 25 (20%). Most recently, adverse events associated with exchange
transfusions were reviewed at two perinatal centers in Cleveland, Ohio, between
1992 and 2002 (542). Over a 10.5-year period in two large perinatal centers, only
67 infants were identified who had exchange transfusions for hyperbilirubinemia
an average of about 3 exchange transfusions per year in each institution. The
gestational ages ranged from less than 32 weeks (n=15) to term infants (n=22).
Adverse events occurred in 74% of the exchanges, with thrombocytopenia (44%),
hypocalcemia (29%), and metabolic acidosis (24%) being the most common.
There were only two serious adverse events, both in infants who had other
preexisting, serious neonatal morbidities. The one infant who died was a critically
ill 25-weeks' gestation infant with a birth weight of 731 g. The investigators also
found that exchange transfusions performed using both umbilical venous and
arterial catheters were significantly more likely to be associated with adverse
events than when done through the umbilical vein alone or via other routes (542).
Although the risk is very low, exchange transfusion nevertheless carries the usual
risk of any blood product. The risk estimates (risk per tested unit) for transfusion
transmitted viruses in the United States for the period from 1991 to 1993 were as
follows: human immunodeficiency virus, 1:493,000; human T-cell lymphotropic
virus, 1:641,000; hepatitis C virus, 1:103,000; and hepatitis B virus, 1:63,000
(544).
Bilirubin Dynamics During and After Exchange

Transfusion
During exchange transfusion, bilirubin from the extravascular space is drawn into
the plasma, and partial equilibration between extravascular and plasma bilirubin
occurs almost instantaneously (545). Thus, by the end of a double-volume
exchange, in which only about 15% of the circulating erythrocytes remain, the
serum bilirubin is still 45% to 60% of the preexchange level (545,546,547).
Immediately after the exchange, further equilibration takes plac, which is
completed within 30 minutes and produces the early rebound of plasma bilirubin
to 60% to 80% of the preexchange level (545).
Repeat Exchange Transfusions

In general, the criteria for repeat exchange transfusions are similar to those used
for the initial exchange.
Pharmacologic Treatment
Pharmacologic agents used in the management of hyperbilirubinemia can
accelerate the normal metabolic pathways for bilirubin clearance, inhibit the
enterohepatic circulation of bilirubin, and interfere with bilirubin formation by
either blocking the degradation of heme or inhibiting hemolysis.
Acceleration of Normal Metabolic Pathways for

Bilirubin Clearance
Phenobarbital
Phenobarbital is a potent inducer of microsomal enzymes that increases bilirubin
conjugation and excretion and increases bile flow. When given in sufficient doses
to the mother, the infant, or both, phenobarbital is effective in lowering serum
bilirubin levels in the first week of life (226,548). However, concerns about longterm toxicity when given to pregnant women militate against its use for this
purpose (549,550).
Decreasing Bilirubin Production by Inhibiting

Heme Oxygenase
Tin Mesoporphyrin
As illustrated in Fig. 35-1, the enzyme microsomal heme oxygenase is necessary
for the conversion of heme to biliverdin, one of the first steps in the formation of
bilirubin from hemoglobin. Certain synthetic metalloporphyrins are powerful
competitive inhibitors of heme oxygenase and suppress the formation of bilirubin.
The inhibition of heme degradation to bilirubin does not result in the accumulation
of heme; the heme is excreted unchanged in the bile in quantities that
compensate for the decreased excretion of bilirubin (226).
In a series of controlled clinical trials in Greece and Argentina, Kappas and
colleagues demonstrated that tin mesoporphyrin (SnMP) is a potent inhibitor of
heme oxygenase and is highly effective in reducing TSB levels and the
requirements for phototherapy in term and preterm
P.835
neonates (370,551,552). They also showed that SnMP in a single dose of 6 mol/
kg was more effective than special blue-light phototherapy in the treatment of

term and near-term neonates with established hyperbilirubinemia (551,552).
SnMP was equally effective in controlling hyperbilirubinemia in infants with G6PD
deficiency (553); a U.S. trial of term and near-term infants showed similar results
(554). The only side effect seen so far has been a transient, non-dose-dependent
erythema that disappeared without sequelae in infants who received (white light)
phototherapy after SnMP administration (555). SnMP has been used in the
treatment of the Crigler-Najjar syndrome and has achieved a temporary reduction
in TSB levels (369,556), and SnMP has prevented the need for exchange
transfusion in Jehovah's Witness newborns with Rh hemolytic disease (557).
Kappas reviewed the published controlled clinical trials performed with SnMP
(370). Of 279 infants in the control groups, 129 (46%) received phototherapy
versus 13 of 443 (3%) of the infants who received SnMP.
The idea of using an inhibitor of bilirubin production is very attractive. It requires
no apparatus, infants do not need to be blindfolded, no hospitalization is required,
and there is no separation of mother from infant. Phototherapy is a relatively
cumbersome and slow method of lowering serum bilirubin levels, whereas SnMP
prevents the production of bilirubin and is more efficient than phototherapy. To
date, more than 500 newborns have received SnMP in controlled trials, but the
drug is still awaiting FDA approval, although it can be obtained for compassionate
use (Wellspring Pharmaceu-tical Corp, Neptune, NJ). If approved, SnMP could find
immediate application in preventing the need for exchange transfusion in infants
who are not responding to phototherapy (557).
Inhibiting the Enterohepatic Circulation of

Bilirubin
A number of agents have been administered to infants in an attempt to interrupt
the process of the enterohepatic absorption of bilirubin. These include charcoal,
agar, and cholestyramine, but they have failed to produce clinically significant
reductions in TSB (226). There is evidence that bilirubin in the neonatal gut may
be bound to unabsorbed fat (262), and that impairing fat absorption could
enhance fecal excretion of unconjugated bilirubin. The administration of orlistat, a
substance that binds fat in the intestine, significantly reduced bilirubin levels in
Gunn rats (31). To date, there have been no clinical trials of this intriguing
concept in the human newborn.
Decreasing Bilirubin Production by Inhibiting

Hemolysis
Intravenous Immunoglobulin
Controlled trials have confirmed that the administration of IVIg to infants with Rh
hemolytic disease will significantly reduce the need for exchange transfusion
(469,558,559). It also is likely that IVIg will help to mitigate the course of severe
ABO hemolytic disease (560) and other isoimmune causes of hemolysis. The
doses used have ranged from 500 mg/kg given over 2 hours soon after birth to
800 mg/kg given daily for 3 days. In Rh hemolytic disease, anti-D-coated
erythrocytes are removed from the circulation through antibody-dependent lysis
by cells of the reticuloendothelial system. The mechanism of action of IVIg is
unknown, but it is possible that it might alter the course of Rh hemolytic disease
by blocking Fc receptors, thereby inhibiting hemolysis. The risks of IVIg therapy
are certainly lower than those of exchange transfusion.
Physiologic Role of Bilirubin

Despite its potential toxicity, bilirubin may have an important and positive
physiologic role (278,561). Bilirubin is a powerful antioxidant in vitro
(278) and there is a positive relationship between serum bilirubin levels
and antioxidant activity in term and preterm infants (562,563,564).
Bilirubin may have a physiologic role as an antioxidant in the human
neonate (279,564,565). In sick neonates who have circulatory failure,
asphyxia or sepsis, the rate of rise of TSB is less than in control infants
(279,565) suggesting that bilirubin is consumed to cope with oxidative
stress (561). Because of the possibility that oxidative injury might play a
role in the development of retinopathy of prematurity, investigators have
evaluated the relationship between bilirubin levels and retinopathy of
prematurity (181,181,566,567,568,569 and 570). Of eight published
studies, however, six show no protective effect of an elevated TSB level
(567,568,569,570,571,572). In adults there is clear evidence that
decreased TSB levels are associated with an increased risk of coronary
artery disease and peripheral vascular disease (561), but we have yet to
identify newborns who have suffered a bad outcome because they have
too little bilirubin in their serum.
Acknowledgment
I thank Anthony McDonagh, PhD, for his critical review of portions of this
chapter.
REFERENCES
1. Hansen TW. Pioneers in the scientific study of neonatal jaundice and
kernicterus. Pediatrics 2000;106:E15.
2. Weiss JS, Gautam A, Lauff JJ, et al. The clinical importance of protein-bod
fraction of serum bilirubin in patients with hyperbilirubinemia. N Engl J Med
1983;309:147-150.
3. Tiribelli C, Ostrow JD. New concepts in bilirubin and jaundice: report of the
third international bilirubin workshop, April 6-8, 1995. Trieste, Italy. Hepatology
1996;24:1296-1311.
4. Berk PD, Noyer C. Structure, formation, and sources of bilirubin and its
transport in plasma. Semin Liver Dis 1994;14:325-330.
5. Berk PD, Noyer C. Hepatic uptake, binding, conjugation, and excretion of

biliriubin. Semin Liver Dis 1994;14:331-343.
6. Berk PD, Noyer C. Clinical chemistry and physiology of bilirubin. Semin Liver
Dis 1994;14:346-351.
P.836
7. Hansen TWR. Fetal and neonatal bilirubin metabolism. In: Maisels MJ,
Watchko JF, eds. Neonatal jaundice. London: Harwood Academic, 2000:3-22.
8. Lightner DA, McDonagh AF. Molecular mechanisms of phototherapy for

neonatal jaundice. Acc Chem Res 1984;17:417-424.
9. Dennery PA, Rodgers P. Ontogeny and developmental regulation of heme

oxygenase. J Perinatol 1996;16:S79-S83.
10. McDonagh AF, Lightner DA. Like a shrivelled blood orangebilirubin,

jaundice and phototherapy. Pediatrics 1985;75: 443-455.
11. Ives NK, Gardner RM. Blood-brain barrier permeability to bilirubin in the rat:
studies using intracarotid bolus injection and in situ brain perfusion techniques.
12. Onishi S, Kawade N, Itoh S, et al. Postnatal development of uridine

diphosphate glucuronyl transferase activity towards bilirubin and o-aminophenol
in human liver. Biochem J 1979;194: 705-707.
13. Kawade N, Onishi S. The prenatal and postnatal development of UDPglucuronyl transferase activity toward bilirubin and the effect of premature birth
on this activity in the human liver. Biochem J 1981;196:257-260.
14. Dubey AP, Garg A, Bhatia BD. Fetal exposure to maternal

hyperbilirubinemia. Indian Pediatr 1983;20:527-528.
15. Smith JF, Baker JM. Crigler-Najjar disease in pregnancy. Obstet Gynecol
1994;84:670-672.
16. Taylor WG, Walkinshaw SA, Farquharson RG, et al. Pregnancy in CriglerNajjar syndrome. Case report. Br J Obstet Gynaecol 1991; 98:1290-1291.
17. Brodersen R. Binding of bilirubin to albumin. Crit Rev Clin Lab Sci
1980;11:305-399.
18. Ostrow JD, Pascolo L, Tiribelli C. Mechanisms of bilirubin neurotoxicity.

Hepatology 2002;35:1277-1280.
19. Rosenthal P, Blanckaert N, Cabra PM, et al. Formation of bilirubin

conjugates in human newborns. Pediatr Res 1986;20:947-950.
20. Bosma PJ, Seppen J, Goldhoorn B, et al. Bilirubin UDPglucuronosyltransferase 1 is the only relevant bilirubin glucuronodating isoform
in man. J Biol Chem 1994;269:17960-17964.
21. van Es HH, Bout A, Liu J, et al. Assignment of the human UDPglucuronosyltransferase gene (UGT1A1) to chromosome region 2q37. Cytogenet
Genome Res 1993;63:114-116.
22. Kaplan M, Hammerman C, Maisels MJ. Bilirubin genetics for the

nongeneticist: hereditary defects of neonatal bilirubin conjugation. Pediatrics
2003;111:886-893.
23. Watchko JF, Daood MJ, Biniwale M. Understanding neonatal

hyperbilirubinaemia in the era of genomics. Semin Neonatol 2002;7:143-152.
24. Kotal P, Van der Veere CN, Sinaasappel M, et al. Intestinal excretion of
unconjugated bilirubin in man and rats with inherited unconjugated
hyperbilirubinemia. Pediatr Res 1997;42:195-200.
25. Kaplan M, Muraca M, Hammerman C, et al. Imbalance between production

and conjugation of bilirubin: a fundamental concept in the mechanism of
neonatal jaundice. Pediatrics 2002;4:110.
26. Bartoletti AL, Stevenson DK, Ostrander CR, et al. Pulmonary excretion of
carbon monoxide in the human infant as an index of bilirubin production. I.
Effects of gestational age and postnatal age and some common neonatal
abnormalities. J Pediatr 1979;94:952-955.
27. Maisels MJ, Pathak A, Nelson NM, et al. Endogenous production of carbon
monoxide in normal and erythroblastotic newborn infants. J Clin Invest
1971;50:1-9.
28. Gourley GR. Pathophysiology of breast-milk jaundice. In: Polin RA, Fox WW,
eds. Fetal and neonatal physiology. Philadelphia: WB Saunders, 1998:1499.
29. Poland RL, Odell GB. Physiologic jaundice: the enterohepatic circulation of
bilirubin. N Eng J Med 1971;284:1-6.
30. Gartner LM, Lee K-S, Vaisman S, et al. Development of bilirubin transport
and metabolism in the newborn rhesus monkey. J Pediatr 1977;90:513.
31. Nishioka T, Hafkamp AM, Havinga R, et al. Orlistat treatment increases fecal
bilirubin excretion and decreases plasma bilirubin concentrations in
hyperbilirubinemic gunn rats. J Pediatr 2003;143:327-334.
32. Fevery J. Fasting hyperbilirubinemia: unraveling the mechanism involved.
Gastroenterology 1997;113:1798-1800.
33. Grtner U, Goeser T, Wolkoff AW. Effect of fasting on the uptake of bilirubin
and sulfobromophthalein by the isolated perfused rat liver. Gastroenterology
1997;113:1707-1713.
34. Wolkoff AW, Goresky CA, Sellin J, et al. Role of ligandin in transfer of
bilirubin from plasma into liver. Am J Physiol 1979; 236:E638.
35. Sackett DL, Haynes RB, Guyatt GH, et al. Clinical epidemiology: a basic
science for clinical medicine. 2nd ed. Boston: Little Brown, 1991.
36. Orth J. Ueber das vorkommen von bilirubinkrystallen bei neugebornen

kindern. Virchows Arch 1875;63:447-462.
37. Schmorl G. Zur kenntnis des ikterus neonatorum, insbesondere der dabei
auftretenden gehirnveranderungen. Verh Dtsch Ges Pathol 1904;15:109-115.
38. Turkel SB. Autopsy findings associated with neonatal hyperbilirubinemia.

Clin Perinatol 1990;17:381.
39. Ahdab-Barmada M, Moossy J. The neuropathology of kernicterus in the

premature neonate: diagnostic problems. J Neuropathol Exp Neurol 1984;43:4556.
40. Volpe JJ. Neurology of the newborn. 4th ed. Philadelphia: WB Saunders,
2001.
41. Ahdab-Barmada M. The neuropathology of kernicterus: definitions and

debate. In: Maisels MJ, Watchko JF, eds. Neonatal jaundice. London: Harwood
Academic, 2000:75-88.
42. Valdes-Dapena MA, Nissim JE, Arey JB, et al. Yellow pulmonary hyaline
membranes. J Pediatr 1976;89:128.
43. Perlman JM, Rogers B, Burns D. Kernicterus findings at autopsy in 2 sick

near-term infants. Pediatrics 1997;99:612-615.
44. Cashore WJ, Oh W. Unbound bilirubin and kernicterus in low birthweight

45. Gartner LM, Snyder RN, Chabon RS, et al. Kernicterus: high incidence in
premature infants with low serum bilirubin concentration. Pediatrics
1970;45:906-917.
46. Turkel SB, Miller CA, Guttenberg ME, et al. A clinical pathologic reappraisal
of kernicterus. Pediatrics 1982;69:267-272.
47. Lipsitz PJ, Gartner LM, Bryla DA. Neonatal and infant mortality in relation to
phototherapy. Pediatrics 1985;75[Suppl]: 422-426.
48. Ahlfors CE. Bilirubin-albumin binding and free bilirubin. J Perinatol 2001;21:
S40-S42.
49. Hansen TWR. The pathophysiology of bilirubin toxicity. In: Maisels MJ,
Watchko JF, eds. Neonatal jaundice. London: Harwood Academic, 2000:89-104.
50. Hansen TWR. Bilirubin brain toxicity. J Perinatol 2001;21: S48-S51.
51. Volpe JJ. Bilirubin and brain injury. In: Volpe JJ, ed. Neurology of the
newborn. Philadelphia: WB Saunders, 2001:521-546.
52. Claireaux A. Icterus of the brain in the newborn. Lancet 1953;2:1226.
53. Brodersen R, Robertson A. Chemistry of bilirubin and its interaction with

albumin. In: Levine RL, Maisels MJ, eds. Hyperbilirubinemia in the newborn:
report of the 85th Ross Conference on Pediatric Research. Columbus, OH: Ross
Laboratories, 1983: 91-101.
54. Brodersen R, Stern L. Deposition of bilirubin acid in the central nervous

system: a hypothesis for the development of kernicterus. Acta Paediatr Scand
1990;79:12.
55. Wennberg RP. Cellular basis of bilirubin toxicity. N Y State J Med

1991;91:493-496.
56. Chuniaud L, Dessante M, Chantoux F, et al. Cytotoxicity of bilirubin for

human fibroblasts and rat astrocytes in culture: effect of the ratio of bilirubin to
serum albumin. Clin Chim Acta 1996;256:103-114.
57. Roger C, Koziel V, Vert P, et al. Regional cerebral metabolic consequences

of bilirubin in rat depend upon post-gestational age at the time of
hyperbilirubinemia. Brain Res Dev Brain Res 1995;87: 194-202.
58. Rodrigues C, Sola S, Brites D. Bilirubin induces apoptosis via the

mitochondrial pathway in developing rat brain neurons. Hepatology
2002;35:1186-1195.
59. Hoffman DJ, Zanelli SA, Kubin J, et al. The in vivo effect of bilirubin on the
N-methyl-D-aspartate-receptor/ion channel complex in the brains of newborn
piglets. Pediatr Res 1996;40: 804-808.
60. Bratlid D. How bilirubin gets into the brain. Clin Perinatol 1990;17:449-465.
61. Wennberg RP, Ahlfors CE, Rasmussen LF. The pathochemistry of

kernicterus. Early Hum Dev 1979;31:353.
62. Wennberg RP, Hance AJ. Experimental encephalopathy: importance of total

bilirubin, protein binding and blood brain barrier. Pediatr Res 1986;20:789.
P.837
63. Silverman WA. A difference in mortality rate and incidence of kernicterus
among premature infants allotted to two prophylactic antibacterial regimens.
Pediatr 1998;102[Suppl]:225-227.
64. Harris RC, Lucey JF, MacLean JR. Kernicterus in premature infants
associated with low concentrations of bilirubin in the plasma. Pediatrics
1958;21:875-883.
65. Ostrow JD, Pascolo L, Tiribelli C. Reassessment of the unbound

concentrations of unconjugated bilirubin in relation to neurotoxicity in vitro.
Pediatr Res 2003;54:98-104.
66. Ostrow JD, Tiribelli C. New concepts in bilirubin neurotoxicity and the need
for studies at clinically relevant bilirubin concentrations. J Hepatology
2001;34:467-470.
67. Ostrow J, Pascolo L, Shapiro S, et al. New concepts in bilirubin

encephalopathy. Eur J Clin Invest 2003;33:988-997.
68. Ahlfors CE. Measurement of plasma unbound unconjugated bilirubin. Anal

Biochem 2000;279:130-135.
69. Hansen TWR, Oyasaeter S, Stiris T, et al. Effects of sulfisoxazole,

hypercarbia, and hyperosmolality on entry of bilirubin and albumin into brain
regions in young rats. Biol Neonate 1989;56:22.
70. Amin SB, Ahlfors CE, Orlando MS, et al. Bilirubin and serial auditory
brainstem responses in premature infants. Pediatrics 2001;107:664-670.
71. Funato M, Tamai H, Shimada S, et al. Vigintiphobia, unbound bilirubin, and

auditory brainstem responses. Pediatrics 1994;93: 50-53.
72. Ahlfors CE, Herbsman O. Unbound bilirubin in a term newborn with

kernicterus. Pediatrics 2003;111:1110-1112.
73. Nakamura H, Takada S, Shimabuku R, et al. Auditory and brainstem

responses in newborn infants with hyperbilirubinemia. Pediatrics 1985;75:703708.
74. Shimabuku R, Nakamura H. Total and unbound bilirubin determination using

a automated peroxidase micromethod. Kobe J Med Sci 1982;28:91-104.
75. Ahlfors CE. Criteria for exchange transfusion in jaundiced newborns.

Pediatrics 1994;93:488-494.
76. Maisels MJ, Baltz RD, Bhutani V, et al. Management of hyperbilirubinemia in

the newborn infant 35 or more weeks of gestation. Pediatrics 2004;114:297316.
77. Cashore WJ. Free bilirubin concentrations and bilirubin-binding affinity in

term and preterm infants. J Pediatr 1980;96: 521-527.
78. Spear ML, Stahl GE, Paul MH, et al. The effect of 15-hour fat infusions of
varying dosage on bilirubin binding to albumin. JPEN J Parenter Enteral Nutr
1985;9:144-147.
79. Nizar L, Vyhmeister N, Ross R, et al. A jaundiced look at intravenous fat

administration and the risk factor for kernicterus. Clin Res 1990;38:197A.
80. Levine RL. Fluorescence quenching studies of the binding of bilirubin to

albumin. Clin Chem 1972;23:2292.
81. Nelson P, Jacobsen J, Wennberg RP. Effect of pH on the interaction of

bilirubin with albumin and tissue culture cells. Pediatr Res 1974;8:963.
82. Kozuki K, Oh W, Widness J, et al. Increase in bilirubin binding to albumin

with correction of neonatal acidosis. Acta Paediatr Scand 1979;68:213.
83. Robertson A, Carp W, Broderson R. Bilirubin displacing effect of drugs used

in neonatology. Acta Paediatr Scand 1991;80: 1119-1127.
84. Brodersen R, Ebbesen F. Bilirubin-displacing effect of ampicillin,

indomethacin, chlorpromazine, gentamicin, and parabens in vitro and in
newborn infants. J Pharm Sci 1983;72: 248-253.
85. Rasmussen LF, Ahlfors CE, Wennberg RP. Displacement of bilirubin from
albumin by indomethacin. J Clin Pharmacol 1978;18: 477-481.
86. Cooper-Peel C, Brodersen R, Robertson A. Does ibuprofen affect bilirubinalbumin binding in newborn infant serum? Pharmacol Technol 1996;79:297-299.
87. Aranda JV, Varvarigou A, Beharry K, et al. Pharmacokinetics and protein

binding of intravenous ibuprofen in the premature newborn infant. Acta Paediatr
1997;86:289-298.
88. Overmeire BV, Touw D, Schepens PJC, et al. Ibuprofen pharmacokinetics in
preterm infants with patent ductus arteriosus. Clin Pharmacol Ther

2001;70:336-343.
89. Jardine DS, Rogers K. Relationship of benzyl alcohol to kernicterus,

intraventricular hemorrhage, and mortality in preterm infants. Pediatrics
1989;83:153-160.
90. Watchko JF. The clinical sequelae of hyperbilirubinemia. In: Maisels MJ,
Watchko JF, eds. Neonatal jaundice. London: Harwood Academic, 2000:115138.
91. Ahlfors CE. Benzyl alcohol, kernicterus, and unbound bilirubin. J Pediatr
2001;139:317-319.
92. Robertson A, Carp W, Broderson R. Effect of drug combinations on bilirubinalbumin binding. Dev Pharmacol Ther 1991;17:95.
93. Hanko E, Tommarello S, Watchko JF, et al. Administration of drugs known

to inhibit p-glycoprotein increases brain bilirubin and alters the regional
distribution of bilirubin in rat brain. Pediatr Res 2003;54:441-445.
94. Valaes T, Hyte M. Effect of exchange transfusion on bilirubin binding.

95. Cifuentes RF, Nelson AJ, Levine J, et al. Cutaneous bilirubinometry during
phototherapy. Pediatr Res 1982;16:282A(abst).
96. Cashore WJ, Oh W, Brodersen R. Reserve albumin and bilirubin toxicity

index in infant serum. Acta Paediatr Scand 1983; 72:415-419.
97. Ritter DA, Kenny JD, Norton HJ, et al. A prospective study of free bilirubin
and other high-risk factors in the development of kernicterus in premature
98. Robertson A, Sharp C, Karp W. The relationship of gestational age to

reserve albumin concentration for binding of bilirubin.J Perinatol 1988;8:17.
99. Jirka JH, Duckrow B, Kendig JW, et al. Effect of bilirubin on brainstem
auditory evoked potentials in the asphyxiated rat. Pediatr Res 1985;19:556-560.
100. Rozdilsky B, Olszewski J. Experimental study of the toxicity of bilirubin in

newborn animals. J Neuropathol Exp Neurol 1961;20:193-205.
101. Bratlid D, Cashore WJ, Oh W. Effect of serum hyperosmolality on opening

of blood brain barrier for bilirubin in rat brain. Pediatrics 1983;71:909-912.
102. Bratlid D, Jori G. Mechanism of bilirubin entry into the brain in an animal
model. In: Rubaltelli FF, ed. Neonatal jaundice: new trends in phototherapy.
New York: Plenum Press, 1984:23-24.
103. Bratlid D, Cashore WJ, Oh W. Effects of acidosis on bilirubin deposition in

rat brain. Pediatrics 1984;73:431-434.
104. Burgess GH, Oh W, Bratlid D, et al. The effects of brain blood flow on brain
bilirubin deposition in newborn piglets. Pediatr Res 1985;19:691-696.
105. Drummond GS, Kappas A. Chemoprevention of neonatal jaundice: potency

of tin- protoporphyrin in an animal model. Science 1982;217:1250-1252.
106. Levine RL, Fredericks WR, Rapoport SI. Entry of bilirubin into the brain due
to opening of the blood brain barrier. Pediatrics 1982;69:255.
107. Stobie PE, Hansen CT, Hailey JR, et al. A difference in mortality between
two strains of jaundiced rats. Pediatrics 1991;87:5918.
108. Wennberg RP. The blood-brain barrier and bilirubin encephalopathy. Cell
Mol Neurobiol 2000;20(1):97-109.
109. Watchko JF, Daood MJ, Mahmood B, et al. P-glycoprotein and bilirubin
disposition. J Perinatol 2001;21:S43-S47.
110. Watchko JF, Daood MJ, Hansen TWR. Brain bilirubin content is increased in
P-Glycoprotein-Deficient transgenic null mutant mice. Ped Res 1998;44:763766.
111. Cashore WJ. Bilirubin metabolism and toxicity in the newborn. In: Polin
RA, Fox WW, eds. Fetal and neonatal physiology. Philadelphia: WB Saunders,
1998:1493-1498.
112. Saunders NR, Mollgard K. Development of the blood-brain barrier. J Dev

Physiol 1984;6:45.
113. Lee C, Oh W. Permeability of the blood-brain barrier for 125I-albuminbound bilirubin in newborn piglets. Pediatr Res 1989;25:452.
114. Ohsugi M, Sato H, Yamamura H. Transfer of bilirubin covalently bound to

125I-albumin
from blood to brain in the Gunn rate newborn. Biol Neonate
1992;62:47-54.
115. Escher-Graub DC, Ricker HS. Jaundice and behavioral organization in the
full-term neontae. Helv Paediatr Acta 1986;41:425-435.
116. Rapisardi G, Vohr B, Cashore W, et al. Assessment of infant cry variability

in high-risk infants. Int J Ped Otorhinolaryngology 1989;17:19-29.
117. Perlman M, Fainmesser P, Sohmer H, et al. Auditory nerve-brainstem

evoked responses in hyperbilirubinemic neonates. Pediatrics 1983;72:658-664.
118. Vohr BR. New approaches to assessing the risks of hyperbilirubinemia. Clin
Perinatol 1990;17:293-306.
P.838
119. Hansen TWR, Allen JW. Bilirubin-oxidizing activity in rat brain. Biol
Neonate 1996;70:289-295.
120. Hansen TWR, Allen JW. Oxidation of bilirubin by brain mitochondrial

membranes-dependence on cell type and postnatal age. Biochem Mol Med
1997;60:155-160.
121. Johnson L, Brown AK, Bhutani VK. Binda clinical score for bilirubin
induced neurologic dysfunction in newborns. Pediatrics 1999;104:746-747.
122. Gerrard J. Kernicterus. Brain 1952;75:526-570.
123. Van Praagh R. Diagnosis of kernicterus in the neonatal period. Pediatrics

1961;28:870-876.
124. Johnson LH, Bhutani VK, Brown AK. System-based approach to

management of neonatal jaundice and prevention of kernicterus. J Pediatr
2002;140:396-403.
125. Harris M, Bernbaum J, Polin J, et al. Developmental follow-up of breastfed

term and near-term infants with marked hyperbilirubinemia. Pediatrics
2001;107:1075-1080.
126. Johnston WH, Angara V, Baumal R, et al. Erythroblastosis fetalis and

hyperbilirubinemia. A five-year follow-up with neurological, physiological and
audiological evaluation. Pediatrics 1967;39:88-92.
127. Jones MH, Sands R, Hyman CB, et al. Longitudinal study of incidence of
central nervous system damage following erythroblastosis fetalis. Pediatrics
1954;14:346.
128. Byers RK, Paine RS, Crothers V. Extrapyramidal cerebral palsy with
hearing loss following erythroblastosis. Pediatrics 1955; 15:248.
129. Perlstein MA. The late clinical syndrome of post-icteric encephalopathy.

Pediatr Clin North Am 1960;7:665.
130. Chin KC, Taylor MJ, Perlman M. Improvement in auditory and visually
evoked potentials in jaundiced preterm infants after exchange transfusion. Arch
Dis Child 1985;60:714-717.
131. Hyman CB, Keaster J, Hanson V, et al. CNS abnormalities after neonatal
hemolytic disease or hyperbilirubinemia. A prospective study of 405 patients.
Am J Dis Child 1969;117:395-405.
132. Shapiro SM, Nakamura H. Bilirubin and the auditory system. J Perinatol
2001;21:S52-S55.
133. Penn AA, Enzman DR, Hahn JS, et al. Kernicterus in a full-term infant.
Pediatrics 1994;93:1003-1006.
134. Martich-Kriss V, Kollias SS, Ball WS. MR findings in kernicterus. AJNR Am J

Neuroradiol 1995;16:819-821.
135. Grobler JM, Mercer MJ. Kernicterus associated with elevated predominantly
direct-reacting bilirubin. S Afr Med J 1997;87:146.
136. Sugama S, Soeda A, Eto Y. Magnetic resonance imaging in three children

with kernicterus. Pediatr Neurol 2001;25:328-331.
137. Govaert P, Lequin M, Swarte R, et al. Changes in globus pallidus with (pre)
term kernicterus. Pediatrics 2003;112:1256-1263.
138. Hoon AH, Reinhardt EM, Kelley RI, et al. Brain magnetic resonance
imaging in suspected extrapyramidal cerebral palsy: observations in
distinguishing genetic-metabolic from acquired causes. J Pediatr 1997;131:240245.
139. Ip S, Chung M, Kulig J, et al. An evidence-based review of important

issues concerning neonatal hyperbilirubinemia. Pediatrics 2004;114:e130-e153.
140. Ip S, Glicken S, Kulig J, et al. Management of neonatal hyperbilirubinemia.

Rockville, MD: US Department of Health and Human Services, Agency for
Healthcare Research and Quality, 2003.
141. Maisels MJ, Newman TB. Bilirubin and neurological dysfunction do we

need to change what we are doing? Pediatr Res 2001;50(6): 677-678.
142. Hsia DYY, Allen FH, Gellis SS, et al. Erythroblastosis fetalis. VIII. Studies of
serum bilirubin in relation to kernicterus. N Engl J Med 1952;247:668-671.
143. Mollison PL, Cutbush M. Haemolytic disease of the newborn. In: Gairdner
D, ed. Recent Advances in Pediatrics. New York: P Blakiston, 1954:110.
144. Newman TB, Maisels MJ. Does hyperbilirubinemia damage the brain of
healthy full-term infants? Clin Perinatol 1990;17:331-358.
145. Newman TB, Maisels MJ. Evaluation of jaundice in the term newborn: a
kinder, gentler approach. Pediatrics 1992;89:809-818.
146. American Academy of Pediatrics. Provisional Committee for Quality

Improvement and Subcommittee on Hyperbilirubinemia. Practice parameter:
management of hyperbilirubinemia in the healthy term newborn. Pediatrics
1994;94:558-562.
147. Gartner LM. Management of jaundice in the well baby. Pediatrics

1992;89:826-827.
148. Merenstein GB. New bilirubin recommendations questioned. Pediatrics

1992;89:822-823.
149. Watchko JF, Oski FA. Bilirubin 20 mg/dL =vigintiphobia. Pediatrics

1983;71:660-663.
150. Watchko J, Claassen D. Kernicterus in premature infants: current

prevalence and relationship to NICHD phototherapy study exchange criteria.
Pediatrics 1994;93:996-999.
151. Watchko JF, Maisels MJ. Jaundice in low birth weight infantspathobiology
and outcome. Arch Dis Child Fetal Neonatol Ed 2003;88;456-459.
152. Watchko JF, Oski FA. Kernicterus in preterm newborns: past, present and
future. Pediatrics 1992;90:707-715.
153. Mollison PL, Walker W. Controlled trials of the treatment of haemolytic

disease of the newborn. Lancet 1952;1:429-433.
154. Ozmert E, Erdem G, Topcu M. Long-term follow-up of indirect

hyperbilirubinemia in full-term Turkish infants. Acta Paediatr 1996;85:14401444.
155. Johnson L, Boggs TR. Bilirubin-dependent brain damage: incidence and

indications for treatment. In: Odell GB, Schaffer R, Simopoulos AP, eds.
Phototherapy in the newborn: an overview. Washington, DC: National Academy
of Sciences, 1974: 122-149.
156. Valaes T. Severe neonatal jaundice associated with glucose-6-phosphate

dehydrogenase deficiency: pathogenesis and global epidemiology. Acta Paediatr
1994;394[Suppl]:58-76.
157. Martin DH, Thompson W, Pinkerton JHM, et al. A randomized controlled

trial of selective planned delivery. Br J Obstet Gynecol 1978;85:109-113.
158. Raghubir KV, Fox GF, Inwood S, et al. Follow up of term neonates with
extremely high unconjugated bilirubin. Pediatr Res 1996; 39:276A.
159. Newman TB, Liljestrand P, Gabriel J, et al. Infants with bilirubin levels of
30 mg/dL or more in a large managed care organization. Pediatrics
2003;6:1303-1310.
160. Newman TB, Klebanoff MA. Neonatal hyperbilirubinemia and long-term

outcome: another look at the collaborative perinatal project. Pediatrics
1993;92:651-657.
161. Seidman DS, Paz I, Stevenson DK, et al. Neonatal hyperbilirubinemia and
physical and cognitive performance at 17 years of age. Pediatrics 1991;88:828833.
162. Seidman DS, Laor A, Paz I, et al. Neonatal hyperbilirubinemia in healthy

term infants and cognitive outcome in late adolescence. Pediatr Res
1998;43:194A.
163. Grimmer I, Berger-Jones K, Buhrer C, et al. Late neurological sequelae of

non-hemolytic hyperbilirubinemia of healthy term neonates. Acta Paediatr
1999;88:661-663.
164. Soorani-Lunsing I, Woltil H, Hadders-Algra M. Are moderate degrees of

hyperbilirubinemia in healthy term neonates really safe for the brain? Pediatr
Res 2001;50:701-705.
165. Naeye RL. Amniotic fluid infections, neonatal hyperbilirubinemia, and

psychomotor impairment. Pediatrics 1978;62:497- 503.
166. Odell GB, Storey GNB, Rosenberg LA. Studies in kernicterus: III. The
saturation of serum proteins with bilirubin during neonatal life and its
relationship to brain damage at five years. J Pediatr 1970;76:12-21.
167. Scheidt PC, Graubard BI, Nelson KB, et al. Intelligence at six years in
relation to neonatal bilirubin level: follow-up of the National Institute of Child
Health and Human Development Clinical Trial of Phototherapy. Pediatrics
1991;87:797-805.
168. Nwaesei CG, Van Aerde J, Boyden M, et al. Changes in auditory brainstem
responses in hyperbilirubinemic infants before and after exchange transfusion.
Pediatrics 1984;74:800-803.
169. Nakamura H, Yonetani M, Uetani Y, et al. Determination of serum unbound

bilirubin for prediction of kernicterus in low birth weight infants. Acta Paediatr
Jpn 1992;54:642-647.
170. Scheidt PC, Bryla DA, Nelson KB, et al. Phototherapy for neonatal
hyperbilirubinemia: six year follow-up of the NICHD clinical trial. Pediatrics
1990;85:455-463.
171. Nilsen ST, Finne PH, Bergsjo P, et al. Males with neonatal
hyperbilirubinemia examined at 18 years of age. Acta Paediatr Scand
1984;73:176-180.
172. Shapiro SM. Evoked potentials and bilirubin. In: Maisels MJ, Watchko JF,
eds. Neonatal jaundice. London: Harwood Academic, 2000:105-114.
173. Suresh G, Lucey J. Lack of deafness in Crigler-Najjar syndrome type 1: a

patient survey. Pediatrics 2000;100:E9.
174. Johnson L. Hyperbilirubinemia in the term infant: when to worry, when to

treat. N Y State J Med 1991;91:483-489.
P.839
175. Department of Clinical Epidemiology and Biostatistics MUHSC. How to read

journals. IV. To determine etiology or causation. Can Med Assoc J
1981;124:985.
176. Scheidt PC, Mellits ED, Hardy JB, et al. Toxicity to bilirubin in neonates:
infant development during first year in relation to maximum neonatal serum
bilirubin concentration. J Pediatr 1977;91:292-297.
177. Brown AK, Kim MH, Wu PYK, et al. Efficacy of phototherapy in prevention
and management of neonatal hyperbilirubinemia. Pediatrics 1985;75:393-400.
178. Kim MH, Yoon JJ, Sher J, et al. Lack of predictive indices in kernicterus. A
comparison of clinical and pathologic factors in infants with or without
kernicterus. Pediatrics 1980;66:852- 858.
179. Turkel SB, Guttenberg ME, Moynes DR, et al. Lack of identifiable risk
factors for kernicterus. Pediatrics 1980;66:502-506.
180. van de Bor M, Ens-Dokkum M, Schreuder AM, et al. Hyperbilirubinemia in

low birthweight infants and outcome at 5 years of age. Pediatrics 1992;89:359364.
181. Yeo KL, Perlman M, Hao Y, et al. Outcomes of extremely premature infants
related to their peak serum bilirubin concentrations and exposure to
phototherapy. Pediatr 1998;102(6):1426-1431.
182. O'Shea TM, Dillard RG, Klinepeter KD, et al. Serum bilirubin levels,
intracranial hemorrhage, and the risk of developmental problems in very low
birth weight infants. Pediatrics 1992; 90:888-892.
183. Graziani LJ, Mitchell DG, Kornhauser M, et al. Neurodevelopment of

preterm infants: neonatal neurosonographic and serum bilirubin studies.
184. van de Bor M, van Zeben-van der Aa TM, Verloove-Vanhorick SP, et al.
Hyperbilirubinemia in very preterm infants and neurodevelopmental outcome at
two years of age: results of a national collaborative survey. Pediatrics
1989;83:915-920.
185. Oh W, Tyson JE, Fanaroff AA, et al. Association between peak serum
bilirubin and neurodevelopmental outcomes in extremely low birth weight
186. Ikonen RS, Kuusinen EJ, Janas MO, et al. Possible etiologic factors in
extensive periventricular leucomalacia of preterm infants. Acta Paeditr Scand
1988;77:489-495.
187. Trounce J, Shaw DE, Levine MI, et al. Clinical risk factors and
periventricular leucomalacia. Arch Dis Child 1988;63:17-22.
188. Ikonen RS, Koivkko MJ, Laippala, et al. Hyperbilirubinemia, hypocarbia and
periventricular leukomalacia in preterm infants: relationship to cerebral palsy.
Acta Paediatr 1992;81:802-807.
189. Hack M, Wilson-Costello D, Friedman H, et al. Neurodevelopment and

predictors of outcomes of children with birth weights of less than 1000 g 19921995. Arch Pediatr Adolesc Med 2000;154: 725-731.
190. DeVries KL, Lary S, Dubowitz LMS. Relationship of serum bilirubin levels to
ototoxicity and deafness in high-risk low-birth-weight infants. Pediatrics
1985;76:351-354.
191. DeVries L, Lary S, Whitelaw A. Relationship of serum bilirubin levels and

hearing impairment in newborn infants. Early Hum Dev 1987;15:269-277.
192. Morris B. A randomized trial of aggressive or conservative phototherapy

for extremely low birth weight infants. National Institute of Child Health and
Development Neonatal Research Network. Personal communication, 2002.
193. Ahvenainen EK, Call JD. Pulmonary hemorrhage in infants. A descriptive

study. Am J Pathol 1952;28:1-18.
194. Zuelzer WW, Mudgett RT. Kernicterus: etiologic study based on an analysis
of 55 cases. Pediatrics 1950;6:452-474.
195. Johnson L, Sarmiento F, Blanc WA, et al. Kernicterus in rats with an

inherited deficiency of glucouronyl transferase. Am J Dis Child 1960;97:591-608.
196. Horiguchi T, Bauer C. Ethnic differences in neonatal jaundice: comparison

of Japanese and Caucasian newborn infants. Am J Obstet Gynecol 1975;121:7174.
197. Linn S, Schoenbaum SC, Monson RR, et. al. Epidemiology of neonatal
hyperbilirubinemia. Pediatrics 1985;75:770-774.
198. Newman TB, Easterling MJ, Goldman ES, et al. Laboratory evaluation of
jaundiced newborns: frequency, cost and yield. Am J Dis Child 1990;144:364368.
199. Munroe M, Shah CP, Badgley R, et al. Birthweight, length, head

circumference and bilirubin level in Indian newborns in the Sioux Lookout Zone,
Northwestern Ontario. Can Med Assoc J 1984;131:453-456.
200. Johnson JD, Angelus P, Aldrich M, et al. Exaggerated jaundice in Navajo

neonates: the role of bilirubin production. Am J Dis Child 1986;140:889-890.
201. Otterbein L, Mantell L, Choi A. Carbon monoxide provides protection

against hyperoxic lung injury. Am J Physiol Lung Cell Mol Physiol 1999;20:L688L694.
202. Engle WD, Jackson GL, Sendelbach D, et al. Assessment of a

transcutaneous device in the evaluation of neonatal hyperbilirubinemia in a
primarily Hispanic population. Pediatrics 2002;110:61-67.
203. Newman TB, Escobar GJ, Gonzales VM, et al. Frequency of neonatal
bilirubin testing and hyperbilirubinemia in a large health maintenance
organization. Pediatrics 1999;104:1198-1203.
204. Maisels MJ, Ostrea EJ Jr, Touch S, et al. Evaluation of a new

transcutaneous bilirubinometer. Pediatrics 2004;113:1628-1635.
205. Hardy JB, Drage JS, Jackson EC. The first year of life: the Collaborative
Perinatal Project of the National Institutes of Neurological and Communicative
Disorders and Stroke. Baltimore, MD: Johns Hopkins University, 1979.
206. Newman TB, Xiong B, Gonzales VM, et al. Prediction and prevention of
extreme neonatal hyperbilirubinemia in a mature health maintenance
organization. Arch Pediatr Adolesc Med 2000;154:1140-1147.
207. Khoury MJ, Calle EE, Joesoef RM. Recurrence risk of neonatal
hyperbilirubinemia in siblings. Am J Dis Child 1988;142: 1065-1069.
208. Nielsen HE, Haase P, Blaabjerg J, et al. Risk factors and sib correlation in
physiological neonatal jaundice. Acta Paediatr Scand 1987;76:504-511.
209. Diwan VK, Vaughan TL, Yang CY. Maternal smoking in relation to the
incidence of early neonatal jaundice. Gynecol Obstet Invest 1989;27:22-25.
210. Knudsen A. Maternal smoking and the bilirubin concentration in the first
three days of life. Eur J Obstet Gynecol Reprod Biol 1991;25:37-41.
211. Gale R, Seidman DS, Dollberg S, et al. Epidemiology of neonatal jaundice

in the Jerusalem population. J Pediatr Gastroenterol Nutr 1990;10:82-86.
212. Jones JB. The smoking disease. Br Med J 1971;1:228.
213. Jhrig D, Jhrig K, Striet S, et al. Neonatal jaundice in infants of diabetic

mothers. Acta Paediatr Scand 1989;360:101-107.
214. Stevenson DK, Ostrander CR, Cohen RS, et al. Pulmonary excretion of
carbon monoxide in he human infant as an index of bilirubin production. II.
Evidence for the possible effect of maternal prenatal glucose metabolism on
postnatal bilirubin production in a mixed population of infants. Eur J Pediatr
1981;137:255-259.
215. Widness JA, Susa JB, Garcia JF, et al. Increased erythropoiesis and
elevated erythropoietin in infants born to diabetic mothers and in
hyperinsulinemic rhesus fetuses. J Clin Invest 1981;67:637- 642.
216. Berk MA, Mimouni F, Miodovnik M, et al. Macrosomia in infants of insulin-
dependent diabetic mothers. Pediatrics 1989;83: 1029-1034.
217. Maisels MJ. Neonatal Jaundice. In: Sinclair JC, Bracken MB, eds. Effective
care of the newborn infant. Oxford: Oxford University, 1992:507-561.
218. Maisels MJ, Gifford K, Antle CE, et al. Jaundice in the healthy newborn
infant: a new approach to an old problem. Pediatrics 1988;81:505-511.
219. Friedman L, Lewis PJ, Clifton P, et al. Factors influencing the incidence of
neonatal jaundice. Br Med J 1978;1:1235-1237.
220. Brodersen R, Friis-Hansen B, Stern L. Drug-induced displacement of

bilirubin from albumin in the newborn. Dev Pharmacol Ther 1983;6:217-229.
221. Pedersen H, Morishima HO, Finster M. Uptake and effects of local

anesthetics in mother and fetus. Int Anesthesiol Clin 1978;16:73-89.
222. Ferguson JE, II, Schutz TE, Stevenson DK. Neonatal bilirubin production
after preterm labor tocolysis with nifedipine. Dev Pharmacol Ther 1989;12:113117.
223. Caritis SN, Toig G, Heddinger LA, et al. A double-blind study comparing
ritodrine and terbutaline in the treatment of preterm labor. Am J Obstet Gynecol
1984;150:7-14.
224. Drew JH, Kitchen WH. The effect of maternally administered drugs on
bilirubin concentrations in the newborn infant. J Pediatr 1976;89:657-661.
P.840
225. Nathenson G, Cohen MI, Litt IF, et al. The effect of maternal heroin
addition on neonatal jaundice. J Pediatr 1972;81:899-903.
226. Valaes T, Harvey-Wilkes K. Pharmacologic approaches to the prevention

and treatment of neonatal hyperbilirubinemia. Clin Perinatol 1990;17:245-274.
227. Yamauchi Y, Yamanouchi I. Difference in TcB readings between full term

newborn infants born vaginally and by cesarean section. Acta Paediatr Scand
1989;79:824-828.
228. Wallace RL, Schifrin BS, Paul RH. The delivery route for very-low-birthweight infants. A preliminary report of a randomized, prospective study. J
Reprod Med 1984;29:736-740.
229. Dell DL, Sightler SE, Plauche WC. Soft cup vacuum extraction: a
comparison of outlet delivery. Obstet Gynecol 1985;66:624-628.
230. Vacca A, Grant A, Wyatt G, et al. Portsmouth operative delivery trial: a

comparison of vacuum extraction and forceps delivery. Br J Obstet Gynaecol
1983;90:1107-1112.
231. Black VD, Lubchenco LO, Luckey DW, et al. Developmental and neurologic
sequelae in the neonatal hyperviscosity syndrome. Pediatrics 1982;69:426-431.
232. Black VD, Lubchenco LO, Koops BL, et al. Neonatal hyperviscosity:
Randomized study of effect of partial plasma exchange transfusion on long-term
outcome. Pediatrics 1985;75:1048-1053.
233. Goldberg K, Wirth FH, Hathaway WE, et al. Neonatal hyperviscosity II.
Effect of partial plasma exchange transfusion. Pediatrics 1982;69:419-425.
234. Saigal S, O'Neill A, Surainder Y, et al. Placental transfusion and

hyperbilirubinemia in the premature. Pediatrics 1972;49: 406-419.
235. Davidson LT, Merritt KK, Weech AA. Hyperbilirubinemia in the newborn.
Am J Dis Child 1941;61:958-980.
236. Knudsen A, Lebech M. Maternal bilirubin, cord bilirubin and placental

function at delivery in the development of jaundice in mature newborns. Acta
Obstet Gynecol Scand 1989; 68:719-724.
237. Rosenfeld J. Umbilical cord bilirubin levels as a predictor of subsequent

hyperbilirubinemia. J Fam Pract 1986;23:556-558.
238. Risemberg HM, Mazzi E, MacDonald MG, et al. Correlation of cord bilirubin
levels with hyperbilirubinemia in ABO incompatibility. Arch Dis Child

1977;52:219-222.
239. Soskolne El, Schumacher R, Fyock C, et al. The effect of early discharge
and other factors on readmission rates of newborns. Arch Pediatr Adolesc Med
1996;150:373-379.
240. Wood B, Culley P, Roginski C, et al. Factors affecting neonatal jaundice.

Arch Dis Child 1979;54:111-115.
241. Maisels MJ, Kring EA. Length of stay, jaundice and hospital readmission.
Pediatrics 1998;101:995-998.
242. Frishberg Y, Zelicovic I, Merlob P, et al. Hyperbilirubinemia and influencing

factors in term infants. Isr J Med Sci 1989; 25:28-31.
243. Gourley GR, Kreamer B, Arend R. The effect of diet on feces and jaundice
during the first three weeks of life. Gastroenterology 1992;103:660.
244. Gourley GR. Breastfeeding, neonatal jaundice and kernicterus. Semin

Neonatol 2002;7:135-141.
245. Rubaltelli FF. Unconjugated and conjugated bilirubin pigments during

perinatal development. IV. The influence of breast-feeding on neonatal
hyperbilirubinemia. Biol Neonate 1993;64: 104-109.
246. Bertini G, Dani C, Trochin M, et al. Is breast feeding really favoring early
neonatal jaundice? Pediatrics 2001;107.
247. Schneider AP. Breast milk jaundice in the newborn. A real entity. JAMA
1986;255:3270-3274.
248. Seidman DS, Stevenson DK, Ergaz Z, et al. Hospital readmission due to
neonatal hyperbilirubinemia. Pediatrics 1995;96:727-729.
249. Maisels MJ, Kring E. Risk of sepsis in newborns with severe

hyperbilirubinemia. Pediatrics 1992;90:741-743.
250. Gartner L. Breastfeeding and jaundice. J Perinatol 2001;21:S25-S29.
251. De Carvalho M, Klaus MH, Merkatz RB. Frequency of breastfeeding and

serum bilirubin concentration. Am J Dis Child 1982;136:737-738.
252. Yamauchi Y, Yamanouchi I. Breast-feeding frequency during the first 24

hours after birth in full-term neonates. Pediatrics 1990;86:171-175.
253. Auerbach KG, Gartner LM. Breast feeding and human milk: Their
association with jaundice in the neonate. Clin Perinatol 1987;14:89-107.
254. Maisels MJ, Gifford K. Normal serum bilirubin levels in the newborn and the
effect of breast feeding. Pediatrics 1986;78:837- 843.
255. Kivlahan C, James EJP. The natural history of neonatal jaundice. Pediatrics
1984;74:364-370.
256. Maisels MJ, D'Archangelo MR. Breast feeding and jaundice in the first six
weeks of life. Pediatr Res 1983;17:324A(abst).
257. Alonso EM, Whitington PF, Whitington SH, Rivard WA, Given G.
Enterohepatic circulation of non-conjugated bilirubin in rats fed with human
milk. J Pediatr 1991;118:425-430.
258. Gathwala G, Sharma S. Phototherapy induces oxidative stress in

premature neonates. Indian J Gastroenterol 2002;21:153-154.
259. Monaghan G, McLellan A, McGeehan A, et al. Gilbert's syndrome is a

contributory factor in prolonged unconjugated hyperbilirubinemia of the
newborn. J Pediatr 1999;134:441-446.
260. Maruo Y, Nishizawa K, Sato H, et al. Prolonged unconjugated

hyperbilirubinemia associated with breast milk and mutations of the bilirubin
uridine diphosphate glucuronosyltransferase gene. Pediatrics 2000;106:e59.
261. De Carvalho M, Robertson S, Klaus M. Fecal bilirubin excretion and serum

bilirubin concentration in breast-fed and bottle-fed infants. J Pediatr
1985;107:786-790.
262. Verkade HJ. A novel hypothesis on the pathophysiology of neonatal

jaundice. J Pediatr 2002;141:594-595.
263. Hamosh M. Digestion in the newborn. Clin Perinatol 1996;23: 191-209.
264. Yoshioka H. Development and differences of intestinal flora in the neonatal

period in breast-fed and bottle-fed infants. Pediatrics 1983;72:317-321.
265. Gourley GR, Arend RA. Beta-glucuronidase and hyperbilirubinemia in

breast-fed and formula-fed babies. Lancet 1986;1:644-646.
266. Wilson DC, Afrasiabi M, Reid MM. Breast-milk beta-glucuronidase and

exaggerated jaundice in the early neonatal period. Biol Neonate 1992;61:232234.
267. Cottrell BH, Anderson GC. Rectal or axillary temperature measurement:

Effect on plasma bilirubin and intestinal transit of meconium. J Pediatr
268. Weisman LE, Merenstein GB, Digirol M, et al. The effect of early meconium
evacuation on early-onset hyperbilirubinemia. Am J Dis Child 1983;137:666668.
269. Daum F, Cohen MI, McNamara H. Experimental toxicologic studies on a

phenol detergent associated with neonatal hyperbilirubinemia. J Pediatr
1976;89:853.
270. Wysowski DK, Flynt JW, Goldfield M, et al. Epidemic neonatal

hyperbilirubinemia and use of a phenolic disinfectant detergent. Pediatrics
1978;61:165.
271. Moore LG, Newberry MA, Freeby GM, Crnic LS. Increased incidence of
neonatal hyperbilirubinemia at 3,100 m in Colorado. Am J Dis Child
1984;138:157-161.
272. Atland PD, Parker MG. Bilirubinemia and intravascular hemolysis during
acclimatization to high altitude. Int J Biometeorol 1977;21:165-170.
273. Berendsohn S. Hepatic function at high altitudes. Arch Intern Med

1962;109:256-264.
274. Barron ESG. Bilirubinemia. Medicine (Baltimore) 1931;10:114-115.
275. Freeman J, Lesko S, Mitchell AA, et al. Hyperbilirubinemia following

exposure to pancuronium bromide in newborns. Dev Pharmacol Ther
1990;14:209-215.
276. Lambert GH, Muraskas J, Anderson CL, et al. Direct hyperbilirubinemia

associated with chloral hydrate administration in the newborn. Pediatrics
1990;86:277-281.
277. Reimche LD, Sankaran K, Hindmarsh KW, et al. Chloral hydrate sedation in
neonates and infantsclinical and pharmacologic considerations. Dev Pharmacol
Ther 1989;12:57-64.
278. McDonagh AF. Is bilirubin good for you? Clin Perinatol 1990;17:359-369.
279. Benaron DA, Bowen FW. Variation of initial serum bilirubin rise in newborn
infants with type of illness. Lancet 1991;338:78-81.
280. Saigal S, Lunyk O, Bennett KJ, et al. Serum bilirubin levels in breast- and
formula-fed infants in the first 5 days of life. Can Med Assoc J 1982;127:985989.
281. Knudsen A. Prediction of the development of neonatal jaundice by

increased umbilical cord blood bilirubin. Acta Pediatr Scand 1989;78:217-221.
282. Lo S, Doumas BT, Ashwood E. Performance of bilirubin determinations in

US laboratoriesrevisited. Clin Chem 2004;50: 190-194.
P.841
283. Vreman HJ, Verter J, Oh W, et. al. Interlaboratory variability of bilirubin
measurements. Clin Chem 1996;42:869-873.
284. Doumas BT, Eckfeldt JH. Errors in measurement of total bilirubin: A

perennial problem. Clin Chem 1996;42:845-848.
285. Stevenson DK, Fanaroff AA, Maisels MJ, et al. Prediction of

hyperbilirubinemia in near-term and term infants. Pediatrics 2001;108:31-39.
286. Maisels MJ, Kring EA. The natural history of neonatal bilirubinemia. Pediatr
Res 2004;55:458A.
287. Li R, Zhao Z, Mokdad A, et al. Prevalence of breastfeeding in the United

States: The 2001 national immunization survey. Pediatrics 2003;111:11981201.
288. Lee K-S, Perlman M, Ballantyne M. Association between duration of

neonatal hospital stay and readmission rate. J Pediatr 1995;127:758-766.
289. Bhutani VK, Johnson L, Sivieri EM. Predictive ability of a predischarge hourspecific serum bilirubin for subsequent significant hyperbilirubinemia in healthyterm and near-term newborns. Pediatrics 1999;103:6-14.
290. Maisels MJ, Fanaroff AA, Stevenson DK, et al. Serum bilirubin levels in an
international, multiracial newborn population. Pediatr Res 1999;45:167A.
291. Okolo AA, Omene JA, Scott-Emaukpor AB. Physiologic jaundice in the
Nigerian neonate. Biol Neonate 1988;53:132-137.
292. Yamauchi Y, Yamanouchi I. Transcutaneous bilirubinometry in normal

Japanese infants. Acta Paediatr Jpn 1989;31:65-72.
293. Maisels MJ, Newman TB. Predicting hyperbilirubinemia in newborns: the

importance of timing. Pediatrics 1999;103:493- 495.
294. Koenig JM. Evaluation and treatment of erythroblastosis fetalis in the

neonate. In: Christensen RD, ed. Hematologic problems of the neonate.
295. Bowman JM. RhD hemolytic disease of the newborn. N Engl J Med
1998;339:1775-1777.
296. Bowman JM. The management of alloimmune fetal hemolytic disease. In:
Maisels MJ, Watchko JF, eds. Neonatal jaundice. London, UK: Harwood
Academic Publishers, 2000:23-36.
297. Bowman JM. Immune hemolytic disease. In: Nathan DG, Orkin SH, eds.
Hematology of infancy and childhood. Philadelphia: WB Saunders, 1998:54-78.
298. Chavez GF, Mulinare J, Edmonds LD. Epidemiology of Rh hemolytic disease

of the newborn in the United States. JAMA 1991;265:3270-3274.
299. Sebring ES, Polesky HF. Detection of fetal hemorrhage in Rh immune

globulin candidates. A rosetting technique using enzyme-treated Rh2Rh2
indicator erythrocytes. Transfusion 1982;22:468-471.
300. Voto LS, Mathet ER, Zapaterio JL, et al. High-dose gammaglobulin (IVIG)
followed by intrauterine transfusions (IUTs): a new alternative for the treatment
of severe fetal hemolytic disease. J Perinat Med 1997;25:85-88.
301. Blair DK, Vander Straten MC, Gest AL. Hydrops fetalis in sheep from rapid
induction of anemia. Pediatr Res 1994;35:560-564.
302. Nicolaides KH, Warenski JC, Rodeck CH. The relationship of fetal plasma
protein concentration and hemoglobin level to the development of hydrops in
rhesus isoimmunization. Am J Obstet Gynecol 1985;152:341-344.
303. Grannum PA, Copel JA, Moya FR, et al. The reversal of hydrops fetalis by
intravascular intrauterine transfusion in severe isoimmune fetal anemia. Am J
304. Moya FR, Grannum PA, Riddick L, et al. Atrial natriuretic factor in hydrops
fetalis caused by Rh isoimmunization. Arch Dis Child 1990;65:683-686.
305. Weiner C. Nonhematologic effects of intravascular transfusion on the

human fetus. Semin Perinatol 1989;13:338-341.
306. Barss VA, Doubilet PM, St. John-Sutton M, et al. Cardiac output in a fetus
with erythroblastosis fetalis: assessment using pulsed Doppler. et al
1987;70:442-444.
307. Ohls RK, Wirkus PEN, Christensen RD. Recombinant erythropoieitin as

treatment of late hypo-regenerative anemia of Rh hemolytic disease. Pediatrics
1992;90:678-680.
308. Dallacasa P, Ancora G, Miniero R, et al. Erythropoietin course in newborns

with Rh hemolytic disease transfused and not transfused in utero. Pediatr Res
1996;40:357-360.
309. Giblet ER. Blood groups and blood transfusion. In: Braunwald E,
Isselbacher KJ, Petersdordf RG, et al, eds. New York: Harrison's principles of
internal medicine, 1987:1483-1489.
310. Katz MA, Kanto WP, Korotkin JH. Recurrence rate of ABO hemolytic disease
of the newborn. Obstet Gynecol 1982;59:611- 614.
311. Ozolek J, Watchko J, Mimouni F. Prevalence and lack of clinical significance

of blood group incompatibility in mothers with blood type A or B. J Pediatr
1994;125:87-91.
312. Kanto WP, Marino B, Godwin AS, et al. ABO hemolytic disease: a
comparative study of clinical severity and delayed anemia. Am J Dis Child
1978;62:365-369.
313. Osborn LM, Lenarsky C, Oakes RC, et al. Phototherapy in full-term infants
with hemolytic disease secondary to ABO incompatibility. Pediatrics
1984;74:371-374.
314. Quinn MW, Weindling AM, Davidson DC. Does ABO incompatibility matter?
Arch Dis Child 1988;63:1258-1260.
315. Serrao PA, Modanlou HD. Significance of anti-A and anti-B

isohemagglutinins in cord blood of ABO incompatible newborn infants:
correlation with hyperbilirubinemia. J Perinatol 1989; 9:154-158.
316. Herschel M, Karrison T, Wen M, et al. Isoimmunization is unlikely to be the

cause of hemolysis in ABO-Incompatible but direct antiglobulin test-negative
neonates. Pediatrics 2002;110: 127-130.
317. Kaplan M, Hammerman C, Renbaum P, et al. Gilbert's syndrome and

hyperbilirubinaemia in ABO-incompatible neonates. Lancet 2000;356:652-653.
318. Leistikow EA, Collin MF, Savastano GD, et al. Wasted health care dollars.
Routine cord blood type and Coombs' testing. Arch Pediatr Adolesc Med
1995;149:1147-1151.
319. Maisels MJ, Kring EA. Early discharge from the newborn nursery: effect on
scheduling of follow-up visits by pediatricians. Pediatrics 1997;100:72-74.
320. Gallagher PG, Forget BG, Lux SE. Disorders of the erythrocyte membrane.
In: Nathan DG, Orkin SH, eds. Hematology of infancy and childhood.
321. Gallagher PG. Disorders of erythrocyte metabolism and shape. In:

Christensen RD, ed. Hematologic problems of the neonate. Philadelphia: WB
Saunders, 2000:209-237.
322. Watchko JF. Indirect hyperbilirubinemia in the neonate. In: Maisels MJ,
Watchko JF, eds. Neonatal jaundice. London, UK: Harwood Academic Publishers,
2000:51-66.
323. Iolascon A, Faienza MF, Moretti A, et al. UGT1 promoter polymorphism

accounts for increased neonatal appearance of hereditary spherocytosis. Blood
1998;91:1093.
324. Valaes T. Neonatal jaundice in glucose-6-phosphate dehydrogenase

deficiency. In: Maisels MJ, Watchko JF, eds. Neonatal jaundice. London, UK:
Harwood Academic Publishers, 2000:67-74.
325. Brown AK, Damus K, Kim MH, et al. Factors relating to readmission of
term and near-term neonates in the first two weeks of life. Early discharge
survey group of the Health Professional Advisory Board of the Greater New York
Chapter of the March of Dimes. J Perinat Med 1999;27:263-275.
326. Kaplan M, Hammerman C. Glucose-6-phosphate dehydrogenase-deficient

neonates: A potential cause for concern in North America. Pediatrics
2000;106:1478-1480.
327. Kaplan M, Hammerman C. Severe neonatal hyperbilirubinemia: a potential

complication of glucose-6-phosphate dehydrogenase deficiency. Clin Perinatol
1998;25:575-590.
328. MacDonald M. Hidden risks: early discharge and bilirubin toxicity due to
glucose-6-phosphate dehydrogenase deficiency. Pediatrics 1995;96:734-738.
329. Washington EC, Ector W, Abboud M, Ohning B, Holden K. Hemolytic

jaundice due to G6PD deficiency causing kernicterus in a female newborn.
South Med J 1995;88:776-779.
330. Herschel M, Kaplan M, Hammerman C, et al. Increased hemolysis in G6PD

deficient African American Neonates. Pediatr Res 2003;53:399A.
331. Kaplan M, Vreman H, Hammerman C, et al. Favism by proxy in nursing

glucose-6-phosphate dehydrogenase-deficient neonates. J Perinatol 1998;18
(6):477-479.
332. Beutler E. Glucose-6-phosphate dehydrogenase deficiency. Blood

1994;84:3613-3636.
333. Kaplan M., Hammerman C. Glucose-6-phosphate dehydrogenase

deficiency: a potential source of severe neonatal hyperbilirubinaemia and
kernicterus. Semin Neonatol 2002;7:121-128.
334. Kaplan M, Algur N, Hammerman C. Onset of jaundice in glucose-6phosphate dehydrogenase-deficient neonates. Pediatrics 2001;108:956-959.
P.842
335. Kaplan M., Beutler E, Vreman HJ, et al. Neonatal hyperbilirubinemia in
glucose-6-phosphate dehydrogenase deficient heterozygotes. Pediatrics
1999;104:68-74.
336. Herschel M, Ryan M, Gelbart T, Kaplan M. Hemolysis and

hyperbilirubinemia in an African American neonate heterozygous for glucose-6phosphate dehydrogenase deficiency. J Perinatol 2002;22:577-579.
337. Kaplan M, Abramov A. Neonatal hyperbilirubinemia associated with

glucose-6-phosphate dehydrogenase deficiency in Sephardic-Jewish infants:
incidence, severity and the effect of phototherapy. Pediatrics 1992;90:401-405.
338. Meloni T, Cutillo S, Testa U, Luzatto L. Neonatal jaundice and severity of

glucose-6-phosphate dehydrogenase deficiency in Sardinian babies. Early Hum
Dev 1987;15:317-322.
339. Slusher TM, Vreman HJ, McLaren D, et al. Glucose-6-phosphate

dehydrogenase deficiency and carboxy hemoglobin concentrations associated
with bilirubin related morbidity and death in Nigerian infants. J Pediatr
1995;126:102-108.
340. Necheles TF, Rai US, Valaes T. The role of haemolysis in neonatal
hyperbilirubinemia as reflected by carboxyhemoglobin levels. Acta Paediatr
Scand 1976;65:361-367.
341. Kaplan M, Vreman HJ, Hammerman C, et al. Contribution of haemolysis to

jaundice in Sephardic Jewish glucose-6-phosphate dehydrogenase deficient
neonates. Br J Haematol 1996;93:822-827.
342. Kaplan M, Hammerman C, Renbaum P, et al. Differing pathogenesis of

perinatal bilirubinemia in glucose-6-phosphate dehydrogenase-deficient versus
normal neonates. Pediatr Res 2001; 50:532-537.
343. Muraca M, Rubaltelli FF, Blanckaert N, et al. Unconjugated and conjugated

bilirubin pigments during perinatal development. II. Studies on serum of
healthy newborns and of neonates with erythroblastosis fetalis. Biol Neonate
1990;57:1-9.
344. Kaplan M, Muraca M, Hammerman C, et al. Bilirubin conjugation, reflected

by conjugated bilirubin fractions, in glucose-6-phosphate dehydrogenasedeficient neonates: a determining factor in the pathogenesis of
hyperbilirubinemia. Pediatrics 1998;102(3):e37.
345. Kaplan M, Renbaum P, Levi-Lahad E, et al. Gilbert syndrome and glucose6-phosphate dehydrogenase deficiency: a dose-dependent genetic interaction
crucial to neonatal hyperbilirubinemia. Proc Natl Acad Sci U S A 1997;94:1212812132.
346. Iolascon A, Faienza MF, Perrotta S, et al. Gilbert's syndrome and jaundice
in glucose-6-phosphate dehydrogenase deficient neonates. Haematologica
1999;84:99-102.
347. Brown WR, Boon WH. Hyperbilirubinemia and kernicterus in glucose-6phosphate dehydrogenase deficient infants in Singapore. Pediatrics
1968;41:1055-1062.
348. Gibbs WN, Gray R, Lowry M. G6PD deficiency and neonatal jaundice in
Jamaica. Br J Hematol 1979;43:263-274.
349. Bainbridge R, Khoury J, Mimounie F. Jaundice in neonatal sickle cell

disease: a case controlled study. Am J Dis Child 1988;148: 569.
350. Bard H. Hemoglobin synthesis and metabolism during the neonatal period.
In: Christensen RD, ed. Hematologic problems of the neonate. Philadelphia: WB
Saunders, 2000:365-388.
351. Rajagopalan I, Katz BZ. Hyperbilirubinemia secondary to hemolysis of

intrauterine intraperitoneal blood transfusion. Clin Pediatr 1984;23:511-512.
352. Rose J, Berdon WE, Sullivan T, et al. Prolonged jaundice as presenting sign
of massive adrenal hemorrhage in newborn. Radiology 1971;98:263-272.
353. Wright K, Tarr PI, Hickman RO, et al. Hyperbilirubinemia secondary to

delayed absorption of intraperitoneal blood following intrauterine transfusion. J
Pediatr 1982;100:302-304.
354. Epstein MF, Leviton A, Kuban KC, et al. Bilirubin intraventricular

hemorrhage and phenobarbital in very low birth weight babies. Pediatrics
1988;82:350.
355. Pasnick M, Lucey JF. Serum bilirubin in preterm infants following

intracranial hemorrhage. Pediatr Res 1983;17:329A.
356. Kemper K, Forsyth B, McCarthy P. Jaundice, terminating breast-feeding,

and the vulnerable child. Pediatrics 1989;84:773-779.
357. Bleicher MA, Reiner MA, Rapaport SA, et al. Extraordinary

hyperbilirubinemia in a neonate with idiopathic hypertrophic pyloric stenosis. J
Pediatr Surg 1979;14:527.
358. Wooley MM, Felsher BF, Asch MJ, et al. Jaundice, hypertrophic pyloric
stenosis, and glucuronyl transferase. J Pediatr Surg 1974;9:359.
359. Trioche P, Chalas J, Francoual J, et al. Jaundice with hypertrophic pyloric

stenosis as an early manifestation of Gilbert syndrome. Arch Dis Child
1999;81:301-303.
360. Labrune P, Myara A, Huguet P, et al. Jaundice with hypertrophic pyloric

stenosis: a possible early manifestation of Gilbert syndrome. J Pediatr
1989;115:93-95.
361. Job H, Hart G, Lealman G. Improvements in long term phototherapy for

patients with Crigler-Najjar syndrome type I. Phys Med Biol 1996;41(11):25492556.
362. Blumenschein SD, Kallen RJ, Storey P, et al. Familial non-hemolytic

jaundice with late onset of neurological damage. Pediatrics 1968;42:786.
363. Chalasani N, Roy-Chowdhury N, Roy-Chowdury J, et al. Kernicterus in an

adult who is heterozygous for Crigler-Najjar syndrome and homozygous for
Gilbert-type genetic defect. Gastroenterology 1997;112:2099-2103.
364. Ahmad P, Pratt A, Land VJ, et al. Multiple plasma exchanges successfully
maintained a young adult patient with Crigler-Najjar syndrome type I. J Clin
Apheresis 1989;5:17.
365. Shevell MI, Bernard B, Adelson JW, et al. Crigler-Najjar syndrome type I:
Treatment by home phototherapy followed by orthotopic hepatic
transplantation. J Pediatr 1987;110:429.
366. Fox IJ, Roy-Chowdhury J, Kaufman SS, et al. Treatment of the CriglerNajjar syndrome type I with hepatocyte transplantation. N Engl J Med
1998;338:1422-1426.
367. Roy Chowdhury J, Strom SC, Kaufman SS, et al. Human hepatocyte
transplantation: gene therapy and more? Pediatrics 1998; 647-648.
368. Kim BH, Takahashi M, Tada K, et al. Cell and gene therapy for inherited
deficiency of bilirubin glucuronidation. J Perinatol 1996;16:S67-S72.
369. Rubaltelli FF, Guerrini P, Reddi E, et al. Tin-protoporphyrin in the

management of children with Crigler-Najjar disease. Pediatrics 1989;84:728731.
370. Kappas A. A method for interdicting the development of severe jaundice in

newborns by inhibiting the production of bilirubin. Pediatrics 2004;113:119-123.
371. Van der Veere CN, Jansen PLM, Sinaasappel M, et al. Oral calcium
phosphate: a new therapy for Crigler-Najjar disease? Gastroenterology
1997;112:455-462.
372. Rubaltelli FF, Novello A, Zancan L, et al. Serum and bile bilirubin pigments
in the differential diagnosis of Crigler-Najjar disease. Pediatrics 1994;94:553556.
373. Lee W, McKiernan P, Beath S, et al. Bile bilirubin pigment analysis in

disorders of bilirubin metabolism in early infancy. Arch Dis Child 2001;85:38-42.
374. Sampietro M, Iolascon A. Molecular pathology of Crigler-Najjar type I and

II and Gilbert's syndromes. Haematologica 1999;84:150-157.
375. Clarke DJ, Moghrabi N, Monaghan G, et al. Genetic defects of the UDPglucoronosyltransferase-1 (UGT1) gene that cause familial nonhaemolytic
unconjugated hyperbilirubinemias. Clin Chim Acta 1997;266:63-74.
376. Sugatani J, Kojima H, Ueda A, et al. The phenobarbital response enhancer

module in the human bilirubin UDP-glucuronosyltransferase UGT1A1 gene and
regulation by the nuclear receptor CAR. Hepatology 2001;33:1232-1238.
377. Van der Veere CN, Sinaasappel M, McDonagh AF, et al. Current therapy for
Crigler-Najjar syndrome type I: report of a world registry. Hepatology
1996;24:311-315.
378. Kaplan M, Hammerman C, Rubaltelli F, et al. Hemolysis and bilirubin

conjugation in association with UDP-glucuronosyltransferase 1A1 promoter
polymorphism. Hepatology 2002;35:905-911.
379. Bosma PJ, Roy-Chowdhury J, Bakker C, et al. The genetic basis of the
reduced expression of bilirubin UDP-glucuronosyl transferase 1 in Gilbert's
syndrome. N Engl J Med 1995;333:1171-1175.
380. Kadakol A, Ghosh SS, Sappal BS, et al. Genetic lesions of bilirubin uridinediphospho-glucuronate glucuronosyltransferase (UGT1A1) causing Crigler-Najjar
and Gilbert syndromes: Correlation of genotype to phenotype. Hum Mutat
2000;16:297-306.
381. Monaghan G, Ryan M, Seddon R, et al. Genetic variation in bilirubin UDPglucuronosyltransferase gene promoter and Gilbert's syndrome. Lancet
1996;347:578-581.
382. Kadakol A, Sappal BS, Ghosh SS, et al. Interaction of coding area
mutations and the Gilbert-type promoter abnormality of the UGT1A1 gene
causes moderate degrees of unconjugated hyperbilirubinemia and may lead to
neonatal kernicterus. J Med Genet 2001;38:244-249.
P.843
383. Bancroft JD, Kreamer B, Gourley GR. Gilbert's syndrome accelerates
development of neonatal jaundice. J Pediatr 1998;132(4): 656-660.
384. Roy-Chowdhury N, Deocharan B, Bejjanki HR, et al. The present of a

Gilbert-type promotor abnormality increases the level of neonatal
hyperbilirubinemia. Hepatology 1997;26:370A.
385. Laforgia N, Faienza MF, Rinaldi A, et al. Neonatal hyperbilirubinemia and

Gilbert's syndrome. J Perinat Med 2002;30:166-169.
386. Maruo Y, Nishizawa K, Sato H, et al. Association of neonatal

hyperbilirubinemia with bilirubin UDP-glucuronosyltransferase polymorphism.
Pediatrics 1999;103(6):1224-1227.
387. Akaba K, Kimura T, Sasaki A, et al. Neonatal hyperbilirubinemia and

mutation of the bilirubin uridine diphosphate-glucuronosyltransferase gene: A
common missense mutation among Japanese, Koreans and Chinese. Biochem
Mol Biol Int 1998;46:21-26.
388. Hsieh S-Y, Wu Y-H, Lin D-Y, et al. Correlation of mutational analysis to
clinical features in Taiwanese patients with Gilbert's syndrome. Am J
Gastroenterol 2001;96:1188-1192.
389. Maruo Y, Sato H, Yamano T, et al. Gilbert syndrome caused by a

homozygous missnse mutation (Tyr486Asp) of bilirubin UDPglucuronosyltransferase gene. J Pediatr 1998;132:1045-1047.
390. Maisels MJ, Newman TB. Kernicterus in otherwise healthy, breast-fed term
newborns. Pediatrics 1995;96:730-733.
391. Berry GT. Inborn errors of carbohydrate, ammonia, aminoacid, and organic
acid metabolism. In: Taeusch HW, Ballard RA, eds. Avery's diseases of the
newborn. Philadelphia: WB Saunders, 1998:245-274.
392. Akerren Y. Prolonged jaundice in the newborn associated with congenital

myxedema. A syndrome of practical importance. Acta Paediatr 1954;43:411425.
393. Labrune P, Myara A, Huguet P, et al. Bilirubin uridine diphosphate

glucuronosyltransferase hepatic activity in jaundice associated with congenital
hypothyroidism. J Pediatr Gastroenterol Nutr 1992;14:79-82.
394. Smith DW, Klein AM, Henderson JR, et al. Congenital hypothyroidism signs and symptoms in the newborn period. J Pediatr 1975;87:958-962.
395. Weldon AP, Danks DM. Congenital hypothyroidism and neonatal jaundice.
Arch Dis Child 1972;(47):469-471.
396. Lees MH, Ruthven CRJ. The effects of triiodothyronine on neonatal

hyperbilirubinemia. Lancet 1959;2:371-373.
397. Shrand H, Ruthven CRJ. Effect of triiodothyronine on serum bilirubin level

in neonatal development of the premature infant. Lancet 1960;2:1274-1275.
398. Van Steenbergen W, Fevery J, DeVos R, et al. Thyroid hormones and the
hepatic handling of bilirubin. I. Effects of hypothyroidism and hyperthyroidism
on the hepatic transport of bilirubin mono- and diconjugates in the Wistar rat.
Hepatology 1989;9: 314-321.
399. Chavalitdhamrong P-O, Escobedo MB, Barton LL, et al. Hyperbilirubinemia

and bacterial infection in the newborn. Arch Dis Child 1975;50:652-654.
400. Linder N, Yatsiv I, Tsur M, et al. Unexplained neonatal jaundice as an early

diagnostic sign of septicemia in the newborn. J Perinatol 1988;8:325-327.
401. Rooney JC, Hill DJ, Danks DM. Jaundice associated with bacterial infection
in the newborn. Am J Dis Child 1971;122:39-41.
402. Garcia FJ, Nager AL. Jaundice as an early diagnostic sign of urinary tract
infection in infancy. Pediatrics 2002;109:846-851.
403. Maisels MJ, Newman TB. Neonatal jaundice and urinary tract infections.
Pediatrics 2003;112:1213-1214.
404. Copeland KC, Franks RC, Ramamurthy R. Neonatal hyperbilirubinemia and

hypoglycemia and congenital hypopituitarism. Clin Pediatr 1981;20:523.
405. Crabtree N, Gerrard J. Perceptive deafness associated with severe

neonatal jaundice. J Laryngol Otol 1950;64:482-506.
406. Ebbesen F. Recurrence of kernicterus in term and near-term infants in

Denmark. Acta Paediatr 2000;89:1213-1217.
407. Maisels MJ, Baltz RD, Bhutani VK, et al. Neonatal jaundice and kernicterus.
Pediatrics 2001;108:763-765.
408. Madden JM, Soumerai SB, Lieu TA, et al. Length-of-stay policies and
ascertainment of postdischarge problems in newborns. Pediatrics 2004;113:4249.
409. Varimo P, Simil S, Wendt L, et al. Frequency of breast feeding and

hyperbilirubinemia. Clin Pediatr 1986;25:112.
410. Eidelman AL, Hoffman MW, Kaitz M. Cognitive deficits in women after
childbirth. Obstet Gynecol 1993;81:764-767.
411. Newman TB, Escobar GJ, Branch PT, et al. Incidence of extreme
hyperbilirubinemia in a large HMO. Amb Child Health 1997; 3:203(abst).
412. Li R, Zhao Z, Mokdad A, et al. Prevalence of breastfeeding in the United

States: the 2001 national immunization survey. Pediatrcs 2003;111:1198-1201.
413. Ryan AS, Wenjun MS, Acosta A. Breastfeeding continues to increase into
the new millennium. Pediatrics 2002;110:1103- 1109.
414. Institue of Medicine. Crossing the quality chasm: a new health system for
the 21st century. Washington, DC: National Academy Press, 2001.
415. Palmer H, Clanton M, Ezhuthachan S, et al. Applying the 10 simple rules

of the Institute of Medicine to management of hyperbilirubinemia in newborns.
Pediatrics 2003;112:1388-1393.
416. Escobar GJ, Gonzales VM, Armstrong MA, et al. Rehospitali-zation for
neonatal dehydration: a nested case-control study. Arch Pediatr Adolesc Med
2002;156:155-161.
417. American Academy of Pediatrics and the American College of Obstetricians

and Gynecologists. Guidelines for perinatal care. 5th ed. Elk Grove Village, IL:
American Academy of Pediatrics, 2002:220-224.
418. Lawrence, RA. Management of the Mother-Infant Nursing Couple. In

Lawrence RA. Breast. A guide for the medical profession. ST. Louis: Mosby
1994:215-277.
419. De Carvalho M, Hall M, Harvey D. Effects of water supplementation on

physiological jaundice in breast fed babies. Arch Dis Child 1981;56:568-569.
420. Nicoll A, Ginsburg R, Tripp JH. Supplementary feeding and jaundice in

newborns. Acta Pediatr Scand 1982;71:759-761.
421. Butler DA, MacMillan JP. Relationship of breast feeding and weight loss to
jaundice in the newborn period: review of the literature and results of a study.
Cleve Clin Q 1983;50:263-268.
422. Maisels MJ, Gifford K. Breast feeding, weight loss and jaundice. J Pediatr
1983;102:117-118.
423. Laing IA, Wong CM. Hypernatraemia in the first few days: is the incidence
rising? Arch Dis Child Neonatal Ed 2002;87:F158-F162.
424. Kuhr M, Paneth N. Feeding practices and early neonatal jaundice. J Pediatr
425. Madlon-Kay DJ. Recognition of the presence and severity of newborn

jaundice by parents, nurses, physicians, and icterometer. Pediatrics 1997;100:
e3.
426. Tayaba R, Gribetz D, Gribetz I, et al. Noninvasive estimation of serum

bilirubin. Pediatrics 1998;102:e28.
427. Ebbesen F. The relationship between the cephalo-pedal progress of clinical

icterus and the serum bilirubin concentration in newborn infants without blood
type sensitization. Acta Obstet Gynaecol Scand 1975;54:329-332.
428. Kramer LI. Advancement of dermal icterus in the jaundiced newborn. Am J

Dis Child 1969;118:454-458.
429. Hegyi T, Hiatt M, Gertner I, et al. Transcutaneous bilirubinometry: the

cephalocaudal progression of dermal icterus. Am J Dis Child 1981;135:547.
430. Knudsen A. The cephalocaudal progression of jaundice in newborns in

relation to the transfer of bilirubin from plasma to skin. Early Hum Dev
1990;22:23-28.
431. Knudsen A. The influence of the reserve albumin concentration and pH on

the cephalocaudal progression of jaundice in newborns. Early Hum Dev 1991;25
(37).
432. Knudsen A, Broderson R. Skin colour and bilirubin in neonates. Arch Dis
Child 1989;64:605.
433. Jacobsen J, Brodersen R. Albumin-bilirubin binding mechanism: kinetic and

spectroscopic studies of binding of albumin and xanthobilirubic acid to human
serum albumin. J Biol Chem 1983;10:6319.
434. Moyer VA, Ahn C, Sneed S. Accuracy of clinical judgment in neonatal

jaundice. Arch Pediatr Adolesc Med 2000;154:391-394.
435. Riskin A, Kuglman A, Abend-Weinger M, et al. In the eye of the beholder:

how accurate is clinical estimation of jaundice in newborns? Acta Paediatr
2003;92:574-576.
436. Madlon-Kay DJ. Home health nurse clinical assessment of neonatal

jaundice. Arch Pediatr Adolesc Med 2001;155:583-586.
437. Bhutani V, Gourley GR, Adler S, et al. Noninvasive measurement of total

serum bilirubin in a multiracial predischarge newborn population to assess the
risk of severe hyperbilirubinemia. Pediatrics 2000;106:e17.
438. Yasuda S, Itoh S, Isobe K, et al. New transcutaneous jaundice device with
two optical paths. J Perinat Med 2003;31:81-88.
P.844
439. Schumacher RE, Thornbery J, Gutcher GR. Transcutaneous
bilirubinometry: A comparison of old and new methods. Pediatrics 1985;76:1014.
440. Madlon-Kay DJ. Maternal assessment of neonatal jaundice after hospital

discharge. J Fam Prac 2002;51:445-448.
441. Rubaltelli FF, Gourley G. R., Loskamp N, et al. Transcutaneous bilirubin

measurement: a multicenter evaluation of a new device. Pediatrics
2001;107:1264-1271.
442. Ebbesen F, Rasmussen LM, Wimberley PD. A new transcutaneous

bilirubinometer, bilicheck, used in the neonatal intensive care unit and the
maternity ward. Acta Paediatr 2002;91:203-211.
443. Knupfer M, Pulzer F, Braun L, et al. Transcutaneous bilirubinometry in

preterm infants. Acta Paediatr 2001;90:899-903.
444. Tan KL, Dong F. Transcutaneous bilirubinometry during and after

phototherapy. Acta Paediatr 2003;92:327-331.
445. Schumacher R. Transcutaneous bilirubinometry and diagnostic tests: the

right job for the tool. Pediatrics 2002;110:407-408.
446. Sykes E, Epstein E. Laboratory measurement of bilirubin. Clin Perinatol

1990;17:397.
447. Doumas BT, Wuu TW. The measurement of bilirubin fractions in serum.
Crit Rev Lab Sci 1991;29:415-445.
448. Doumas BT, Kwok-Cheung PP, Perry BW, et al. Candidate reference
method for determination of total bilirubin in serum: development and
validation. Clin Chem 1985;31:1779-1789.
449. Watkinson LR, St John A, Penberthy LA. Investigation into paediatric

bilirubin analyses in Australia and New Zealand. J Clin Pathol 1982;35:52-58.
450. Newman TB, Hope S, Stevenson D. Direct bilirubin measurements in
jaundiced term newborns: a re-evaluation. Am J Dis Child 1991;145:1305-1309.
451. Eidelman AI, Schimmel, Algur N, et al. Capillary and venous bilirubin
values: they are differentand how! Am J Dis Child 1989;143:642.
452. Leslie GI, Philips JB, Cassady G. Capillary and venous bilirubin values: are
they really different? Am J Dis Child 1987;141: 1199-1200.
453. Sykes E, Maisels MJ, Kusack S. the effect of ambient light on serum
bilirubin levels. Pediatr Res 1971;29:326A.
454. Newman TB, Liljestrand P, Escobar GJ. Jaundice noted in the first 24 hours
after birth in a managed care organization. Arch Pediatr Adolesc Med
2002;156:1244-1250.
455. Javier MC, Krauss A, Nesin M. Corrected end-tidal carbon monoxide closely
correlates with the corrected reticulocyte count in coombs' test-positive term
neonates. Pediatrics 2003;112: 1333-1337.
456. Davenport M, Kerkar N., Mieli-Vergani G, et al. Biliary atresia: the King's
College Hospital experience (1974-1995). J Pediatr Surg 1997;32:479-485.
457. Alpay F, Sarici S, Tosuncuk HD, et al. The value of first-day bilirubin
measurement in predicting the development of significant hyperbilirubinemia in
healthy term newborns. Pediatrics 2000;106:E16.
458. Carbonell X, Botet F, Figueras J, et al. Prediction of hyperbilirubinaemia in

the healthy term newborn. Acta Paediatr 2001;90: 166-170.
459. Kaplan M, Hammerman C, Feldman R, et al. Predischarge bilirubin

screening in glucose-6-phosphate dehydrogenase-deficient neonates. Pediatrics
2000;105:533-537.
460. Stevenson DK, Vreman HJ. Carbon monoxide production in neonates.

Pediatrics 1997;100:252-254.
461. Herschel M, Karrison T, Wen M, et al. Evaluation of the direct antiglobulin
(Coombs') test for identifying newborns at risk for hemolysis as determined by

end-tidal carbon monoxide concentration (ETCOc); and comparison of the
Coombs' test with ETCOc for detecting significant jaundice. J Perinatol
2002;22:341-347.
462. Martinez JC, Maisels MJ, Otheguy L, et al. Hyperbilirubinemia in the breastfed newborn: a controlled trial of four interventions. Pediatrics 1993;91:470473.
463. Maisels M. J. Phototherapytraditional and nontraditional. J Perinatol

2001;21:S93-S97.
464. Stern L, Denton RL. Kernicterus in small, premature infants. Pediatrics

1965;35:486-485.
465. Maisels M. J., Kring EA, Shumer D. What is the contribution of hemolysis to
jaundice in the normal newborn? Pediatr Res 2002;51:329A.
466. Newman TB, Easterling MJ. Yield of reticulocyte counts and blood smears
in term infants. Clin Pediatr 1994;33:71-76.
467. Maisels MJ, Gifford K. Neonatal jaundice in full-term infants: Role of

breastfeeding and other causes. Am J Dis Child 1983;137:561-562.
468. Vreman HJ, Wong RJ, Stevenson DK. Light-emitting diodes: A novel light
source for phototherapy. Pediatr Res 1998;44: 804-809.
469. Rubo J, Albrecht K, Lasch P, et al. High-dose intravenous immune globulin

therapy for hyperbilirubinemia caused by Rh hemolytic disease. J Pediatr
1992;121:93-97.
470. Maisels MJ, Vain N, Acquavita AM, et al. The effect of breast-feeding
frequency on serum bilirubin levelsa randomized controlled trial. Am J Obstet
Gynecol 1994;170:880-883.
471. Amato M, Howald H, von Muralt G. Interruption of breast-feeding vs.

phototherapy as treatment of hyperbilirubinemia in full term infants. Helv
Paediatr Acta 1985;40:127-131.
472. Maisels MJ. Why use homeopathic doses of phototherapy? Pediatrics

1996;98:283-287.
473. Maisels MJ, Watchko JF. Treatment of jaundice in low birthweight infants.
Arch Dis Child Fetal Neonatol Ed 2003;88:F459-F463.
474. Clark CF, Torii S, Hamamoto Y, et al. The bronze baby syndrome:
postmortem data. J Pediatr 1976;88:461-464.
475. Ebbesen F. Low reserve albumin for binding of bilirubin in neonates with
deficiency of bilirubin excretion and bronze baby syndrome. Acta Paediatr Scand
1982;71:415-410.
476. Catterton Z, Carp W, Bunyaten C, et al. Bilirubin binding capacity in ABO

hemolytic disease of the newborn. Clin Res 1979;27:817A.
477. Phibbs RH, Johnson P, Tooley WH. Cardiorespiratory status of

erythroblastotic newborn infants. II. Blood volume, hematocrit, and serum
albumin concentration in relation to hydrops fetalis. Pediatrics 1974;53:13-23.
478. Nicolaides KH, Clewell WH, Rodeck CH. Measurement of human

fetoplacental blood volume in erythroblastosis fetalis. Am J Obstet Gynecol
1987;157:50-53.
479. Jhrig K, Jhrig D, Meisel P, eds. Phototherapy: treating neonatal jaundice

with visible light. Mnchen: Quintessenz Verlags-GmbH, 1998.
480. Maisels MJ. Phototherapy. In: Maisels MJ, Watchko JF, eds. Neonatal
jaundice. London, UK: Harwood Academic Publishers, 2000:177-204.
481. Agati G, Fusi F, Donzelli GP, et al. Quantum yield and skin filtering effects
on the formation rate of lumirubin. J Photochem Photobiol B 1993;18(2-3):197203.
Chapter 36
Calcium and Magnesium Homeostasis
Winston W. K. Koo
Reginald C. Tsang
Calcium (Ca) is the most abundant mineral in the body and, together with phosphorus (P),
forms the major inorganic constituent of bone. Magnesium (Mg) is the fourth most
abundant cation and is the second most common intracellular electrolyte in the body. The
maintenance of Ca and Mg homeostasis requires a complex interaction of hormonal and
nonhormonal factors; adequate functioning of various body systems, particularly the renal,
gastrointestinal, and skeletal systems; and adequate dietary intake. From a clinical
perspective, mineral homeostasis is reflected in the maintenance of circulating
concentrations of Ca and Mg in the normal range and integrity of the skeleton.
In the circulation, the amount of Ca and Mg is less than 1% of their respective total body
content; however, disturbances in serum concentrations of these minerals are associated
with disturbances of physiologic function manifested by numerous clinical symptoms and
signs. Chronic and severely lowered serum concentrations of these minerals also may
reflect the presence of a deficiency state.
At all ages, the total body content of Ca and Mg in the skeleton are about 99% and 60%,
respectively. Thus, the skeleton is a reservoir for mineral homeostasis in addition to its role
providing structural and mechanical support; disturbances in mineral homeostasis can
result in osteopenia and rickets in infants and children, and osteomalacia and osteoporosis
in adults.
The mechanisms to maintain mineral homeostasis in neonates are the same as for children
and adults. However, the newborn infant has a number of unique challenges to maintain
mineral homeostasis during adaptation to extrauterine life and to continue a rapid rate of
growth. These include an abrupt discontinuation of the high rate of intrauterine accretion of
Ca (approximately 120 mg/kg/day) and Mg (approximately 4 mg/kg/day) during the third
trimester; a smaller skeletal reservoir available for mineral homeostasis; a delay in
establishment of adequate nutrient intake for at least a few days or longer, particularly in
sick and preterm infants; and a high requirement for Ca and Mg for the most rapid period
of postnatal skeletal growth, with an average gain in length of more than 25 cm during the
first year. There also may be diminished end-organ responsiveness to hormonal regulation
of mineral homeostasis, although the functional capacity of the gut and kidney improves
rapidly within days after birth. The effects of these issues are exaggerated in infants with
heritable disorders of mineral metabolism, such as extracellular calcium-sensing receptor
(CaR) mutations, and in infants who have experienced adverse antenatal events such as
maternal diabetes, intrapartum problems such as perinatal asphyxia or maternal Mg
therapy, or postnatal problems such as immature functioning of multiple organs from
premature birth.
Increased understanding of the physiology and molecular basis of mineral metabolism
allows a better understanding of the pathophysiology of the resultant clinical disorder. This
in turn allows a more rational management to minimize the adverse impact from disturbed
mineral homeostasis and to prevent iatrogenic causes precipitating or prolonging these
problems.
TISSUE DISTRIBUTION
In the fetus, about 80% of minerals are accrued between 25 weeks of gestation and term.
During this period, the estimated daily accretion per kilogram of fetal body weight is 2.3 to
2.98 mmol (92 to 119 mg) Ca and 0.1 to 0.14 mmol (2.51 to 3.44 mg) Mg. The peak
accretion rates occur at 36 to 38 weeks of gestation. In newborn term infants, the total
body Ca and Mg contents average approximately 28 grams and 0.7 grams, respectively
(1,2).
After birth, 99% of total body Ca is in bone. The tissue distribution of Mg varies according
to the extent of bone mineralization and rate of soft tissue growth. Near the end of the
third trimester, however, about 60% of the body's Mg is in bone, 20% in muscle, and most
of the remainder is found in the intracellular space of other tissues.
P.848
CIRCULATING CONCENTRATION
Calcium
Serum Ca (1 mmol/L =4 mg/dL) occurs in three forms: approximately 40% is bound,
predominantly to albumin; approximately 10% is chelated and complexed to small
molecules such as bicarbonate, phosphate, or citrate; and approximately 50% is ionized.
Complexed and ionized Ca are ultrafiltrable.
Total Ca concentrations (tCa) in cord sera increase with increasing gestational age. Serum
tCa may be as high as 3 mmol/L in cord blood of infants born at term, and the
concentrations are significantly higher than paired maternal values at delivery (3,4,5,6).
Serum tCa reaches a nadir during the first 2 days after birth (7,8,9,10,11,12,13);
thereafter, concentrations increase and stabilize at a level generally above 2 mmol/L (14).
In infants exclusively fed human milk, the mean serum tCa increases from 2.3 to 2.7 mmol/
L over the first 6 months postnatally. Normally serum tCa in children and adults remains
stable, with a diurnal range of less than 0.13 mmol/L. During the third trimester of
pregnancy, a modest reduction in maternal serum tCa concentration (average 0.1 mmol/L)
is associated with a decrease in serum albumin concentration.
Serum ionized Ca (iCa) concentration is the best indicator of physiologic blood Ca activity.
Measurement of serum iCa is firmly established in clinical medicine, and simple, rapid, and
direct determination of iCa from whole blood, plasma, and serum by ion-selective
electrodes is freely available. Whole-blood iCa analyzers are gaining popularity because
they are adaptable for point of care testing. However, some differences exist in values
from different iCa analyzers, particularly for whole-blood iCa values (15), as a result of
differences in the design of the reference electrode, formulation of calibrating solutions,
and lack of a reference system for iCa. Thus, normative data should be generated
according to the subject's age, the instrument, and the type of sample used for iCa
measurement.
Serum iCa decreases in the presence of high serum albumin, P, bicarbonate, and heparin.
Serum iCa increases with increased Mg and is inversely related to blood pH. The effect of
the latter may be minimized by the immediate analysis of serum samples for iCa. Freezing
serum samples in 5% carbon dioxide-containing tubes may minimize the impact of pH
variations if measurement of iCa is delayed for 1 week.
One report showed a wide range of cord whole-blood iCa of 0.4 to 1.85 mmol/L from
apparently normal pregnancies (16). This is a much wider range compared with multiple
reports based on cord sera, although the range for whole-blood iCa becomes much
narrower within hours after birth and similar to serum iCa values. Cord-serum iCa
increases with increasing gestational age and is higher than values in paired maternal sera.
In healthy term neonates, serum iCa averages 1.25 mmol/L with 95% confidence limits of
1.1 to 1.4 mmol/L (4.4 to 5.6 mg/dL), and there is a decline in serum iCa in the first 48
hours of life with a nadir at 24 hours (17). Serum iCa generally changes in parallel with tCa
in healthy humans. However, serum iCa is stable and normal during pregnancy in contrast
to a slight reduction in tCa. Serum tCa and iCa are correlated in infants and adults, but the
correlation is inadequate to predict one from the other with sufficient accuracy.
The concentration of iCa is critical to many important biologic functions with the
extracellular Ca concentration normally maintained within a narrow range. The Ca ion is
well established as an intracellular second messenger, but identification of the extracellular
CaR has established that iCa also functions as a messenger outside cells. Ca homeostasis
also involves the maintenance of an extremely large Ca concentration gradient across the
cellular plasma membrane. In the cell, distribution of Ca is not uniform. The cytosolic
compartment contains 50 to 150 nmol of Ca per liter of water; a larger intramitochondrial
Ca pool contains 500 to 10,000 nmol of Ca per liter of cell water. In contrast, the
concentration of iCa in extracellular fluid is 1 million nmol/L (1 mmol/L). There are at least
two adenosine triphosphate-dependent mechanisms involved in the maintenance of the Ca
concentration gradient across the plasma membrane. The measurement of intracellular Ca
continues to improve with better instrumentation and probes, but it is not yet freely
available.
Magnesium
Approximately 30% of serum Mg (1 mmol/L =2.4 mg/dL) is in the protein-bound form,
with the remainder in the ultrafiltrable portion. Seventy percent to 80% of ultrafiltrable Mg
is in ionic form, and the remainder is complexed to anions, particularly phosphate, citrate,
and oxalate. Cord-serum total Mg (tMg) is higher than paired maternal values. Serum tMg
of 0.92 0.13 mmol/L (mean 2 SD) in children is slightly higher than the adult values of
0.88 0.13 mmol/L (18). Ion-selective electrodes are used in the measurement of ionized
Mg (iMg) in whole blood and sera. iMg concentrations average 62% to 70% of the tMg
concentration in cord and postnatal sera. Cord-serum iMg is also higher than that in
maternal serum (19,20,21). The clinical role of iMg (versus tMg) in a number of disease
states appears limited (22).
Cellular Mg content of most tissues is 6 to 9 mmol/kg wet weight, and most of this Mg is
localized in membrane structures (e.g., microsome, mitochondria, plasma membrane). The
much smaller pool of free Mg in the cell is maintained at about 1 mmol/L and is in an
exchanging equilibrium with membrane-bound Mg. Unbound intracellular Mg has a critical

role in cellular physiology and catalyzes enzymatic processes concerned with the transfer,
storage, and use of energy. Intracellular Mg usually remains stable despite wide
fluctuations in serum Mg. In Mg-deficient states, however, the intracellular content of Mg
can be low despite normal serum concentrations. The measurement of intracellular Mg
continues to improve with better instrumentation and probes but is not freely available.
P.849
PHYSIOLOGIC CONTROL
Calciotropic hormones, including parathyroid hormone (PTH), 1,25 dihydroxyvitamin D
[1,25(OH)2D], and possibly calcitonin (CT), appear to maintain Ca homeostasis by
intermodulation of their physiologic effects on each other and on the classic target organs:
kidney, intestine, and bone. Dietary intake of Ca, Mg, P, and other nutrients including
sodium, glucose, and protein also may significantly contribute to the regulation of mineral
homeostasis. PTH serves as the major component of the rapid response to hypocalcemia,
whereas 1,25(OH)2D, with its major effect on elevating intestinal absorption of Ca, is
responsible for a slower but more sustained contribution to the maintenance of
normocalcemia. CT, on the other hand, appears to function in the opposite role to PTH but
with the capacity to stimulate the production of 1,25(OH)2D, which in theory may serve an
additional regulatory role in the maintenance of Ca homeostasis.
In contrast, the control of Mg homeostasis by calciotropic hormones under physiologic
conditions appears to be limited. However, Mg is critical to the maintenance of Ca
homeostasis since Mg regulates the production and secretion of PTH, acts as a cofactor for
the 25-hydroxyvitamin D 1 hydroxylase enzyme in the production of 1,25(OH)2D, and
maintains adequate sensitivity of target tissues to PTH. Furthermore, Mg is considered a
mimic/antagonist of Ca as it often functions synergistically with Ca, yet competes with it in
the gut and kidney for transport and other metabolic pathways.
Parathyroid Hormone
In humans, parathyroid glands are derived from the third and fourth pharyngeal pouch.
The PTH gene, along with the genes for insulin,-globulin, and CT, is located on
chromosome 11p15 (23), and restriction site polymorphisms near the PTH gene have been
detected. The initial translational product of the mRNA is a 115-amino-acid prepro-PTH.
Prepro-PTH then undergoes proteolytic cleavage in the endoplasmic reticulum to remove a
25-residue amino-terminal signal sequence to form pro-PTH. The prohormone-specific
region is cleaved further during subsequent intracellular processing to generate the 84amino-acid secreted form of the intact hormone with a relative molecular mass (Mr) of
9,500. PTH is synthesized by the chief cells and stored in secretory granules. It is colocated
and secreted with chromogranin A, a protein that may act in autocrine- or paracrineregulated release of PTH.
About 50% of the newly generated PTH is proteolytically degraded intracellularly, and
some of the inactive fragments are also secreted. After release into the circulation, the
intact PTH molecule has a serum half-life of 5 to 8 minutes and undergoes a series of
cleavages by endopeptidases in the liver and kidney. The amino-terminal fragments

contain the biologically active fractions, with the 1-34 fragment having the most calcemic
activity; modifications at the amino terminal, particularly at the first two residues, can
abolish its biologic activity. The midregion and carboxyl-terminal fragments are biologically
inert, although the latter may have some in vitro biological activity.
Circulating immunoreactive PTH is a complex mixture of intact 1-84 PTH, multiple peptide
fragments from the amino- and carboxyl-terminals, and mid-molecular regions. Normally
there are greater amounts of middle and carboxyl fragments than intact hormone in the
circulation because of metabolic breakdown of the short-lived, intact hormone, coupled
with glandular secretion of inactive fragments. The fragments are cleared from the blood
virtually exclusively by glomerular filtration. Intact PTH and amino-terminal fragments
constitute less than 20% of PTH immunoreactivity in the peripheral circulation. PTH
molecules reactive in the widely used commercial immunoradiometric assays (IRMA),
designed to detect both amino- and carboxyl-terminal epitopes of the peptide, have been
considered as intact PTH (IPTH). However, there have been reports that the large 7-84
fragment of PTH is also detected by these assays. This large fragment is biologically
inactive and present in greater concentrations in uremic states or hyperparathyroidism.
The conventional IPTH technique increases the PTH concentration by 30% to 50%,
compared to the latest chemiluminescence or IRMA techniques that measure the whole
or bio-intact PTH. Therefore, the treatment of secondary hyperparathyroidism based on
data from conventional IPTH assays theoretically may lead to overtreatment and
oversuppression of biologically active PTH, although the clinical significance of this
possibility remains to be defined. In any case, consistency of the PTH assay methodology
and serial measurements are critical to the interpretation and management of pathologic
states.
PTH concentrations in cord blood frequently are low and do not correlate with PTH
concentrations in maternal sera (4,24). Earlier studies of higher levels of bioactive PTH in
cord sera from cytochemical assays may be related to elevated concentrations of
parathyroid hormone-related protein (PTHrP), since PTH-like bioactivity was tightly
correlated with levels of PTHrP in sheep and pig. Small amounts (approximately 5%) of
perfused fragments (35-84, 44-68, and 65-84 amino acids) but probably not the whole
PTH molecule are reported to cross the human placenta. Serum PTH concentrations
increase postnatally coincidentally with the fall in serum Ca in both term and preterm
infants (4,24-27). The rise in serum IPTH is greater for preterm infants with hypocalcemia
compared to term infants reflecting appropriate PTH response. Serum PTH concentrations
are similar for children and adults but are increased in the elderly. Serum concentrations of
intact PTH as measured by IRMA showed no change during normal pregnancy. In adults,
serum intact PTH is present in picomolar concentrations. It has a significant circadian
periodicity, spontaneous episodic pulsatility with distinct peak property and a significant
temporal coupling with serum iCa and P concentrations and prolactin secretion.
P.850
PTH effects on end-organ systems appear to be mediated through its binding to specific
receptors. The type 1 PTH/PTHrP receptor has been identified in bone, cartilage, kidney,
intestine, aorta, urinary bladder, adrenal gland, brain, and skeletal muscle. It binds equally
to PTH and PTHrP and belongs to a superfamily of guanine-nucleotide-binding (G) proteincoupled cell membrane receptors (GPCR) including those for CT, secretin, growth hormone-
releasing hormone, corticotrophin-releasing hormone, glucagon, vasoactive intestinal

polypeptide, and others. Another PTH receptor (type 2) responds only to PTH, although its
main endogenous ligand appears to be a 39-amino-acid peptide, hypothalamic tubular
infundibular peptide. It has been found in the brain, pancreas, and intestines, but the
physiologic significance of this receptor remains ill defined.
The gene for the PTH/PTHrP receptor is located on chromosome 3p21.1-p24.2. It contains
17 exons and encodes a mature glycoprotein of 593 amino acids (28). The type 1 PTH
receptor consists of extended extracellular, ligand-binding amino-terminal and intracellular
G-protein-associated carboxyl-terminal domains and seven transmembrane domains.
Signal transduction mediated by G proteins results in multiple second messenger pathways
to affect both stimulatory and inhibitory end-organ responses. The strongest and bestcharacterized second messenger signaling pathway is the PTH-stimulated coupling of the
type 1 PTH receptor to the stimulating G protein (Gs; composed of 3 subunits: , , and ,
and encoded by the GNAS1 gene localized to 20q13.3), which activates adenyl cyclase, an
enzyme that generates cyclic AMP (cAMP). However, coupling of the type 1 PTH receptor to
the Gq class protein activates phospholipase C, which generates inositol phosphate (IP3)
and diacylglycerol (DAG). These second messengers in turn lead to stimulation of protein
kinases A and C and Ca transport channels and result in a variety of hormone-specific
tissue responses.
In physiologic terms, PTH acts synergistically with 1,25(OH)2D and is the most important
regulator of extracellular Ca concentration. PTH acts directly on bone and kidney, and
indirectly on intestine. Immediate control of blood Ca is probably due to PTH-induced
mobilization of Ca from bone and increased renal distal tubular reabsorption of Ca. PTH
also decreases proximal tubular and thick ascending limb reabsorption of sodium, Ca,
phosphate, and bicarbonate. PTH effects on kidney are mediated primarily through
stimulation of sodium/calcium exchange, calcium transport proteins and renal 25 OHD-1hydroxylase, but a decrease in sodium-dependent phosphate cotransporter, NPT-2.
Maintenance of steady-state Ca balance is probably from increased intestinal Ca absorption
secondary to increased 1,25(OH)2D production. PTH increases acutely within minutes the
rate of Ca release from bone into blood. Chronic effects of PTH are to increase the numbers
of osteoblasts and osteoclasts and to increase the remodeling of bone. Continuous
exposure to elevated concentrations of PTH leads to increased osteoclastic resorption of
bone. In contrast to its classic action on Ca mobilization from bone, the amino-terminal
fragments of PTH and PTHrP, and small pulses of PTH have an anabolic effect on bone,
independent of its resorptive action. Other PTH effects on bone include enhanced collagen
synthesis, activities of alkaline phosphatase, ornithine and citrate decarboxylases, and
glucose-6-phosphate dehydrogenase; DNA, protein and phospholipid synthesis.
Extracellular Ca is the most potent regulator of PTH secretion and is mediated by the cellsurface CaR, which detects minute perturbations in the extracellular iCa concentration and
responds with alterations in cellular function that normalize iCa. Thus, iCa functions as
extracellular as well as intracellular messenger. The human CaR gene is located on
chromosome 3q13.3-q21 and encodes a cell-surface protein of 1,078 amino acids. The CaR
gene is developmentally upregulated, and CaR transcripts are present in numerous tissues
including chief cells of the parathyroid glands, kidneys (in particular the thick ascending
limb), brain and nerve terminals, breast, lung, intestine, adrenal and skin, and also the
precursor and mature osteoblasts and osteoclasts. CaR is a member of the GPCR
superfamily with a seven-member membrane-spanning domain. It contains at least seven
exons, of which six encode the large (approximately 600-amino-acid) amino-terminal
extracellular domain and/or its upstream untranslated regions, while a single exon codes
for the remainder of the receptor including a cytoplasmic carboxyl-terminal intracellular
domain. Signal transduction mediated by G proteins results in activation of phospholipase
C that generates IP3 and DAG, and subsequent stimulation of protein kinase C (PKC) and
Ca transport channels.
Low or falling serum Ca concentrations result in active secretion of preformed PTH within
seconds. There is a sigmoidal type of PTH secretion in response to decreased serum Ca,
which is most pronounced when serum Ca is in the mildly hypocalcemic range. PTH
secretion is 50% of maximal at a serum iCa of about 1 mmol/L (4 mg/dL); this is
considered the calcium set point for PTH secretion (29). Sustained hypocalcemia increases
PTH mRNA within hours. Protracted hypocalcemia leads within days to cellular replication
and increased gland mass. High serum Ca suppresses PTH secretion via activation of CaR.
It in turn activates phospholipase C and generation of IP3 and DAG and probably increases
the proteolytic destruction of preformed PTH. Hyperphosphatemia stimulates PTH
secretion, probably by lowering the serum Ca concentration.
In the kidney, CaR decreases the basal and PTH-stimulated paracellular reabsorption of Ca,
Mg, and sodium via multiple mechanisms including inhibition of cAMP accumulation;
stimulation of phospholipase A2 activity. The release of free arachidonic acid is promoted
which is metabolized via the lipooxygenase pathway to P450 metabolites that inhibit the
activities of the sodium-potassium-chloride cotransporter and potassium channel; and
possibly affect renal water regulation by inhibition of vasopressin-abated water flow. In
chronic renal failure, downregulation in the
P.851
expression of renal CaR may account for the development of secondary
hyperparathyroidism (30), and downregulation of PTH receptors may account for the
skeletal resistance to the calcemic effect of PTH (31). Extracellular Ca exerts numerous
other actions on parathyroid function, including modulation of the intracellular degradation
of PTH, cellular respiration, membrane voltage, and the hexose monophosphate shunt.
Maintenance of Ca homeostasis through other organs also may be possible, for example,
through the presence of CaR in intestinal cells (32), and probable modulation of CT
secretion from changes in intracellular Ca (33). Furthermore, expression of CaR in gastrinsecreting G cells and acid-secreting parietal cells, together with data indicating that CaR
exhibits selectivity for L-aromatic amino acids, appears to provide a molecular explanation
for amino acid sensing in the gastrointestinal tract, regulation of PTH secretion and urinary
Ca excretion, and the physiologic interaction between Ca and protein metabolism.
Decrease in serum Mg concentration stimulates PTH secretion (34,35), although chronic
hypomagnesemia inhibits secretion of PTH (34,36). Hypomagnesemia is also associated
with an increased target tissue resistance to PTH probably from inactivity of adenylate
cyclase, a Mg-requiring enzyme. Hypermagnesemia rapidly decreases the secretion of PTH
in vivo in human subjects, and PTH concentration remains depressed despite concomitant
hypocalcemia, presumably in part due to stimulation of CaR by other divalent cations
such as Mg. Vitamin D and its metabolites, 25-hydroxyvitamin D (25-OHD) and 1,25(OH)
2D,
acting through vitamin D receptors, decrease the level of PTH mRNA. Additional
systemic factors (growth hormone, insulin-like growth factor 1, estrogen, progesterone,

CT, cortisol, catecholamines, prostaglandins, and somatostatin) and local factors
(interleukin-l [IL-1]) modulate PTH secretion and function, although their role in the
regulation of Ca and Mg metabolism under physiologic conditions is not clear.
Calcitonin
CT is secreted primarily from thyroid C cells and also from many extrathyroidal tissues
including placenta, brain, pituitary, mammary gland, and other tissues. Developmentally,
CT-containing cells and parathyroid gland cells are thought to derive from the same tissue
source as the neural crest. In the rat, the number of thyroid C cells and secretion of CT
increase from fetal life to suckling, a period of rapid growth (37). There is probably no
placental crossover of CT; the human placental tissue is able to produce CT in response to
the presence of Ca in the culture medium. In human neonates, the CT content in crude
tissue preparations of thyroid is larger than that of the adult thyroid (38).
There are two calcitonin genes, and , located on chromosome 11p15.2 near the genes
for -globulin and PTH. Two different RNA molecules are transcribed from the gene. It is
comprised of six exons with the fourth exon translated into the precursor for CT and the
fifth translated into the precursor for CT gene-related peptide-I (CGRP-I). The initial
translational product of the mRNA is prepro-CT, a 141-amino-acid peptide. It is cleaved by
endopeptidase at the endoplasmic reticulum to form pro-CT, a 13-kDa 116-amino-acid
peptide. CT (between the sixtieth and ninety-first positions of the pro-CT peptide) and
equimolar amounts of non-CT secretory peptides, corresponding to the flanking peptides
linked to the amino and carboxyl terminals of the prohormone, are generated during
precursor processing. Further structural modifications to the CT molecule occur
intracellularly prior to release into the circulation. These modifications include formation of
a disulfide bridge between two cysteine remnants in position 1 and 7, and hydroxylation of
the C-terminal proline residue; both are essential for binding of CT to its receptor. The CT
monomer is a 32-amino-acid peptide (Mr 3,400). CGRP-I is synthesized wherever the CT
mRNA is expressed, e.g., in medullary carcinoma of thyroid, although there is no
translational product from the CGRP-I sequence.
The or CGRP-II gene is transcribed into the mRNA for CGRP predominantly in nerve fibers
in the central and peripheral nervous systems, blood vessels, thyroid and parathyroid
glands, liver, spleen, heart, lung, and possibly bone marrow. CGRP, a 37-amino-acid
peptide (Mr 4,000), is also generated from the larger precursor molecule pro-CGRP, a 103amino-acid peptide. Seventy-five amino-terminal residues of the preprohormones for CT
and CGRP are predicted to be identical.
Classic bioactivity of human CT (hCT) is present in the full 32-amino-acid structure or its
smaller fragments, such as hCT 8-32 and hCT 9-32; the ring structure of CT enhances, but
is not essential for, hormone action. Basic amino acid substitutions confer a helical
structure in this region as found in salmon and other nonmammalian CT, resulting in
greater potency in lowering serum Ca and probably a longer circulating half life. The kidney
appears to be the dominant organ in the metabolism of human CT. A small percentage of
the metabolic clearance rate of CT in humans may be accounted for by enzymatic
degradation in blood. Injected hCT monomer disappears from the blood in vivo with a half-
life of approximately 10 minutes; in contrast, the half-life of hCT in plasma incubated in

vitro at 37C may be longer than 20 hours (39). Depending on the animal species, other
sites such as the liver, intestine, and bone may be involved in the metabolism of CT.
Circulating immunoreactive CT and CGRP are a heterogeneous mixture of different
molecular forms and are recognized as long as the antigenic epitopes recognized by the
antiserum are expressed. Immunoreactive CT or CGRP concentration is expressed in
gravimetric or molar equivalents of synthetic CT or CGRP. Sample preparation with initial
extraction, gel chromatography, and high-performance liquid chromatography separation,
and the use of two-site immunoassay can improve the sensitivity and specificity of CT
measurements. Serum CT concentrations during pregnancy are variable. They are high at
birth compared to paired maternal CT concentrations (40).
P.852
Serum CT further increases during the first few days after birth and may reach levels fiveto ten-fold higher than adult CT concentrations. Serum CT concentrations decrease
progressively during infancy; however, in preterm infants up to 3 months after birth, the
mean serum CT concentrations may remain twice the adult value. There is also a small
peak of serum CT concentration during late childhood. In human adults, the basal serum
CT concentration may be lower in women than in men, but the concentration is not
affected by increasing age. The CT secretory response to Ca infusion is lower in women
and with increasing age. In human adults, serum CT and CGRP concentrations are found in
the picomolar range. Diurnal variability has been reported for serum CGRP but not for
serum CT. In normal individuals, larger precursor molecules of CT such as procalcitonin are
not detected.
CT function is mediated by binding to receptors linked to G proteins, members of the GPCR
superfamily, and by activation of adenylate cyclase and phospholipase C (41). CT receptors
(CTR) have been identified in the central nervous system, testes, skeletal muscle,
lymphocytes, and placenta. The function of CTR can be influenced by accessory proteins,
receptor isoforms, genetic polymorphisms, developmental and/or transcriptional
regulation, feedback inhibition, and the specific cellular or tissue background. The CTR
gene is located on chromosome 7q2l.2-q21.3 and encodes a 490-amino-acid G-proteinlinked receptor with seven transmembrane domains. Two isoforms of human CTR arise by
alternative splicing of an exon of 48 nucleotides that encodes a l6-amino-acid insertion
within the first intracellular loop. The isoform with the insertion (hCTR-l) activates only
adenylate cyclase, whereas the other isoform (hCTR-2) activates both adenylate cyclase
and phospholipase C. CGRP functions are also mediated by receptors (42).
The presence of receptor-activity-modifying proteins (RAMPs) can posttranslationally
modify the initially orphan calcitonin receptor-like (CL) receptor and CTR to exhibit
different receptor functions, i.e., functional modification of G-protein-coupled receptors is
possible. RAMP1, 2, and 3 thus far identified are single transmembrane domain proteins
with intracellular C-terminals of up to 10 amino acids and extracellular N-terminals of
about 120 amino acids. Noncovalent association of a RAMP with the CL receptor or CTR
results in heterodimeric RAMP/receptor complexes at the cell surface. The CL receptor
functions as a CGRP receptor when coexpressed with RAMP1, whereas CL receptor/RAMP2
and CL receptor/RAMP3 are adrenomedullin 1 and 2 receptor subtypes, respectively. CTR,
with 60% homology to the CL receptor, predominantly recognizes CT in the absence of
RAMP. When CTR is coexpressed with RAMP1, it transforms into an amylin/ CGRP receptor.
When CTR is coexpressed with RAMP3, it interacts only with amylin. Thus, two class II Gprotein- coupled receptors, the CL receptor and CTR, are associated with three RAMPs to
form high-affinity receptors for CGRP, adrenomedullin, or amylin.
Secretion of CT is stimulated by an increase in serum Ca and Mg concentrations and by
gastrin, glucagon, and cholecystokinin (along with several other structural analogues of
these hormones, e.g., pentagastrin, prostaglandin E2), glucocorticoid, norepinephrine, and
CGRP; secretion is suppressed by hypocalcemia, propranolol and other adrenergic
antagonists, somatostatin, chromogranin A and vitamin D. CT gene transcription is
positively regulated by glucocorticoids and negatively regulated by PKC, Ca, and vitamin D.
Calcitonin may activate the l-hydroxylase system independent of PTH (43), whereas 1,25
(OH)2D decreases CT gene expression in adult rats but is ineffective in 13-day-old suckling
rats (44). The latter observation may be related to fewer 1,25(OH)2D receptors in C cells
of immature rats. Calcitonin induces refractoriness to its own actions by downregulation,
and functional reduction of receptor mRNA is a well-known phenomenon. Clinically, it is
manifested as the escape phenomenon during therapy with calcitonin.
In humans, changes in Ca (and P) metabolism are not seen despite extreme variations in
CT production. In the neonate, there is neither an identifiable hypocalcemic response to
the postnatal surge in serum CT nor a blunting of CT secretion in the presence of
hypocalcemia. In adults, there are no definite effects attributable to CT deficiency (e.g.,
totally thyroidectomized patients receiving only replacement thyroxin) or CT excess (e.g.,
patients with medullary carcinoma of thyroid), except for the chronic suppression of bone
remodeling. The clinical significance of CT is related to its use as a tumor marker in the
management of medullary carcinoma of the thyroid and its pharmacologic effect to inhibit
osteoclast-mediated bone resorption and increase renal Ca clearance. The pharmacologic
activities of CT are useful for the suppression of bone resorption in Paget disease, for
limited use in the treatment of osteoporosis, and for early phase treatment of severe
hypercalcemia. In addition, CT also increases renal clearance of Mg, P, and sodium and
free water clearance. The net effect of CT is a lowering of serum Ca and P concentrations.
Thus, the bioactivity of CT on calcium metabolism frequently is opposite that of PTH; CT
probably modulates the effect of PTH on target organs.
Noncalcium-related actions of CT and associated molecules are increasingly being
discovered. For example, CT and CTR may play important roles in a variety of processes as
wide ranging as embryonic development and sperm function/physiology. In addition, proCT detectable in the plasma is not produced in the C cells of the thyroid and is being
explored as a marker of bacterial-induced inflammation/sepsis. Production of pro-CT after
exposure to bacterial endotoxin and inflammatory cytokines tumor necrosis factor (TNF)
and IL-6 appears to be primarily from neuroendocrine cells in the lung and intestine. Cells
of neuroendocrine origin express all proteins related to CT (CGRP-I and -II and amylin)
derived from the same family of genes, and it is speculated that inflammatory pro-CT
may be coded by the same gene family. There are no enzymes in the plasma that can
break down pro-CT, and when it is secreted into the circulation, it has a half-life of 25 to
30 hours, thus increasing serum pro-CT. After
P.853
administration of endotoxin, the peak circulating concentrations of TNF, IL-6, pro-CT, and
C-reactive protein occur at approximately 90 minutes, 180 minutes, 6 to 8 hours, and 24
hours, respectively.
CGRP primarily affects catecholamine release, vascular tone and blood pressure, and
cardiac contractility. Its clinical role probably also lies in its potential pharmacologic effect.
The influence of CGRP on Ca and P homeostasis is minor compared to that of CT. However,
amylin, a pancreatic islet-derived or synthetic 37-amino-acid peptide, is a member of the
CGRP family with a potent hypocalcemic effect despite sharing only 15% of its amino acid
sequence with human CT. The hypocalcemic effect of amylin is probably mediated by the
CTR on osteoclasts, and it is 100-fold more potent than CGRP (45). Both CT and CGRP
inhibit gastric acid secretion and food intake.
Vitamin D
Vitamin D (Mr 384) can be obtained from diet or synthesized endogenously. It must
undergo several metabolic transformations primarily in the liver and kidney to form the
physiologically most important metabolite, 1,25(OH)2D, which functions as a hormone in
the maintenance of mineral homeostasis. Under in vivo conditions, there are more than 30
other vitamin D metabolites, with and without putative functions.
Dietary vitamin D (1 g =40 IU) is derived from plants as ergocalciferol (vitamin D2) and
from animals as cholecalciferol (vitamin D3). Dietary vitamin D is absorbed from the
duodenum and jejunum into lymphatics, and about 50% of the vitamin D in chylomicron is
transferred to vitamin D-binding protein (DBP) in blood before uptake by the liver.
In animals, vitamin D3 can be synthesized endogenously in the skin (46). During exposure
to sunlight, the high-energy UV photons (290 to 315 nm) penetrate the epidermis and
photochemically cleave the bond between carbons 9 and 10 of the sterol B-ring of 7dehydrocholesterol (7-DHC or provitamin D3) to produce previtamin D3. It then undergoes
a thermally induced isomerization to vitamin D3 that takes 2 to 3 days to reach completion.
Thus, cutaneous synthesis of vitamin D3 continues for many hours after a single sun
exposure. Previtamin D3 is photolabile; continued exposure to sunlight causes the
isomerization of previtamin D3 to biologically inert products, principally lumisterol. No more
than 10% to 20% of the initial provitamin D3 concentrations ultimately end up as
previtamin D3, thus preventing excessive production of previtamin D3 and vitamin D3.
Vitamin D3 synthesis in the skin is directly dependent on the amount of sunlight exposure
and is affected by time of day, season, and latitude. Peak sunlight at midday, in summer,
and at lower latitudes are optimal conditions; the amount of skin area exposed and
duration of sunlight exposure directly affect vitamin D3 synthesis. Melanin in the skin
competes with 7-DHC for ultraviolet photons, but the production of vitamin D3 can be
adjusted by increasing the duration of sunlight exposure. Use of topical sunscreen blocks
ultraviolet photons, and aging decreases the capacity for cutaneous synthesis of vitamin D3.
The term vitamin D is frequently used generically to describe vitamins D2 and D3 and
their metabolites. In mammals, vitamins D2 and D3 appear to metabolize along the same
pathway, and there is little functional difference between their metabolites. However,
differences in affinity to DBP and receptors between D2 and D3 and their metabolites
support the contention that vitamin D3 is more bioavailable than D2.
In the circulation, vitamin D and its metabolites are bound to proteins, mainly DBP
(approximately 85%) and albumin (approximately 15%). The DBP gene is located on
chromosome 4q11-13. It is a member of the albumin multigene family of proteins that
includes albumin and -fetoprotein. DBP is an approximately 53-kDa globulin in humans,
and its x-ray crystallographic structure has been determined (47). Plasma DBP
concentration (4 to 8 M) is approximately 20-fold higher than that of the total circulating
vitamin D metabolites (approximately 100 nM), i.e., it is normally less than 5% saturated
with vitamin D metabolite. The amount of unbound or free 25-OHD and 1,25(OH)2D,
important in determining bioactivity, is less than 1% of the total concentration.
In the liver, vitamin D is hydroxylated at carbon 25 to 25-OHD. Quantitatively, 25-OHD (1
nmol/L =0.4 ng/mL) is the most abundant vitamin D metabolite in the circulation and is a
useful index of vitamin D reserve. Regulation of 25-hydroxylase activity is limited, and
there are few limitations to the production of 25-OHD. However, in vivo administration of
1,25(OH)2D (48) inhibits hepatic production of 25-OHD, and Ca deficiency (49) increase
the metabolic clearance of 25-OHD with subsequently decrease circulating 25-OHD.
In the kidney, 25-OHD is hydroxylated further to 1,25(OH)2D by 25-OHD-1-hydroxylase
(CYP1), and to 24,25-dihydroxyvitamin D (24,25(OH)2D) by 25-OHD-24-hydroxylase
(CYP24). The hydroxylation occurs primarily in the mitochondria of renal proximal tubules.
The genes for these enzymes have been localized to chromosome 12q13-14 and 20q13.3,
respectively. The human gene encoding CYP1 is 5 kb in length and is comprised of nine
exons and eight introns; its exon/intron organization is similar to other cloned
mitochondrial P450 enzymes.
The activity of CYP1 and therefore production of 1,25(OH)2D are tightly regulated. It is
the rate-limiting hormonally regulated step in the bioactivation of vitamin D. PTH increases
transcriptional activity of the CYP1 gene promoter and therefore increases mRNA for 1,25
(OH)2D. Decreases in serum or dietary Ca or P increase mRNA for and serum concentration
of 1,25(OH)2D independent of PTH (49,50,51). However, hypophosphatemia in renal
wasting disorders does not elicit appropriate phosphate conservation or an increase in 1,25
(OH)2D production. These disorders include X-linked hypophosphatemic rickets (XLH),
autosomal-dominant hypophosphatemic rickets (ADHR), and tumor-induced osteomalacia.
They have
P.854
similar phenotypic manifestations characterized by hypophosphatemia, decreased renal
phosphate reabsorption, normal (and thus inappropriately low) or low serum calcitriol
concentrations, normal serum Ca and PTH, and defective skeletal mineralization.
XLH results from mutations in the PHEX (phosphate-regulating gene with homologies to
endopeptidases on the X chromosome, Xp22.1) gene, which encodes a membrane-bound
endopeptidase (52), whereas ADHR is associated with mutations of the gene encoding
fibroblast growth factor 23 (FGF23) and is linked to chromosome 12p13.3 (53). The latter
is a small heat-sensitive molecule less than 25 kDa that inhibits sodium-dependent

phosphate wasting and probably inhibits CYP1. The endopeptidase, PHEX, degrades native
FGF23 which provides the biochemical link among these clinical syndromes. XLH rickets
also has been associated with mutations in CLCN5, a voltage-gated chloride channel gene
located on Xp11.22.
Other factors that enhance 1,25(OH)2D production include estrogen, prolactin, growth
hormone, insulin-like growth factor I, and PTHrP. 1,25(OH)2D production is feedback
regulated and is inhibited by chronic deficiency of or low circulating Mg (36). Mg deficiency
also lowers serum 1,25(OH)2D response to a low-Ca diet but does not appear to limit 1,25
(OH)2D production in animals (54). The effect of Mg on 1,25(OH)2D metabolism is
presumably related in part to its role as a cofactor of the CYP1 enzyme. In contrast to the
rapid increase within minutes in the secretion and serum concentration of PTH, measurable
alterations in serum 1,25(OH)2D concentrations usually occur hours after exposure to an
appropriate stimulus. Extrarenal production of 1,25(OH)2D in macrophages, particularly in
granulomatous disease states, may not be tightly regulated; is stimulated by -interferon
(55), but is not responsive to changes in dietary Ca intake.
The degradation of 1,25(OH)2D is also tightly regulated. 1,25(OH)2D strongly induces the
enzyme 25-hydroxyvitamin D-24 hydroxylase (CYP24) in all target cells for vitamin D.
CYP24 catalyzes several steps of 1,25(OH)2D degradation, collectively known as the C24
oxidation pathway, which starts with 24-hydroxylation and culminates in the formation of
the biliary excretory form, calcitroic acid. CYP24 expression is inhibited by PTH and by
dietary phosphate restriction. In the kidney and intestine in particular, upregulation of the
CYP24 enzyme in response to 1,25(OH)2D treatment is rapid and occurs within 4 hours
(56). Physiologic production of 24,25(OH)2D is therefore an important means to regulate
the circulating concentration of 1,25(OH)2D and catabolism of vitamin D, although it may
also have a role in bone integrity and fracture healing in the chick model. Most of the other
vitamin D metabolites are derived primarily from further metabolic alterations to 25-OHD
and 1,25(OH)2D through oxidation or side chain cleavage and have poorly defined
physiologic functions. However, many analogues of vitamin D metabolites are being
studied for the numerous potential pharmacologic actions that involve less calcemicinducing and more maturation and differentiation effects.
Like other steroid hormones, 1,25(OH)2D function is mediated primarily through
modulation of the cellular genome by binding to a specific nuclear receptors, vitamin D
receptor, VDR, a 424-amino-acid phosphoprotein for which the x-ray crystallographic
structure has been determined (47). The VDR gene contains nine exons and is located on
chromosome 12q13-14 near the site of the gene for CYPl (57). VDR is a member of the
subfamily of nuclear receptors with ligand-binding domains that bind classic hormones
including thyroid hormone, androgens, estrogens, progesterone, glucocorticoids,
aldosterone, hormonal forms of vitamin A, and 1,25(OH)2D. It has several functional
domains including a 110-residue N-terminal DNA-binding domain with two zinc fingers, Cterminal hormone-binding domain, and hinge region important for nuclear localization. The
VDR interacts with the 9-cis retinoic acid nuclear receptor retinoid X receptor (RXR) to form
a heterodimeric RXR-VDR complex that binds to specific DNA sequences, termed vitamin Dresponsive elements (VDREs). After 1,25(OH)2D binds to the receptor, it induces
conformational changes (58) that result in the recruitment of a multitude of transcriptional
coactivators that stimulate the transcription of target genes. VDR also can adopt a dual
role as a repressor in the absence of ligand and then subsequently as a coactivator when a
ligand binds. VDR is upregulated by 1,25(OH)2D at both the mRNA and protein levels and
is increased during growth, gestation, and lactation; however, it shows an age-dependent
decrease in mature animals and humans, supporting the notion that VDR may be up- or
downregulated, depending on Ca needs.
Although 1,25(OH)2D regulates more than 60 genes whose actions include those
associated with Ca homeostasis and immune responses, as well as cellular growth,
differentiation, and apoptosis, two basic clinical functions define the major classic actions
of vitamin D. The first is that vitamin D is required to prevent rickets in children and
osteomalacia in adults. The second is the prevention of hypocalcemic tetany. These
functions are maintained by 1,25(OH)2D through its effect on a number of target tissues,
primarily intestine, kidney, and bone, with modulating effects from other hormones
including PTH and CT.
The genomic action of 1,25(OH)2D can be preceded by more rapid nongenomic actions that
occur in minutes and involve membrane-associated events such as increased Ca transport,
and PKC and mitogen-activated protein kinase activation. This nongenomic rapid increase
in cytosolic Ca within seconds to minutes is reported to occur in various cell types from the
intestine and parathyroid, osteoblasts, myocytes, and leukemic cells (59).
Quantification of vitamin D and its metabolites has been achieved by several different
methods, including high-performance liquid chromatography, with detection by ultraviolet
absorbance or binding assays, and immunoas- says based on antibodies raised to vitamin
D metabolite conjugates. Values from different laboratories cannot be compared without
making direct comparison of their assay procedures. Interlaboratory coefficients of
variation for the
P.855
measurement of 25-OHD, 24,25(OH)2D, and l,25(OH)2D may range between 35% and
52%. Furthermore, differences between the affinities of vitamins D2 and D3 to DBP and
VDR, and different chromatographic behavior on various preparative chromatographic
systems demand that great care be taken with assay techniques when dealing with
patients who have significant vitamin D2 intake. To ensure reliable results, appropriate
vitamin D standards must be used for standard curve generation in performing competitive
protein binding assays of these compounds.
Maternofetal transfer of vitamin D and its metabolites varies, depending on the species. In
humans, the cord-serum vitamin D concentration is very low and may be undetectable; the
25-OHD concentration is directly correlated with, but is lower than, maternal values,
consistent with placental crossover of this metabolite; and 1,25(OH)2D concentrations also
are lower than maternal values, but there is no agreement on the maternofetal relationship
of this and other dihydroxylated vitamin D metabolites (3,60-62). However, the placenta,
like the kidney, produces 1,25(OH)2D, making it difficult to ascertain just how much fetal
1,25(OH)2D results from placental crossover versus placental synthesis. 24,25(OH)2D also
crosses the placenta, and its concentration in the sera of mothers and newborns varies
with the seasons, being highest in autumn. It appears that the human fetus receives the
bulk of its vitamin D already metabolized to 25-OHD.
Seasonal and racial variations in serum 25-OHD concentrations occur, presumably from
variations in endogenous production. The concentration of serum 25-OHD, as with 24,25
(OH)2D, is lower in winter. These changes may be reflected in cord-serum values. In
normal adults, serum 1,25(OH)2D concentrations are relatively constant and maintained
within approximately 20% of the overall 24-hour mean, and show no seasonal variation,
which is consistent with the tightly regulated CYPl activity. African American mothers,
infants, and young children tend to have lower 25-OHD and higher 1,25(OH)2D
concentrations than their white counterparts. Serum 1,25(OH)2D in the newborn becomes
elevated within 24 hours after delivery and appears to vary with to Ca and P intake.
The circulating half-life of vitamin D is about 24 hours and for 25-OHD is 2 to 3 weeks,
although the latter half-life is decreased in vitamin D-deficient individuals. 1,25(OH)2D has
a much shorter half-life of 3 to 6 hours. Metabolites of 25-OHD and 1,25(OH)2D may
undergo enterohepatic circulation after exposure to intestinal -glucuronidase. The
physiologic role of enterohepatic circulation of vitamin D metabolites has not been
precisely quantitated.
NONCLASSIC CONTROL OF CALCIUM AND MAGNESIUM

HOMEOSTASIS
Factors other than the classic calciotropic hormones, PTH, CT, and 1,25(OH)2D, whether
acting systemically or locally on multiple effector organs, are increasingly being recognized
as important to the maintenance of mineral homeostasis in certain circumstances. The
ultimate effect on mineral homeostasis often involves bone formation and/or resorption
and flux of Ca and Mg between extracellular fluid and bone, with or without direct
involvement by calciotropic hormones. Skeletal health, particularly in the growing skeleton,
requires the integrated actions of classic calciotropic hormones, endocrine modulators of
growth, numerous cytokines and growth factors and their receptors and endogenous
modulators.
Many factors, such as growth hormone, insulin-like growth factor I, estrogen,
progesterone, cortisol, and TNF, can affect the secretion or function of one or more of the
calciotropic hormones. In turn, many factors such as insulin-like growth factor I;
transforming growth factor-1; IL-1, -2, -4, and -6; TNF-, and interferon- (IFN-) can be
modulated by calciotropic hormones.
Local factors such as transforming growth factor-1, lymphotoxin, TNF-, IFN-, IL-1, and
IL-6 acting in a paracrine (i.e., cell-to-cell) or autocrine (i.e., cell-to-self) fashion may
influence Ca flux of bone cells. The effects on Ca flux based on these interactions are
probably more important under pathologic situations. IFN- from activated macrophages
(55,63) stimulates CYP1 mRNA and enzyme production, with little or no feedback
inhibition by 1,25(OH)2D, which potentially may compromise Ca homeostasis.
Interaction between systemic and local factors can occur, and some factors such as PTHrP
may act both systemically and locally (64). PTHrP is also known as PTH-like peptide, PTHlike protein, or human humoral hypercalcemic factor. PTHrP and PTH genes appear to be
members of the same gene family. PTHrP mRNA encodes a 177-amino-acid protein
consisting of a 36-amino-acid precursor segment and 141-amino-acid mature peptide. The
mature PTHrP contains several structural or functional domains. The N-terminal 1-13
region has eight of 13 residues in common with PTH. The amino acids 34-111 segment is
highly conserved among species while amino acid 118 to the C-terminus is poorly
conserved. PTHrP gene expression is found in an extensive variety of endocrine and
nonendocrine tissues. PTHrP biologic activity and immunoreactivity for PTHrP mRNA have
been found in many tissues, by as early as 7 weeks of gestation, including the fetus,
placenta, lactating breasts, and milk in human and in various tissues in the sheep (65) and
pig (66). Both PTH and PTHrP appear to bind to the same G-protein-linked receptor.
Synthetic and recombinant PTHrPs can mimic the effects of PTH on the classic PTH target
organs, involving activation of adenylate cyclase and other second messenger systems.
Several PTHrP assays with varying sensitivities and specificities have been developed which
account for the variability reported between assays (67). The stability of PTHrP in plasma
samples may be enhanced if sample collection is done in the presence of protease
inhibitors. Circulating immunoreactive PTHrP concentrations are low or undetectable in
normal subjects. Serum PTHrP is
P.856
increased during pregnancy (5,6) and is similar to or lower than umbilic cord PTHrP
concentrations. In cord sera, PTHrP concentration is 10- to 15-fold higher than that of PTH.
Amniotic fluid PTHrP concentrations at mid-gestation and at term are 13- to 16-fold higher
than the cord or maternal levels (68), and the concentration of PTHrP in milk is 100-fold
higher. PTHrP concentrations correlate positively with total milk calcium (69).
PTHrP concentrations in the circulation of individuals with humoral hypercalcemia of
malignancy (HHM) are elevated (67). The amino-terminal fragment PTHrP 1-74 appears to
be specific for HHM, whereas the carboxyl-terminal fragment PTHrP 109-138 is elevated in
the serum of patients with HHM or renal failure. The levels of PTHrP in these patients are
similar to the concentration of PTH (10-12 to 10-11 mol/L).
Clinically, PTHrP is the humoral mediator secreted by tumors that results in the syndrome
of HHM, and the measurement of PTHrP is of clinical utility primarily as a tumor marker in
HHM. Physiologically, PTHrP is an important paracrine regulator of several tissue-specific
functions that may directly or indirectly affect fetal and neonatal mineral homeostasis,
probably through its effect on smooth muscle relaxation, placental Ca transport, lactation,
fetal bone development, and control of cellular growth and differentiation.
DISTURBANCES IN SERUM MINERAL

CONCENTRATIONS
Hypocalcemia
Neonatal hypocalcemia may be defined as a serum tCa concentration of less than 2 mmol/
L (8 mg/dL) in term infants and 1.75 mmol/L (7 mg/dL) in preterm infants with iCa below
1.0 to 1.1 mmol/L (4.0 to 4.4 mg/dL), depending on the particular ion-selective electrode
used. Whole-blood iCa shows similar values to serum iCa and is often used to determine
hypocalcemia. However, the appropriate range used is also subject to the type of
instrument used (15).
The definition of hypocalcemia is based on the clinical perspective because serum Ca
concentrations are maintained within narrow ranges under normal circumstances, and the
potential risk for disturbances of physiologic function increases as the serum Ca
concentration decreases below the normal range. Furthermore, improvements in
physiologic function, e.g., changes in cardiac contractility, blood pressure, and heart rate,
are reported in hypocalcemic infants undergoing Ca therapy (70,71,72), and a higher
mortality rate has been reported for children with hypocalcemia in pediatric intensive care
settings (73).
Clinically, there are two peaks in the occurrence of neonatal hypocalcemia. An early form
typically occurs during the first few days after birth, with the lowest concentrations of
serum Ca being reached at 24 to 48 hours of age; late neonatal hypocalcemia occurs
toward the end of the first week. These findings reflect in part the traditional clinical
practice of screening for biochemical abnormalities in small or sick hospitalized infants
during the first few days, and in symptomatic infants during hospitalization and after
hospital discharge. However, the nadir of serum Ca concentration may occur at less than
12 hours (9 to 12) or not until some weeks after birth (74,75), and many neonates,
particularly those with genetic defects in Ca metabolism, may be hypocalcemic but remain
asymptomatic and undetected during the early neonatal period. This also may contribute to
the less frequent diagnosis of late neonatal hypocalcemia compared to early neonatal
hypocalcemia. Additionally, increased understanding of the mechanisms of disturbed Ca
metabolism would support the approach to neonatal hypocalcemia based on risk factors
and pathophysiologic basis rather than the traditional early or late onset.
Pathophysiology
Multiple risk factors for neonatal hypocalcemia (Table 36-1) support the existence of varied
and frequently interrelated pathophysiologic mechanisms (Table 36-2). However, the
pathophysiologic mechanisms are not fully defined for all cases of hypocalcemia. In most
cases of neonatal hypocalcemia, there is a decrease in both tCa and iCa, although iCa may
be decreased without lowering tCa.
There are common bases for the occurrence of hypocalcemia, particularly for early onset
hypocalcemia. These include the abrupt discontinuation of placental Ca supply at birth,
limited or no dietary Ca, transient limited increase
P.857
in the serum PTH concentration, possibly end-organ resistance to PTH and 1,25(OH)2D,
and elevated serum CT concentration. Many illnesses may preclude early enteral feeding
but many clinicians do not use parenteral nutrition that contains Ca for 1 or more days
after birth, thus increasing the risk for hypocalcemia. Even in healthy term infants, the
amount of Ca retention from milk feeds probably is less than 20 mg/kg body weight on the
first day, rising to at least 45 to 60 mg/kg on the third day; these amounts are significantly
lower than the daily in utero Ca accretion of more than 100 mg/kg during the third
trimester (2).
TABLE 36-1 RISK FACTORS FOR NEONATAL HYPOCALCEMIA
Maternal
Insulin-dependent diabetes
Hyperparathyroidism
Vitamin D or magnesium deficiency
Medications: calcium antacid and anticonvulsant (?)
Narcotic use (?)
Peripartum
Birth asphyxia
Infant
Intrinsic
Prematurity
Malabsorption
Malignant infantile osteopetrosis
Parathyroid hormone: impaired synthesis, secretion, regulation, or responsiveness

Mitochondrial fatty acid disorder (?)
Extrinsic
Diet
Inadequate calcium
Excess phosphorus
Enema: phosphate
Exchange transfusion with citrated blood
Infectious diarrhea (?)
Clinical therapy (?): phototherapy, alkali, high rate of intravenous lipid
TABLE 36-2 PATHOPHYSIOLOGY OF NEONATAL HYPOCALCEMIA

Physiologic Basis
Calcium (Ca)
Magnesium (Mg)
Mechanism
Clinical Association
Decreased reserves
Prematurity
Decreased intake or
Prematurity, malabsorption syndrome
absorption
Increased Ca complex Chelating agent (e.g., citrated blood
for exchange transfusion, long-chain
free fatty acid)
Decreased tissue store IDM, maternal hypomagnesemia
Decreased intake or
absorption
Prematurity, malabsorption
syndrome, specific Mg malabsorption
(rare)
Increased loss
Intestinal fistula, enterostomy, or

renal (primary or secondary)
Phosphorus (P)
Increased load
Exogenous (e.g., dietary, enema)

phosphate loading
pH
Increased
Respiratory or metabolic alkalosis (e.

g., shifts Ca from ionized to protein-
Parathyroid hormone Inadequate or

(PTH)
bound fraction)
Maternal hypercalcemia, DiGeorge
defective
synthesis or
association,
hypoparathyroidism,
secretion
Impaired regulation
hypomagnesemia, PTH gene mutations

CaR activating mutations, autosomaldominant or
sporadic hypocalcemia with
Impaired
responsiveness
hypercalciuria
Chronic hypomagnesemia, type 1 PTH
receptor
inactivating mutation (?),
pseudohypoparathyroidism
IDM, birth asphyxia, prematurity
Severe maternal deficiency
Calcitonin
Vitamin D
Increased
Deficiency
Decreased response to Prematurity
1,25(OH)2D
Osteoclast activity
Miscellaneous
Absent
Increased anabolism
Others ?
Malignant infantile osteopetrosis

Hungry bone/refeeding syndrome
Mitochondrial fatty acid disorder,
rotavirus diarrhea,
phototherapy, narcotic withdrawal
1,25(OH)2D, 1,25-dihydroxyvitamin D; IDM, infant of insulin-dependent diabetic
mother;
CaR, calcium-sensing receptor
Hypocalcemia (in varying degrees of severity) may occur in association with transient
congenital hypoparathyroidism (TCHP), i.e., suppression of fetal and neonatal parathyroid
function from maternal hyperparathyroidism (74,75) and maternal use of high doses of
calcium carbonate (76) as antacid, and thus, impaired PTH response to the interruption of
placental supply of Ca at birth. Neonatal hypocalcemia is often the first manifestation that
leads to the diagnosis of maternal hyperparathyroidism.
In the neonate, hypocalcemia frequently occurs in the presence of rising concentrations of
PTH in the circulation. It is possible that this represents either a relative inadequate
response of the parathyroid gland or end-organ resistance to PTH. Resistance to
pharmacologic doses of 1,25(OH)2D, demonstrated in vitro (77) and in vivo in infants

(11,12), may also contribute to hypocalcemia.
Despite the hypocalcemic effects of CT, the role of CT in the development of neonatal
hypocalcemia remains uncertain. Serum CT concentrations continue to increase after birth
in neonates of normal and diabetic pregnancies (9,26) irrespective of the variation in
serum Ca; in neonates with birth asphyxia (9); and in preterm infants (25). The stimulus
for the postnatal rise in serum CT, despite falling serum Ca, is unknown. There are
conflicting reports on the effect of Ca supplementation to suppress the postnatal surge in
CT secretion. However, serum CT is increased after an intravenous bolus of Ca during
exchange blood transfusion (78).
The above problems are exaggerated in the preterm infant and account for the inverse
relationship between the frequency of hypocalcemia versus birth weight and gestational
age; more than 50% of preterm very-low-birth-weight neonates may have hypocalcemia
(10,11,12). Infants with intrauterine growth retardation may have hypocalcemia if they are
also preterm or have birth asphyxia; otherwise, there is apparently no increased incidence
of hypocalcemia related to growth retardation (13).
P.858
Hypomagnesemia may be contributory to hypocalcemia in infants of mothers with insulindependent diabetes (79), although gestational diabetics may (80) or may not (81) have
disturbed mineral metabolism. Both hypocalcemia and hypomagnesemia may be the
result of a common insult from the diabetic pregnancy, and rigid control of maternal
glucose levels during pregnancy may significantly diminish these complications (82).
Severe and persistent hypomagnesemia from any cause can result in hypocalcemia (see
hypomagnesemia section). Deficiency of various minerals, including Ca, Mg, and trace
minerals such as zinc, can result from chronic intestinal malabsorption and fistula or
enterostomy loss. During infancy, congenital or acquired short-bowel syndrome and any
chronic diarrheal condition, especially if associated with steatorrhea, are leading causes of
malabsorption and possibly impaired enterohepatic circulation of vitamin D and vitamin D
metabolites.
Excessive P load can result in hypocalcemia. Cow milk and even humanized cow-milkderived formulae (83,84) with lower P content compared with cow milk but higher
compared with human milk, and cereals which usually have high P content are typical
sources of dietary P load. Accidental overdose of oral phosphate supplement (85) or
phosphate-containing enema (86) are less frequent causes of excessive P load.
Neonatal hypocalcemia from impaired synthesis or secretion of PTH in the newborn may be
secondary to maternal hypercalcemia or to developmental defects of the parathyroid gland.
A variety of mutations of PTH or CaR genes, some with Mendelian modes of inheritance,
can affect the synthesis, metabolism, and function of PTH and result in hypocalcemia.
Relative inadequacy or the transient nature of the PTH response to the abrupt withdrawal
of the placental transfer of Ca contributes to the fall in serum Ca after birth. This also may
be responsible for the hypocalcemia induced by exchange transfusion using citrated blood
(78,87) or feeding of the relatively high P content of cow-milk formula (83,84). The ability
of the neonatal parathyroid glands to respond to hypocalcemic stress increases with
postnatal age. Neonates with TCHP may have prolonged hypocalcemia that requires
treatment until late infancy or early childhood, and hypoparathyroidism may recur in later
childhood (88,89,90).
Hypoparathyroidism in the infant is a heterogeneous group of disorders and may occur
sporadically or with differing Mendelian modes of inheritance (91,92,93). Synthesis of
defective PTH can occur in the autosomal-dominant form with a point mutation in the
signal peptide-encoding region for prepro-PTH. The autosomal-recessive form is associated
with a mutation in the donor splice site leading to transcriptional loss of the second exon
and prevention of translation. The X-linked recessive form is associated with embryonic
dysgenesis of parathyroid glands. Hypo-parathyroidism from fetal parathyroid hypoplasia
or dysgenesis usually requires lifelong treatment to prevent hypocalcemia.
Deletion of chromosome 22q11.2 is associated with varied phenotypic manifestations
including DiGeorge and velocardiofacial/Shprintzen syndromes. Both syndromes may
represent different degrees of the same disorder with partial or complete absence of
derivatives of the third, fourth, and possibly fifth pharyngeal pouches, and are often
associated with defective development of the third, fourth, and sixth aortic arches. It is
estimated that up to 30% of these patients may have hypoparathyroidism, although far
fewer patients develop hypocalcemia (94). Delayed motor development, cognition and
neurodevelopment, and behavior and temperament problems are frequently reported in
more than 50% of affected patients (95,96). Early screening and intervention for these
problems are advised. Multiple other organ systems (94,97) may be involved and include
some combination of congenital heart disease, primarily involving the aortic arch,
decreased T-cell number or function, and possibly thyroid C-cell deficiency. DiGeorge
association may be inherited in an autosomal-dominant fashion (98).
Dysregulation of PTH can result from activating mutations of CaR with a reduction in EC50
(concentration of extracellular Ca required to elicit half of the maximal increase in
intracellular inositol phosphate) to suppress PTH synthesis. It is manifested as autosomal
dominant or as sporadic cases of hypocalcemia with hypercalciuria (99,100). The latter is
an effect of the mutated CaR in the kidneys. Hypocalcemia is usually mild and
asymptomatic, and diagnosis is often delayed beyond the neonatal period, although
hypocalcemia was likely present during the immediate newborn period.
Relative defective response to PTH can result in neonatal hypocalcemia. The inactivating
mutation of the type 1 PTH receptor gene, as documented in Blomstrand
chondrodystrophy, is present in the prenatally lethal form of short-limb dwarfism (101).
Theoretically this defective response to PTH may result in hypocalcemia but the regulation
of serum Ca has not been evaluated in vivo.
Impaired end-organ response to PTH occurs with chronic hypomagnesemia and may
involve simultaneous impairment in both PTH and 1,25(OH)2D pathways (36). End-organ
unresponsiveness to PTH associated with genetic defect is classically manifested as
pseudohypoparathyroidism type 1a (PHP-1a) or Albright hereditary osteodystrophy. The
biochemical basis of the defect is proximal to cAMP production (102). It is inherited in an
autosomal-dominant fashion with heterozygous inactivating mutations in the maternal
GNAS1 exons that encode the subunit of Gs (Gs). The gene GNASl is located on
chromosome 20q13.3 and encodes 13 exons that are alternatively spliced to yield four Gs
proteins. Multiple mutations have been reported and include abnormalities in splice
junctions associated with deficient mRNA production and point mutations that result in
diminished amounts and activities of the G proteins. The inactivating mutation of the gene
impairs the production of the adenylate cyclase second messenger system, leading to
resistance to multiple hormones (including PTH, vasopressin, and
P.859
thyrotropin) that activate Gs. Clinical manifestations include short stature, round face,
brachymetacarpals and brachymetatarsals, dental dysplasia, subcutaneous calcifications,
abnormalities in taste, smell, hearing, and vision, and developmental delay. Biochemical
abnormalities include hypocalcemia, hyperphosphatemia, increased circulating PTH, and
insensitivity to the administration of exogenous PTH (unaltered urinary Ca, P, and cAMP) in
the absence of compromised renal function. The extent of resistance to other hormones is
variable, and the complete biochemical picture is usually not evident until 2 to 3 years
after birth.
Parent-specific methylation with parental imprinting of the GNAS1 gene, involving selective
inactivation of either the maternal or paternal allele, is possible and leads to different
phenotypic expression. In the case of the Gs gene, it is paternally imprinted (silenced) so
that the disease PHP-1a is not inherited from the father carrying the defective allele but
only from the mother (103). However, the defective allele is not imprinted or silenced in all
tissues and reflects haplotype insufficiency. For example, PHP-1b is characterized by
isolated resistance to PTH without the accompanying skeletal manifestations. Paternal
isodisomy of chromosome 20q in patients that lack the maternal-specific methylation
pattern within GNAS1 results in normal Gs protein and activity in the fibroblast but not in
the renal proximal tubules (104). There is a third type, PHP-1c, reported in a few patients
that differs from PHP-1a only in having normal erythrocyte levels of Gs; presumably there
is a post-Gs defect in adenyl cyclase stimulation. All type 1 PHP individuals show a
deficient urinary cAMP response to the administration of exogenous PTH. Individuals with
pseudopseudohypoparathyroidism have typical clinical manifestation of PHP-1a but have
normal serum Ca and normal response of urinary cAMP to exogenous PTH. The mutated
GNAS1 gene is inherited from the father, i.e., paternal imprinting, with suppression of the
mutant copy in selected tissues, and there is a 50% reduction in the amount of Gs.
Infants with neonatal hypocalcemia seizures and transient biochemical features of
pseudohypoparathyroidism have been reported (105). These infants have elevated serum
PTH and P with hypocalcemia at diagnosis. Administration of exogenous human PTH 1-34
showed little phosphaturic effect although there was brisk response in plasma and urine
cAMP and alkaline phosphatase. After initial treatment for hypocalcemia, the serum Ca and
PTH spontaneously normalized before 6 months of age.
Maternal anticonvulsant therapy with phenytoin and phenobarbital also may result in
neonatal hypocalcemia, presumably from increased clearance of vitamin D secondary to
the induction of the hepatic cytochrome P450 enzyme system. However, other maternal
factors including seasonal variation in sunlight exposure, increased maternal age and
parity, and poor socioeconomic status, may contribute to development of neonatal
hypocalcemia, presumably in part from varied and probably deficient maternal vitamin D.
Furthermore, there is no seasonal variation in the rate of early neonatal hypocalcemia
(106) despite seasonal variation in maternal and fetal vitamin D status, as indicated by
maternal and cord 25-OHD concentrations. Thus, maternal vitamin D or Mg deficiency

probably predisposes to but is not the primary cause of hypocalcemia in the neonate.
Malignant infantile osteopetrosis may present with neonatal hypocalcemia, presumably
reflecting continued Ca uptake from unopposed bone formation (107). Rapid replenishment
of nutrients in severe deficiency, including prolonged starvation, often leads to disturbed
blood biochemistries including hypokalemia, -phosphatemia, -magnesemia and -calcemia.
This is known as the refeeding syndrome or hungry bone syndrome with excessively
rapid shift of electrolytes and minerals intracellularly in various tissues, particularly muscle
and bone (108,109).
The pathophysiology in some situations with hypocalcemia remains ill defined.
Approximately 40% of infants with severe diarrhea from rotavirus have hypocalcemia, and
it resolves with symptomatic support and improvement in diarrhea (110). Mitochondrial
fatty acid disorders have been associated with severe metabolic anomalies including
hypoglycemia, hypocalcemia, hyperkalemia and metabolic acidosis, and organ dysfunction
including hepatic and cardiac failure (111).
Decreases in serum iCa can occur without decreases in serum tCa. Agents that complex Ca
in the blood would be expected to decrease iCa. Such agents include citrate, which is used
as an anticoagulant for blood storage. During exchange blood transfusion, iCa can
decrease to 0.5 mmol/L in spite of administration of conventional amounts of Ca (i.e., 0.5
to 1 mL of 10% Ca gluconate for each 100 mL of blood exchanged) during the transfusion.
Increased levels of long-chain free fatty acids from intravenous lipid emulsion can complex
Ca and reduce iCa in vitro; thus, hypocalcemia potentially can occur with excessive rates of
intravenous lipid infusion. Alkalosis can result in shifts of Ca from the ionized state to the
protein-bound fraction. Because alkalosis increases neuromuscular hyperirritability, the
combination of decreased serum iCa and alkalosis may precipitate clinical tetany in an
infant with borderline serum Ca status. In clinical practice, administration of sodium
bicarbonate in the therapy of metabolic acidosis often occurs in situations with high risk of
hypocalcemia such as prematurity or perinatal asphyxia, but it is not known if it has an
independent role in the development of hypocalcemia. The mechanisms for hypocalcemia
in some situations are not known. For example, neonates with severe hyperbilirubinemia
tend to have lower iCa (112), the use of phototherapy may be associated with
hypocalcemia (113), and infants born to narcotic-using mothers are reported to have a
lower serum iCa if they manifest withdrawal symptoms (114).
Diagnosis
Suspicion of hypocalcemia must be confirmed by measurement of serum tCa and iCa since
clinical manifestations are
P.860
many and varied and may be indistinguishable from other common neonatal diseases
(Table 36-3). Confirmation of hypocalcemia as the cause of clinical manifestations is its
reversibility when serum tCa or iCa has been normalized.
TABLE 36-3 DIAGNOSTIC WORKUP FOR NEONATAL HYPOCALCEMIA
History
Screen for risk factors (Table 36-1)
Physical examination
General examination with focus on peripheral and central
nervous and cardiovascular systems
Associated features, e.g., infant of a diabetic mother,
prematurity, birth asphyxia, congenital heart disease,
pseudohypoparathyroidism, etc.
abc
Investigations , ,
Serum total and ionized calcium (tCa and iCa), magnesium (Mg),
phosphorus (P), total protein and albumin, and simultaneous
intact or whole parathyroid hormone (PTH)
Acid-base status
Complete blood count (lymphocyte count)
Electrocardiogram (Q-Tc>0.4 sec or Qo-Tc >0.2 sec)
Chest x-ray (thymic shadow, aortic arch)
Urine Ca, P, Mg, and creatinine
Meconium and urine screen for narcotics
Maternal serum +Ca and iCa, Mg and P, urine Ca and P, if
suspect maternal or heritable calcium disorder is suspected,
particularly in persistent neonatal hypocalcemia
Additional workup as indicated: vitamin D metabolites, T-cell
number and function, malabsorption studies, response to
exogenous PTH, molecular genetic studies (deletion of
22q11.2, PTH receptor and end-organ responsiveness
abnormalities, and calcium-sensing receptor defects, etc.) and
family screening.
a
If serum tCa and iCa are normal, diagnostic workup should focus on non-calcium
related causes of clinical symptomatology, e.g., serum glucose, sepsis workup,
screen for excretion of illicit drugs, neuroimaging studies, etc.
b
Resolution of clinical symptomatology when serum tCa or iCa has been

normalized confirms the role of hypocalcemia.
Maternal and family screening for calcium disorders is indicated in the absence of
specific diagnosis for the neonatal hypocalcemia.
The less mature the infant, the more subtle and varied are the clinical manifestations; in
addition, the infant is frequently asymptomatic. Clinical manifestations may include
irritability, jitteriness or lethargy, poor feeding with and without feeding intolerance,
abdominal distention, apnea, cyanosis, and seizures, which may be confused with
manifestations of hypoglycemia, sepsis, meningitis, anoxia, intracranial bleeding, and
narcotic withdrawal. The degree of irritability of the infants does not appear to correlate
with serum Ca values. Frank convulsions are seen more commonly with late neonatal
hypocalcemia. In newborn infants, the classic signs of tetany from peripheral
hyperexcitability of motor nerves, including carpopedal spasm (spasm of the wrists and
ankles, Trousseau sign), facial spasm (Chvostek sign), and laryngospasm (spasm of the
vocal cords), are uncommon.
The level of iCa that determines which feature of tetany will be manifested varies among
individuals and is affected by other components of the extracellular fluid, e.g.,
hypomagnesemia and alkalosis lower, whereas hypokalemia and acidosis raise, the
threshold for tetany. At physiologic concentrations of hydrogen and potassium ions, tetany
may develop in older infants at an iCa less than 0.8 mmol/L (3.2 mg/dL), and will almost
always be manifested (with the possible exception of preterm infants) at an iCa less than
0.6 mmol/L (2.4 mg/dL). If serum albumin concentrations are normal, the corresponding
serum tCa concentrations usually are less than 1.8 mmol/L (7.2 mg/dL). In the preterm
infant, serum iCa may not decrease to the same extent as tCa, presumably in part because
of the sparing effect of lower serum albumin and acidosis found frequently in these infants.
This also may partially explain the frequent lack of clinical signs of hypocalcemia in
preterm infants. The measurement of electrocardiographic QT intervals, corrected for heart
rate, and standard nomograms relating serum tCa and total protein to iCa, have little value
for the prediction of neonatal serum iCa. Serum tCa is correlated with iCa but is also
inadequate for the prediction of one from the other.
Management
Symptomatic hypocalcemia, manifested as seizures for example, should be treated
promptly with parenteral Ca. It is possible that neonatal hypocalcemia may resolve
spontaneously. However, asymptomatic hypocalcemia probably also should be corrected,
as Ca potentially can alter important cellular functions in which Ca serves either as a first
or second messenger in cellular activity.
Any neonate with seizures should have blood drawn for diagnostic tests before therapy.
Intravenous administration of Ca salts is the most effective and most rapid means of
elevating serum Ca concentrations. Gradual or abrupt decreases in heart rate during the
infusion is an indication to slow or stop the infusion. In neonates, 10% Ca gluconate (0.45
mmol [18 mg] elemental Ca/kg) can effectively increase serum iCa, heart rate, cardiac
contractility, and blood pressure (70,71,72) (Table 36-4). In children, small equimolar
doses (0.07 mmol [2.8 mg] elemental Ca/kg) of 10% Ca chloride compared to 10% Ca
gluconate may result in higher mean arterial blood pressure with a slightly greater mean
increase (0.06 mmol/L [0.2 mg/dL]) in the measured serum iCa (115). Prolonged use of
Ca chloride in high doses may be associated with acidosis and probably should be avoided.
With intravenous Ca therapy, bolus infusion may be associated with a transient slight
decrease in blood pH and serum P and with hypercalcemia. Continuous infusion probably is
more efficacious than intermittent therapy because renal loss of Ca may be greater with
the latter method; a dose of 1.25 to 2.0 mmol (50 to 80 mg) elemental Ca/kg/day has
been used successfully in the treatment and prevention of neonatal hypocalcemia.
Intravenous Ca supplements should be rapidly weaned or replaced with Ca-containing
parenteral nutrition if the infant is not expected to tolerate enteral feeding.
Arterial infusion of Ca in high concentrations is potentially fraught with many dangers and
should be avoided.
P.861
Massive sloughing of soft tissue may occur in the distribution of the arterial supply; for
example, inadvertent administration into a mesenteric artery theoretically can lead to

necrosis of intestinal tissues. If an umbilic venous catheter is used, the tip should be in the
inferior vena cava and not intracardiac, as administration of Ca directly into the heart may
result in arrhythmia. Parenteral nutrition solutions containing standard mineral (including
Ca) content can be safely infused through appropriately positioned umbilic venous or
arterial catheters. Direct admixture of Ca preparations with bicarbonate or phosphate
solutions will result in precipitation and must be avoided.
TABLE 36-4 MANAGEMENT OF NEONATAL HYPOCALCEMIA
Acute phase therapy
Correction of hypomagnesemia, acid-base problem, etc, if possible

Intravenous 10 to 20 mg elemental Ca/kg as 10% Ca gluconate or 10% Ca chloride
(provides 9 mg elemental Ca/mL or 27.2 mg/mL, respectively) with dextrose water
or normal saline infused over 5 to 10 minutes under constant ECG monitoring;
repeat as necessary until resolution of severe symptomatology, such as seizures.
In infants who are not fed enterally, this is followed by intravenous continuous
infusion at 50 to 75 mg elemental Ca/kg/d. Alternately, parenteral nutrition
containing 50 mg elemental Ca/100 mL is preferred and continued until feeding
In asymptomatic infants, oral 50 to 75 mg elemental Ca/kg/d in 4 to 6 divided

doses. One mL of calcium carbonate, glubionate, gluceptate, gluconate, lactate, or
chloride contains 40, 23, 18, 9, 13, and 27 mg elemental Ca, respectively.
Once serum tCa is normalized, cut the dose of the Ca supplement in half daily for 2
days, then discontinue
Serial serum tCa (+/-iCa) every 12-24 hours until clinically stable, every 24 hours
until normalized, and at 24 hours after Ca supplement discontinued
Maintenance therapy: treat underlying disorder, if possible
Low phosphorus (P) formula if serum P is high (>2.6 mmol/L or 8 mg/dL) until
serum Ca and P normalized
Prolonged and higher Ca doses, and 1,25(OH)2D may be needed, e.g.,
hypoparathyroidism
1,25 (OH)2D, 1,25-dihydroxyvitamin D; ECG, electrocardiogram.
Oral Ca supplements at a similar dosage to parenteral Ca (1.87 mmol [75 mg] elemental
Ca/kg per day in four to six divided doses) should be started if the infant is expected to
tolerate it, and the serum Ca is normalizing after the initial intravenous Ca therapy. All oral
Ca preparations are hypertonic, and there is a theoretical potential for precipitating
necrotizing enterocolitis in infants at risk for this condition. Oral Ca preparations generally
contain higher Ca concentration than intravenous preparations, for example, Ca
glubionate, gluceptate, and carbonate have respectively 2.88, 2.25 and 2.5 mmol (115,
90, and 200 mg) elemental Ca per 5 mL, and are useful for infants, particularly those
requiring fluid restriction. Syrup-based oral Ca preparations have high sucrose content that
may constitute a significant carbohydrate and osmolar load for small preterm infants, and
may be associated with an increase in frequency of bowel movements. Alternately, an
intravenous preparation can be used orally if the fluid volume is tolerated. Treatment of
asymptomatic hypocalcemia can be instituted with oral Ca supplement in the same dosage.
The duration of supplemental Ca therapy depends on the underlying cause of hypocalcemia
and usually lasts several days for most cases of neonatal hypocalcemia, or may be
prolonged as in the case of hypocalcemia caused by malabsorption or hypoparathyroidism.
The serum Ca concentrations should be measured daily during the first few days of
treatment and for 1 or 2 days after discontinuation, until serum tCa and iCa concentrations
are stabilized. Persistently low serum Ca concentrations should prompt further
investigations even in the absence of suspicious history or physical features associated
with pathologic causes of hypocalcemia.
Vitamin D metabolites, 1,25(OH)2D at 0.05 to 0.2 g/kg/day intravenously or orally and
1-hydroxyvitamin D at 0.33 g bid orally, and exogenous PTH have been used in the
treatment of neonatal hypocalcemia. However, there is no practical advantage to the use
of these agents in place of Ca for the treatment of acute hypocalcemia.
For severe persistent hypocalcemia, vitamin D or one of its analogues is often used in
addition to Ca supplementation. The use of 1,25(OH)2D is preferred because it can raise
serum Ca within 1 to 2 days after initiation of therapy and leaves no residual effects within
several days of its discontinuation. Vitamin D has a slower onset of action of 2 to 4 weeks
and the residual effect also lasts several weeks after its discontinuation, thus making
dosage adjustment more difficult.
Successful management of neonatal hypocalcemia also depends on the resolution, if
possible, of the primary cause of hypocalcemia. For example, a poor response to Ca
therapy may often result from concurrent Mg deficiency. Hypomagnesemia, if present,
must be treated to obtain maximal response to Ca therapy. In phosphate-induced
hypocalcemia, high-phosphate formulae and solids should be discontinued, and human
milk or low-phosphate formula should be substituted. Use of aluminum hydroxide gel to
bind intestinal phosphate should be avoided because of potential risk for aluminum toxicity
(116).
Early milk feeding and the use of Ca-containing parenteral nutrition within hours after birth
are the best means to minimize the development and recurrence of hypocalcemia, and
they may negate the need for Ca supplementation. Delaying premature delivery and
minimizing perinatal asphyxia, judicious use of bicarbonate therapy and mechanical
ventilation, for example, during intentional induction of alkalosis in the treatment of
persistent pulmonary hypertension, are also useful measures to minimize neonatal
hypocalcemia. Maintenance of normal maternal vitamin D status with exogenous vitamin D
supplement, if needed, in theory may be helpful in maintaining normal fetal vitamin D
status and may secondarily prevent hypocalcemia in some neonates. Early feeding and
provision of Ca to the gut in the neonate may be important
P.862
to enhancing the ability of vitamin D metabolites to prevent hypocalcemia.
Pharmacologic prevention of neonatal hypocalcemia has focused primarily on the
prophylactic use of Ca salts or vitamin D metabolites. In newborn infants, Ca
supplementation results in sustained lowering of serum IPTH concentrations compared to
unsupplemented controls (27). Theoretically, Ca supplementation may decrease the
metabolic stress from hypocalcemia and minimize the potential for depletion of tissue Ca
stores. Early studies used up to 1.8 to 2.0 mmol (72 to 80 mg)/kg/day of oral Ca
supplement and about half this amount intravenously to prevent hypocalcemia. However, it
should be noted that a similar amount of Ca can be provided from an intake of 150 to 200
mL/kg/day of standard term infant formula or human milk. Standard preterm infant
formula can provide almost 5 mmol (200 mg) of Ca/kg/day, and parenteral nutrition with
1.25 to 1.5 mmol (50 to 60 mg) Ca/100 mL can easily provide 1.5 mmol (60 mg) of Ca/kg/
day. These amounts of Ca are well tolerated as they have been the standard practice in
most neonatal nurseries for more than a decade. Early feeding or parenteral nutrition must
be considered as the best means to prevent neonatal hypocalcemia, particularly for the
preterm infant. Vitamin D3 and its metabolites have been used in attempts to prevent
neonatal hypocalcemia with variable degrees of success. In small preterm infants, serum
Ca was normalized only at pharmacologic doses of 1,25(OH)2D.
Complications of hypocalcemia vary with the clinical manifestations and may be related to
the therapy and underlying pathophysiology. Acute complications are associated with
clinical manifestations, including seizure, apnea, cyanosis and hypoxia, bradycardia, and
hypotension. Therapy-related complications, such as cardiac arrhythmia, arterial spasm,
tissue necrosis, and extravasation of Ca solution, can be avoided by continuous
electrocardiogram monitoring during Ca infusion, avoiding infusion of Ca into the arterial
line, and checking for venous patency before Ca infusion. There is also a risk for metastatic
calcification from aggressive Ca treatment in the presence of hyperphosphatemia. In
situations in which PTH is absent or nonfunctional, its protective hypocalciuric action
cannot occur; therefore, markedly raising the serum Ca concentration may cause
hypercalciuria, renal stones, nephrocalcinosis, and possible renal damage. These
complications were reported during therapy in patients with an activating CaR mutation,
even while the patients were normocalcemic (100). Isolated transient hypocalcemia even
in symptomatic cases has not been associated with long-term sequelae. Long-term
outcomes depend on the underlying causes, for example, patients with 22q11.2 deletion
syndromes frequently have defects of multiple organ systems and neurodevelopmental
delay unrelated to hypocalcemia (94,95,96,97).
Regular clinical follow-up and laboratory monitoring such as for serum Ca and IPTH are
necessary in infants with transient hypoparathyroidism since some of these infants are at
risk for recurrence of hypoparathyroidism and hypocalcemia as late as adolescence
(88,89,90).
Hypercalcemia
Hypercalcemia in infants occurs much less frequently than hypocalcemia. However, it is
increasingly being diagnosed because serum Ca is usually part of a panel of chemistry tests
and because of increasing knowledge of its pathogenesis. Hypercalcemia is present when
serum tCa is greater than 2.75 mmol/L (11 mg/dL) or when iCa is greater than 1.4 mmol/L
(5.6 mg/dL), depending on the particular ion-selective electrode used. In pathologic
hypercalcemia, elevation of serum iCa usually occurs simultaneously with elevation of tCa;
however, elevated tCa may occur without elevation of iCa. Elevation of protein available to
bind Ca (e.g., prolonged application of tourniquet before venipuncture, with resultant
transudation of plasma water into tissues, shown in adult patients with multiple myeloma
and possibly adrenal insufficiency) may result in elevation of serum tCa. A change in serum
albumin of 1 g/dL generally results in a parallel change in tCa of about 0.2 mmol/L.
Conversely, reduced albumin binding of Ca may result in normal serum tCa in the presence
of elevated iCa.
Pathophysiology
Hypercalcemia may occur within hours after birth or be delayed for weeks or months. It
may result from increased intestinal or renal Ca absorption, increased bone turnover, or
iatrogenic causes.
In the neonatal intensive care setting, hypercalcemia is often iatrogenic from inadequate
provision of dietary phosphate during and after hospitalization, as with the use of low- or
no-phosphate parenteral nutrition or human milk without fortifier in very-low-birth-weight
infants (117,118,119,120,121) (Table 36-5). Phosphate deficiency or hypophosphatemia
stimulates CYP1 and synthesis of 1,25(OH)2D, which enhances intestinal absorption and
renal reabsorption of Ca and P. Increased Ca absorbed in the presence of increased 1,25
(OH)2D cannot be deposited in bone in the absence of phosphate and contributes to
hypercalcemia. Hypercalcemia is more likely if there is concomitant use of Ca supplements,
a common practice for the prevention or treatment of hypocalcemia in preterm infants.
Decreased renal Ca excretion in the neonate or from underlying illness also may
exaggerate the extent of hypercalcemia.
Neonatal hyperparathyroidism frequently results in marked hypercalcemia. It may be a
sporadic congenital occurrence, show Mendelian inheritance, or be secondary to maternal
hypocalcemia.
Hereditary primary hyperparathyroidism manifested in neonates is associated with
inactivating mutations of CaR. The severity of hypercalcemia is related to the extent of CaR
mutation. Mild hypercalcemia (serum tCa less than 3.0 mmol/L [12 mg/dL]) associated
with heterozygous mutated CaR is manifested clinically in most patients with familial
hypocalciuric hypercalcemia (FHH). The normal urinary Ca excretion despite hypercalcemia
is an effect of the mutated CaR in the kidneys. Serum PTH is usually
P.863
within the normal range but is higher than expected for the degree of hypercalcemia. FHH
has been reported in patients from 2 hours to 82 years of age and is usually diagnosed in
infants as part of a screening procedure after diagnosis of a family member with
hypercalcemia or familial multiple endocrine neoplasia. It is inherited as an autosomaldominant trait with a high degree of penetrance (122). There usually is significant
hypophosphatemia and a modest increase in serum Mg concentration, and functional
parathyroid glands are needed for full expression. Neonatal hyperparathyroidism
associated with FHH that resolves spontaneously over several months has been reported
(123). More severe hypercalcemia with serum tCa of 3 to 3.3 mmol/L (12 to 13 mg/dL)
has been attributed to coexpression of the normal and mutated CaR, with the latter having
a functional equivalent of a dominant-negative effect. The most marked hypercalcemia
(serum Ca greater than 4 mmol/L [16 mg/dL]) occurs in neonatal severe
hyperparathyroidism with homozygous inactivating germline mutations of the CaR gene.
This severe disorder can be lethal within the first few weeks after birth (124, 125).
TABLE 36-5 PATHOPHYSIOLOGY OF NEONATAL HYPERCALCEMIA
Phosphate deficiency
Low or no phosphate, but calcium-containing parenteral nutrition
Very-low-birth-weight infants fed human milk or, less commonly, standard
formula
Parathyroid related
Hereditary primary hyperparathyroidism
Calcium-sensing receptor inactivating mutations: familial hypocalciuric

hypercalcemia, neonatal severe hyperparathyroidism
Parathyroid hormone receptor activating mutation

Secondary hyperparathyroidism
Maternal: hypocalcemia, renal tubular acidosis
Neonatal: renal tubular acidosis
Parathyroid hormone-related protein-secreting tumors

Vitamin D
Excessive intake
Mother: high-dose vitamin D

Neonate: high-dose vitamin D prophylaxis, overfortification of milk
Increased 1,25-dihydroxyvitamin D
Subcutaneous fat necrosis
Histiocytic disorders, disseminated tuberculosis with septic shock and
hemophagocytic syndrome (?)
Calcitonin response impairment (?) in congenital hypothyroidism

Vitamin A excess
Uncertain pathophysiologic mechanism
Chromosomal/gene abnormalities
Idiopathic infantile hypercalcemia/Williams syndrome

Severe infantile hypophosphatasia
Microdeletion of 4q
Heritable metabolic defect
Blue diaper syndrome

Glycogen storage disease type 1a, congenital lactase or sucrase-isomaltase
deficiency, disaccharidase deficiency
Extracorporeal membrane oxygenation therapy

Activating mutations of the PTH/PTHrP receptor gene in Jansen metaphyseal dysplasia
presumably have the receptor defects in the kidney, bone, and chondrocytes at the growth
plate. The clinical manifestations include postnatal-onset short-limb dwarfism with
radiographic rachitic changes, and mild hypercalcemia occurs in about 50% of the affected
patients (126).
Neonatal hyperparathyroidism may be secondary to various causes of maternal
hypocalcemia including maternal hypoparathyroidism (127) and maternal (128) or
neonatal (129) renal tubular acidosis. Presence of metabolic acidosis independently
increases bone resorption and enhances the renal effects of hyperparathyroidism; the
hypercalcemic effects are augmented by decreased renal excretory capacity of the neonate.
Elevated serum PTHrP and hypercalcemia are found in increasing numbers of infants with a
variety of tumors (130,131,132,133), including malignant hepatic sarcoma, infantile
fibrosarcoma, renal adenoma, and rhabdoid tumors. There is also associated mortality in
some cases, although the relative contribution to death from hypercalcemia versus the
underlying disease is not clear.
Hypercalcemia was reported in 34% of neonates and infants from intermittent high-dose
vitamin D (600,000 IU every 3 to 5 months) prophylaxis (134). Hypercalcemia also has
been reported in infants given human milk with very high vitamin D content (7,000 IU/L),
from high-dose vitamin D therapy for maternal hypoparathyroidism, from milk with
excessive vitamin D fortification from errors during processing, and in preterm infants
given chronic vitamin D supplementation in addition to high-Ca and high-P milk formulae.
Neonates with extensive subcutaneous fat necrosis often have a history of perinatal
asphyxia and may develop hypercalcemia after a period of low or normal serum Ca
concentrations (135). There is an anecdotal report that body cooling for the treatment of
birth asphyxia may augment the development of subcutaneous fat necrosis. Hypercalcemia
is reported to occur between 2 and 16 weeks, most commonly at 6 to 7 weeks after the
development of subcutaneous fat necrosis. Increased prostaglandin E activity, increased
release of Ca from fat and other tissues, and unregulated production of 1,25(OH)2D from
macrophages infiltrating fat necrotic lesions, have been postulated to be responsible for the
hypercalcemia in this condition. Histiocytic disorders and disseminated tuberculosis with
septic shock and hemophagocytic syndrome may be complicated with hypercalcemia in
infants; it is not known if this is also related to nonrenal production of 1,25(OH)2D. Vitamin
A toxicity is associated with hypercalcemia, presumably secondary to the retinoic acid
stimulation of osteoclastic activity and bone resorption. Vitamin A toxicity in infants may
occur at intakes as low as 2,100 IU/100 kcal and can be fatal (136).
Hypercalcemia may develop before and during thyroxine therapy of infants with congenital
agoitrous hypothyroidism
P.864
(137). In theory, deficient CT response to Ca loading or an increased degradation of CT
may be responsible for the hypercalcemia.
Neonatal hypercalcemia is reported in other situations in which the pathophysiology
remains uncertain. Idiopathic infantile hypercalcemia, often considered as part of Williams
syndrome, is associated with varying manifestations including hypercalcemia, mental
retardation, elfin facies, and supravalvular aortic stenosis. There also may be prenatal and
postnatal growth failure. The presence of hypercalcemia in infants with Williams syndrome
is variable, and serum Ca may be normal, but the presence of nephrocalcinosis and softtissue calcifications in some of these infants suggests that hypercalcemia may have
occurred previously. An exaggerated response to pharmacologic doses of vitamin D2 and a
blunted CT response to Ca loading and PTH infusion may contribute to the pathogenesis of
hypercalcemia in idiopathic infantile hypercalcemia. Several genetic defects in idiopathic
infantile hypercalcemia, including hemizygosity at the elastin gene on the long arm of
chromosome 7, have been reported (138,139). No mutations of the CT/CGRP gene has
been detected. However, the cellular mechanism that leads to the phenotypic expression
remains unknown.
Severe infantile hypophosphatasia is associated with hypercalcemia. It is a rare autosomalrecessive disorder associated with decreased synthesis of tissue-nonspecific alkaline
phosphatase from a deletion or point mutation in its gene located on chromosome 1. These
patients have severe bone demineralization, low serum alkaline phosphatase, and elevated
urinary pyrophosphate and phosphoethanolamine. The condition may be lethal in utero or
shortly after birth because of inadequate bony support of the thorax and skull, although
milder phenotypes are compatible with survival to adulthood (140).
Microdeletion of the long arm of chromosome 4 has been associated with hypercalcemia
and cardiac failure (141).
Blue diaper syndrome is a rare familial disorder with impaired intestinal transport of
tryptophan. The blue discoloration of the urine results from the hydrolysis and oxidation of
urinary indican, an end product of intestinal degradation of unabsorbed tryptophan and
hepatic metabolism of its intermediate metabolites. Blue discoloration of the urine has
been reported within weeks after birth, although hypercalcemia and nephrocalcinosis are
not reported until some months after birth. Glycogen storage disease type 1a, congenital
lactase deficiency, and congenital sucrase-isomaltase deficiency with chronic diarrhea have
been associated with hypercalcemia and nephrocalcinosis. Hypercalcemia apparently
resolves without specific treatment following treatment for disaccharidase deficiency.
Transient hypercalcemia occurs in infants during extracorporeal membrane oxygenation
(ECMO) therapy varying in frequency from less than 5% to about 30%, depending on
whether the cut-off point used is greater than 2.5 or 2.25 mmol (12 mg/dL or 11 mg/dL),
respectively (142, 143).
Diagnosis
Neonates with hypercalcemia may be asymptomatic despite the onset of hypercalcemia at
birth. In these cases, there are often delays of weeks or months before diagnosis is made,
coincidental to a chemistry panel screening during the course of other illness or because of
hypercalcemia in another family member.
The presence of a family history of Ca disorders or anatomic anomalies (e.g., elfin facies,
evidence of congenital heart disease, subcutaneous fat necrosis) on physical examination
of the infant may be helpful in arriving at the diagnosis (Table 36-6).
Symptoms and signs are frequently nonspecific and include lethargy, irritability, poor
feeding with or without feeding intolerance, constipation, polyuria, dehydration, and failure
to thrive. Hypertension associated with hypercalcemia may occur in infants, although it
may be in part linked to treatment-related relative fluid overload, as in many infants who
require ECMO therapy.
Management
Therapy depends on the extent of elevation of serum Ca and whether the infant is
symptomatic. For mildly elevated
P.865
serum tCa (less than 12 mg/dL) in the presence of an iatrogenic cause, e.g., phosphatefree parenteral nutrition or the use of Ca supplements without any dietary phosphate
intake, resolution of the underlying cause should also clear up the Ca problem.
Hypercalcemia induced by dietary P deficiency is becoming less common with the
increasing use of commercial fortifier for human milk fed to preterm infants and the use of
high Ca- and high P-containing infant formulae and parenteral nutrition for the preterm
infant. In patients with low serum P concentrations, large amounts of phosphate
supplement may cause hypocalcemia and the possibility of metastatic calcification.
Phosphate supplements given orally may result in diarrhea.
TABLE 36-6 DIAGNOSTIC WORKUP FOR NEONATAL HYPERCALCEMIA
History
Familial or maternal disturbances in calcium (Ca) or phosphorus (P) metabolism

Gestational age, difficult labor, extracorporeal membrane oxygenation (ECMO)
therapy and pre-ECMO therapy
Mother's and infant's intake of Ca, P, vitamins D and A
Physical examination
General examination with focus on growth parameters, hydration status, heart rate,
blood pressure, cornea for band keratopathy (rare)
Associated features (e.g., subcutaneous fat necrosis, elfin facies, congenital heart
disease, developmental delay)
Investigations
Serum total and ionized Ca, magnesium, P, creatinine (Cr), total protein and
albumin, alkaline phosphatase (total and bone specific), simultaneous intact or
whole parathyroid hormone (PTH), 25-hydroxyvitamin D and 1,25
dihydroxyvitamin D
Acid-base status
Urine Ca, P, Cr, amino acids
X-ray of chest, hands, and long bones
Ultrasound of kidneys and abdomen, ophthalmologic examination,

electrocardiogram (shortened QT interval, bradycardia) for complications
If the above do not yield diagnosis, other tests depend on associated history and
symptomatology
Parental (both parents) serum and urine Ca, P, and Cr
Molecular studies
Family screening depends on the primary diagnosis

Serum PTH-related protein and screen for occult tumors
Screen for metabolic defects and unusual dietary supplements
With moderate to severe hypercalcemia, the initial treatment is nonspecific with expansion
of the extracellular fluid compartment (10 to 20 mL/kg of 0.9% sodium chloride
intravenously) and furosemide (2 mg/kg)-induced diuresis (Table 36-7). Care should be
taken to avoid fluid and electrolyte imbalance with careful monitoring of fluid balance and
serum Ca, Mg, sodium, potassium, and osmolality at 6- to 12-hour intervals. Furosemide
therapy may be repeated at 4- to 6-hour intervals. Prolonged diuresis also requires
replacement of Mg losses.
Minimal information is available on the use of hormonal and other drug therapy for
neonatal hypercalcemia. Nonmammalian sources of CT, e.g., salmon CT (4 to 8 IU/kg
every 12 hours subcutaneously or intramuscularly), have greater hypocalcemic effects and
longer durations of action, compared with recombinant hCT. However, salmon CT has
greater potential for allergic reaction and induction of antibody formation. The
hypocalcemic effect decreases after a few days of any CT treatment. Steroid (prednisone
0.5 to 1 mg/kg per day) therapy may result in significant problems including hypertension,
hyperglycemia, and gastrointestinal hemorrhage, and thus is not recommended for longterm therapy. Bisphosphonates, oral etidronate (25 mg bid), and intravenous pamidronate
(0.5 mg/kg) have been used for hypercalcemia in the mother and neonate. Long-term use
of pamidronate in infants and children with osteogenesis imperfecta decreases serum iCa
with a compensatory increase in PTH (144). The effects on growth plate, bone production,
and mineralization remains unknown, and its use should be restricted to acute short-term
therapy. Dialysis in the neonate is not without technical or metabolic complications. Rarely,
parathyroidectomy may be necessary, although it is not always effective. Development of
calcimimetic agents able to amplify the sensitivity of the CaR to iCa and suppress PTH
levels, with a resulting decrease in blood iCa, offer potential for noninvasive therapy of
hypercalcemia.
Treatment for chronic conditions also includes restriction of dietary intake of vitamin D and
Ca and minimizing exposure to sunlight to decrease endogenous vitamin D production. A
low-Ca, low-vitamin D3, low-iron infant formula is available for the management of
hypercalcemia in infants. This formula contains only trace amounts of Ca (less than 10
mg/100 kcal) and no vitamin D. Long-term use of this formula alone will lead to calcium
depletion; iatrogenic vitamin D deficiency is also a concern in this situation, and both can
result in deleterious consequences.
TABLE 36-7 MANAGEMENT OF NEONATAL HYPERCALCEMIA
Acute
Remove etiologic factor, if possible, e.g., discontinue vitamin D and calcium (Ca)
supplements
Intravenous normal saline (20 mL/kg) and loop diuretic (furosemide 2 mg/kg).
Reassess and repeat every 4 to 6 hours as necessary. Monitor fluid balance and
serum Ca, magnesium (Mg), sodium, potassium, phosphorus (P), and osmolality
every 6 to12 hours. Prolonged diuresis may require Mg and potassium replacement
Use lower Ca content milk or parenteral nutrition, if possible, to maintain nutrition

In neonates with low serum P (<1.3 mmol/L; 4 mg/dL), oral phosphate supplement
at 0.5 to 1 mmol (15 to 30 mg) elemental P/kg/day in 4 divided doses given
judiciously may normalize serum P and Ca. In infants not being fed, use parenteral
nutrition containing usual amount of phosphate (1 to 1.5 mmol [31 to 46 mg]/100

mL) but no Ca until serum Ca returned to normal
Minimal data on the use of hormones, e.g., subcutaneous or intramuscular

recombinant human calcitonin (4 to 8 IU/kg every 6 hours), +/-oral glucocorticoid
(prednisone 0.5 to 1 mg/kg/day). Other drugs, e.g., bisphosphonates (oral
etidronate 25 mg bid, intravenous pamidronate 0.5 mg/kg) are experimental
Peritoneal or hemodialysis with a low-Ca dialysate may be considered in severely
symptomatic patients refractory to medical therapy
Parathyroidectomy may be needed when patient is clinically stablized
Maintenance
Depends on underlying cause

Additional general therapy may be needed: low Ca, no vitamin D infant formula
(Calcilo XD, Abbott Laboratories Columbus, OH); keep in minimize sunlight
exposure to lower endogenous synthesis of vitamin D
Complications of hypercalcemia vary with the clinical manifestations at presentation.

Persistent hypercalcemia may result in ectopic calcification, which involves the ectopic
deposition of a solid phase of calcium and phosphate in walls of blood vessels and in
connective tissue about the joints, gastric mucosa, renal parenchyma, and cornea,
especially when accompanied by normal or elevated levels of serum P. Prolonged therapy,
such as severe limitation of Ca and vitamin D intake, may be associated with hypocalcemia
and bone demineralization (145). Severe hypercalcemia can be fatal, although some of the
infants have other potentially lethal underlying conditions. Long-term complications usually
depend on the underlying cause of hypercalcemia, including failure to thrive and
nephrocalcinosis.
Neonatal hypercalcemia may not develop until some weeks after the onset of the insult and
may resolve spontaneously, as in subcutaneous fat necrosis. Therefore, serum Ca should
be monitored at regular intervals in certain situations to determine the onset of
hypercalcemia and to determine
P.866
the continued need for treatment. Family screening for hypercalcemia should be done
unless a specific nonfamilial cause for hypercalcemia is established in the index case.
Hypomagnesemia
Hypomagnesemia is present when serum tMg is less than 0.6 mmol/L (1.5 mg/dL). There
are no data on the level of iMg during hypomagnesemia. Tissue Mg deficiency, however,
may be present despite normal serum Mg concentrations.
Pathophysiology
Decreased tissue accretion of Mg is a major cause of hypomagnesemia (Table 36-8). The
compensatory response at birth of abrupt termination of placental transfer of Mg will be
impaired if there is reduced tissue Mg. The severity and prevalence of hypomagnesemia in
infants of insulin-dependent diabetic mothers is directly related to the severity of maternal
diabetes, which is thought to reflect the severity of maternal Mg deficiency (146). Mg
infusion in infants results in greater increases in serum Ca and PTH in those with initially
low serum Mg concentrations and in children with insulin-dependent diabetes, compared to
normal control subjects.
Maternal hyperparathyroidism has been associated with neonatal hypomagnesemia (147).
Negative Mg balances may occur with hyperparathyroidism, which may account for
neonatal hypomagnesemia. Alternately, neonatal hypoparathyroidism in this situation may
lead to hypomagnesemia from reduced PTH mobilization of bone Mg to extracellular fluid.
In theory, chronic maternal Mg deficiency from any cause may result in decreased tissue
Mg accretion for the fetus. Hypomagnesemia occurs more frequently in infants with
intrauterine growth retardation compared to infants with appropriate weight for gestational
age.
TABLE 36-8 NEONATAL HYPOMAGNESEMIA
Decreased tissue accretion

Infants of mothers with insulin-dependent diabetes or
hyperparathyroidism
Small-for-gestational-age infants
Chronic maternal magnesium deficiency
Decreased absorption
Extensive small-intestine resection
Specific intestinal magnesium malabsorption
Increased loss
Intestinal fistula or diarrhea
Hepatobiliary disorders
Decreased renal tubular reabsorption
Primary: transient receptor potential channel protein mutation, renal tubulopathies

with hypo- or hypercalciuria
Secondary: extracellular fluid compartment expansion, osmotic diuresis, drugs (e.
g., loop diuretic, aminoglycoside, ibuprofen overdose)
Others
Increased phosphate intake
Exchange transfusion with citrated blood
Intestinal resection, particularly of the jejunum and ileum, the major sites of Mg
absorption; malabsorption; and rapid intestinal transit time may lead to Mg deficiency and
hypomagnesemia. Mg content in bile, gastric fluid, and pancreatic secretion varies from 0.2
to 5.0 mmol/L (0.5 to 12 mg/dL). Diarrheal Mg content may be as high as 7.1 mmol/L (17
mg/dL). Thus, chronic losses from diarrhea, intestinal fistula, or enterostomy may be
associated with significant Mg loss.
Infants with congenital biliary atresia and neonatal hepatitis may have low serum Mg
concentrations. This is thought to be partly related to increased aldosterone-related renal
Mg losses.
Hypomagnesemia can occur as a primary defect in Mg transport in the intestine or kidney
or in conjunction with a variety of inherited hypokalemic salt-losing tubulopathies.
Mutation in a member of the long transient receptor potential channel protein TRPM6,
which encodes for a bifunctional protein that combines Ca- and Mg-permeable cation
channel properties with protein kinase activity and is expressed in intestinal epithelia and
kidney tubules, can result in hypomagnesemia (148). Genetic mapping and analysis of a
balanced translocation breakpoint have localized some cases of recessively inherited
familial hypomagnesemia to chromosome 9q (149).
Renal tubulopathies may be subclassified further into a hypercalciuric group consistent with
classic Bartter syndrome, which usually presents in infancy with failure to thrive and
episodes of dehydration. Mutations in PCLN-1, which encodes the renal tight junction
protein paracellin-1 (claudin-16), resulting in impaired tubular reabsorption of Mg and Ca
in the thick ascending limb of Henle's loop have been reported (150). These patients
typically present with urinary tract infection, polyuria, hematuria, hypomagnesemia,
hypercalciuria, nephrocalcinosis, and progressive renal failure. A variant syndrome with
hypocalciuria is thought to present later with short stature, substantially lower serum Mg,
and more episodes of tetany.
Secondary defects in renal tubular reabsorption of Mg may result from extracellular fluid
expansion caused by excessive glucose, sodium, or fluid intake, or from osmotic diuresis,
diuretics such as furosemide, high doses of aminoglycosides such as gentamicin, and
ibuprofen overdose.
Increased phosphate intake may lead to decreased Mg absorption, and infants on highphosphate milk preparations have lowered serum Mg concentrations. Elevation of serum
phosphate concentrations decreases serum Mg. It is not known if these changes are
related to decreased Mg absorption or to the shift of Mg from extracellular to intracellular
compartments. In infants with uremia, serum Mg concentrations may be decreased,
possibly in relation to higher blood phosphate concentrations (151). Patients
P.867
with renal failure, however, become hypermagnesemic at Mg loads that do not affect
people with normal renal function.
Exchange blood transfusions using citrate as an anticoagulant result in complexing of
citrate with Mg, which leads to hypomagnesemia, especially after multiple exchanges
(87,152).
Diagnosis
Suspicion of hypomagnesemia must be confirmed by measurement of serum tMg and iMg,
if available, since clinical manifestations are many and varied and may be indistinguishable
from those of other common neonatal diseases. The less mature the infant, the more
subtle and varied are the clinical manifestations, and the infants frequently are
asymptomatic.
The typical deficit required to produce symptomatic hypomagnesemia is approximately 0.5
to 1.0 mmol (12 to 24 mg)/kg of body weight. However, critical assessment of Mg
deficiency is difficult because more than 99% of total body Mg is found in intracellular
fluids or is complexed in the skeleton. It has been proposed that high Mg retention after a
Mg load may be a test to reflect Mg deficiency. Infants generally retain large amounts of
infused Mg, however, and there are large variations in response; the clinical utility of this
test thus appears limited in infancy. Confirmation of hypomagnesemia as the cause of
clinical manifestations is its reversibility when serum tMg or iMg has been normalized.
Infants with hypomagnesemia associated with malabsorption, or increased losses from the
gut or kidney, also are at risk for concurrent hypocalcemia, hypokalemia, and possible
disturbance of the acid-base status. The loss of other nutrients such as zinc also may be
considerable. Symptoms and signs of hypomagnesemia, which often coexists with
hypocalcemia, may be indistinguishable. Thus, simultaneous measurement of serum Ca
(total and ionized, if available), P, potassium, sodium, chloride and bicarbonate, urea
nitrogen and creatinine, and zinc status may be indicated. Measurement of urine and
intestinal fluid content of Mg also may be helpful in diagnosis and management. Additional
investigations depend on the underlying etiology, and the status of other nutrients also
may need to be considered.
Typically, hypomagnesemia is associated with decreased circulating PTH concentrations,
decreased production of active vitamin D metabolites, in particular 1,25(OH)2D, and
resistance to PTH and 1,25(OH)2D. When hypomagnesemia coexists with hypocalcemia, a
trial infusion of 6 mg of elemental Mg/kg over 1 hour with pre- and postinfusion
measurements of total and ionized Ca and PTH may be helpful in the diagnosis of the
primary defect. An increase in serum PTH after Mg infusion is indicative of
hypoparathyroidism and hypocalcemia secondary to Mg deficiency, whereas no change or a
decrease in serum PTH supports the diagnosis of hypocalcemia unrelated to Mg deficiency.
Management
Clinical manifestations of symptomatic hypomagnesemia such as seizures should be
treated promptly with parenteral Mg. Asymptomatic hypomagnesemia probably also should
be corrected, as Mg potentially can alter important cellular functions and may lead
secondarily to hypocalcemia with its attendant complications. Hypocalcemia occurring
under this circumstance is corrected only when the Mg disturbance is corrected.
Any neonate with seizures should have blood drawn for diagnostic tests before therapy.
The treatment of choice for acute hypomagnesemic seizures is 50% Mg sulfate, 0.05 to 0.1
mL/kg (0.1 to 0.2 mmol/kg or 2.5 to 5.0 mg of elemental Mg/kg) given by slow
intravenous infusion over 15 to 20 minutes, or by intramuscular route. The frequency of
Mg administration depends on the clinical response and rate of increase in serum Mg.
Repeat doses may be given at 2- to 12-hour intervals. Infants receiving parenteral Mg
therapy should receive continuous cardiorespiratory monitoring. Serum Mg concentrations
should be measured daily or more frequently as clinically indicated to evaluate efficacy and
safety until values are stable.
Concomitantly, oral Mg supplements can be started if oral fluids are tolerated. Fifty percent
Mg sulfate can be given at a dose of 0.2 mL/kg per day. In specific Mg malabsorption, daily
oral doses of 1 mL/kg per day may be required. Oral Mg salts are not well absorbed, and
large doses may cause diarrhea. The maintenance Mg supplement should be diluted five-
to six-fold to allow for more frequent administration, maximizing gut absorption and
minimizing side effects. Some oral preparations of Mg (e.g., Mg L-lactate dihydrate),
especially those in a sustained-release form, may have greater bioavailability than other
sources of Mg (e.g., Mg oxide, hydroxide, citrate). However, practical experience with the
use of Mg salts other than Mg sulfate in infancy is limited.
Potassium and zinc deficiency frequently coexists with Mg-deficient states, especially when
there are abnormal gastrointestinal losses or malabsorption. Appropriate replacement
therapy is needed. Treatment of underlying disorders (e.g., closure of gastrointestinal
fistula) should be pursued actively. Chronic Mg therapy is needed if the underlying cause
persists, such as a genetic defect in Mg transport.
Complications of hypomagnesemia vary with clinical manifestations and may be related to
therapy and underlying pathophysiology. Prolonged dietary Mg deprivation in human adults
leads to personality change, tremor, muscle fasciculations, spontaneous carpopedal spasm,
and generalized spasticity, as well as hypomagnesemia, hypocalcemia, and hypokalemia.
Mg depletion in pregnant rats results in fetal mortality, malformations, hypomagnesemia,
decreased skeletal Mg content, hemolytic anemia, hypoproteinemia, and edema.
In infants, acute complications are associated with clinical manifestations including seizure,
apnea, cyanosis, hypoxia, bradycardia, and hypotension. Possible complications of
P.868
intravenous infusion include systemic hypotension and prolongation or even blockade of
sinoauricular or atrioventricular conduction. Isolated transient hypomagnesemia even in
symptomatic cases has not been associated with long-term sequelae. Long-term outcome
of neonatal hypomagnesemia depends on the underlying cause and adequacy of therapy,
and severe neurodevelopmental deficit has been reported, presumably from suboptimal
therapy and recurrent seizures.
Hypermagnesemia
Hypermagnesemia is present when serum Mg is more than 1.04 mmol/L (greater than 2.5
mg/dL). There are insufficient data to define hypermagnesemia based on the measurement
of serum iMg alone.
Pathophysiology
Hypermagnesemia may result from a combination of excessive Mg load and a relatively low
capacity for renal excretion of Mg (Table 36-9). Neonatal hypermagnesemia most
commonly occurs after maternal Mg sulfate administration for preeclampsia. In mothers
given Mg sulfate, serum Mg concentrations have been reported from 1.1 to 5.8 mmol/L
(2.6 to 14.0 mg/dL), with umbilic cord-serum Mg concentrations from 0.8 to 4.8 mmol/L
(2.0 to 11.5 mg/dL) (153,154); concomitant maternal hypocalcemia also may occur
secondarily to decreased serum PTH concentrations. Variations in parenteral Mg intake
(118,119, 155) resulting from high Mg content or high rate of infusion of parenteral
nutrition fluids may result in hypermagnesemia, particularly in critically ill neonates. The
use of Mg-containing antacids or enemas can cause hypermagnesemia. Prematurity and
perinatal asphyxia may aggravate hypermagnesemia, presumably because of decreased
renal Mg excretion.
Diagnosis
Most neonates with hypermagnesemia, particularly preterm infants, are asymptomatic,
even at serum Mg concentrations of more than 1.25 mmol/L (3 mg/dL) (118,119,153155). Clinical signs may not correlate with serum Mg concentrations, although there does
appear to be a correlation with the duration of maternal Mg sulfate therapy, possibly
representing tissue Mg content. With judicious use of Mg sulfate in the mother, however,
signs of Mg intoxication should be uncommon in the infant. In adults with
hypermagnesemia, hypotension and urinary retention occur at serum Mg concentrations of
1.67 to 2.5 mmol/L (4.0 to 6.0 mg/dL); central nervous system depression, hyporeflexia,
and electrocardiographic abnormalities (i.e., increased atrioventricular and ventricular
conduction time) at 2.5 to 5.0 mmol/L (6.0 to 12.0 mg/dL); and respiratory depression,
coma, and cardiac arrest above 5.0 mmol/L (12.0 mg/dL). Clinical signs of neuromuscular
depression with floppiness, lethargy, and respiratory depression are frequent
manifestations of severe neonatal hypermagnesemia. Acute hypotonia, apnea,
hypotension, and refractory bradycardia mimicking septic shock syndrome have been
reported in premature infants accidentally overdosed with Mg in parenteral nutrition (156).
TABLE 36-9 NEONATAL HYPERMAGNESEMIA
Increased load
Maternal magnesium sulfate administration
Neonatal magnesium therapy
Parenteral nutrition
Antacid
Enema
Decreased excretion
Prematurity and asphyxia
Serum Ca concentrations may be normal, increased, or decreased and should be measured

in all infants with suspected hypermagnesemia. Hypermagnesemia might in theory displace
bound Ca in the circulation and lead to elevation of serum iCa concentration.
Hypermagnesemia may suppress PTH and 1,25(OH)2D production and may result in lower
serum Ca concentrations (157,158). Rickets has been reported when maternal Mg therapy
is prolonged (e.g., in tocolysis to prevent preterm delivery). It is speculated that excess Mg
interferes with normal mineralization of fetal bone.
In newborn infants, a delay in passage of meconium (i.e., meconium plug syndrome) has
been thought to be related to neonatal hypermagnesemia. In pregnant and newborn rats
and dogs, however, hypermagnesemia does not have an effect on intestinal motility or the
consistency of meconium.
Management
In asymptomatic infants with normal renal function, serum Mg generally returns to normal
within several days after adequate hydration and nutritional support and elimination of
further Mg intake. These infants should be cared for in a facility that can provide
cardiorespiratory support in case additional complications develop.
For symptomatic infants, intravenous Ca given in the same dosage as for treatment of
hypocalcemia may be useful for acute therapy, since Ca is a direct antagonist of Mg. Loop
diuretics (e.g., furosemide) with adequate fluid intake may hasten Mg excretion. Exchange
blood transfusion with citrated blood is an effective treatment for severely depressed
hypermagnesemic infants. Citrated donor blood is particularly useful because the
complexing action of citrate will expedite removal of Mg from the infant. Peritoneal and
hemodialysis may be considered in refractory patients.
In infants, acute complications are associated with clinical manifestations including
respiratory depression and hypoxia, bradycardia, and hypotension; and potential
P.869
complications associated with therapy such as exchange transfusion. Isolated transient
hypermagnesemia even in symptomatic cases has not been associated with long-term
sequelae.
SKELETAL MANIFESTATIONS OF DISTURBED MINERAL

HOMEOSTASIS
Pathophysiology
Skeletal manifestations of disturbed mineral metabolism in infants usually present as
osteopenia or rachitic changes on standard radiographs. True fetal or congenital rickets is
rare. It may result from severe maternal nutritional osteomalacia associated with Ca and
vitamin D deficiency, maternal hypo- or hyperparathyroidism, or prolonged maternal
treatment with Mg sulfate or phosphate-containing enemas (74,75,159-163) (Table 36-10).
The most frequent cause of skeletal abnormalities in infancy is nutritional deficiency,
although it may occur secondarily to disorders of metabolism of multiple organs including
gut, pancreas, liver, and kidney. In the Western world, rickets and osteopenia presenting
during infancy occur most frequently in small preterm infants and may occur in more than
30% of extremely low-birth-weight (less than 1 kg) infants. The rate of occurrence
depends on the nutrient intake and is associated most frequently with prolonged low-Ca
and/or low-P parenteral nutrition and prolonged intake of soy formula or unfortified human
milk (2). The primary underlying cause in preterm infants appears to be mineral deficiency,
particularly Ca and P, which was demonstrated more than 2 decades ago (164) and
confirmed by many investigators (2); vitamin D deficiency is of secondary importance.
Unfortified human milk fed to preterm infants or standard milk formula for term infants
often have low serum concentrations of 25-OHD. The major reason for the low serum 25OHD in these situations is the increased metabolism of 25-OHD with mineral deficiency.
Unfortunately, vitamin D deficiency secondary to mineral deficiency is still frequently
misdiagnosed as the primary cause of osteopenia, fracture, and rickets in preterm infants,
and is treated with more vitamin D supplement without improving the mineral and general
nutritional support.
TABLE 36-10 RISK FACTORS FOR DEVELOPMENT OF OSTEOPENIA AND RICKETS

IN INFANTS
in utero
Severe maternal nutritional osteomalacia
(i.e., vitamin D +/-calcium deficiency)
Maternal hypoparathyroidism and hyperparathyroidism
Prolonged maternal magnesium or phosphate treatment
Birth weight <1 kg
Postnatal
Prolonged organ dysfunction: intestine, kidney, liver,
pancreas
Nutritional
Preterm infants
Prolonged low calcium and/or low phosphate total parenteral nutrition
Term infants
Vitamin D deficiency
Calcium deficiency
Macrobiotic diet
Soy formula or unfortified human milk for small preterm infants

Chronic loop diuretic therapy given to preterm infants
Toxin contamination: aluminum (?)
Inherited Defects
Renal tubular disorder

Disorders of vitamin D or parathyroid hormone metabolism
Hypo- or hyperphosphatasia
Chronic diuretic therapy, commonly used in infants with bronchopulmonary dysplasia may
aggravate the Ca deficiency. Contamination of nutrients with toxins such as aluminum is
an added risk factor (116). The extent, however, to which each specific risk factor is
responsible for the development of osteopenia, fractures, and rickets is difficult to define in
critically ill infants receiving multiple therapies and suboptimal nutritional support (2).
Isolated nutritional deficiency of copper and ascorbic acid has been reported in preterm
infants with clinical and radiographic manifestations similar to rickets.
In infants born at term, insufficient endogenous production or exogenous supply of vitamin
D is important in the cause of rickets and osteopenia. In one report, almost all children
with vitamin D deficiency had ethnocultural risk factors, and 80% of the mothers were also
vitamin D deficient (165). However, Ca deficiency also is important in older infants and
young children (166). Clinical risk factors thus include prolonged exclusive human milk
feeding, limited sunshine exposure, macrobiotic diet, and prolonged total parenteral
nutrition.
Acquired and heritable forms of rickets that develop despite adequate availability of
vitamin D usually are associated with renal tubular disorders and metabolic defects in
vitamin D and PTH metabolism. Both hypo- and hyperphosphatasia are autosomalrecessive disorders associated with disturbed bone resorption and formation. These causes
of rickets are rare, but their skeletal manifestations may present during infancy.
Diagnosis
A history of significant nutritional defects in the mother, either from self-selected dietary
restriction or cultural habits, e.g., extensive covering of the body with lack of sunlight
exposure or a family history of metabolic disorders and disturbed bone mineral
metabolism, should raise the awareness of the potential for nutritional and skeletal
problems in both the mother and infant.
Infants with congenital rickets may be asymptomatic at birth, leading to a delay in
diagnosis unless investigations are performed as part of the workup for disturbances in
P.870
maternal mineral metabolism. Most postnatal cases of rickets and osteopenia are
diagnosed incidentally during the radiographic investigation of skeletal complications such
as fractures, or nonskeletal problems such as respiratory illness. Radiographic features
such as generalized bone demineralization and widening, cupping, and fraying of the distal
metaphyses confirm the presence of osteopenia and rickets. Some investigators have
derived a scoring system for the assessment of osteopenia and rickets based on the extent
of radiographic changes (167,168,169). The use of dual-energy x-ray absorptiometry
(DXA) allows a more accurate quantification of the degree of bone mineralization
(170,171), although its role in the diagnosis of bone disorders remains to be defined.
Classic clinical features of rickets such as severe skeletal deformities, including
kyphoscoliosis and bowing of the legs, may not be present if the diagnosis is made early in
infancy, before significant growth and weight-bearing have occurred. This is particularly
true for the preterm infant whose skeletal problems typically are diagnosed between 2 and
6 months postnatally (172). With the current practice of early discharge of preterm infants
from neonatal units, it is possible that some nutritional rickets could be diagnosed after
hospital discharge; if there are associated fractures, it may be misdiagnosed as child
abuse, as is the case with fractures from other medical illnesses. Clinical hypotonia is
probably due to a decrease in the intracellular phosphate pool of the skeletal muscle.
Serial biochemical changes (117,173,174) commonly include persistently low serum
inorganic phosphate, elevated serum alkaline phosphatase activity more than five times
the normal adult upper limit, and elevation of other bone turnover markers in serum and
urine. Serum Ca is usually normal except in late severe nutritional vitamin D deficiency
rickets. Vitamin D deficiency as indicated by low or undetectable serum 25-OHD is
possible; however, it is more likely in preterm infants to be secondary to mineral
deficiency. There may be elevated serum 1,25(OH)2D and IPTH. The elevated IPTH and
1,25(OH)2D still may be relatively insufficient to maintain Ca and P homeostasis if the Ca
and P intake remain low. Urine changes may reflect increased serum IPTH with increased
urine P excretion and Ca conservation. However, in chronic P deficiency, urine findings may
reflect changes of P-deficiency-related PTH resistance, in which case urine P would be
minimal while there is calciuria. Measurement of specific trace mineral status may be
useful if deficiency is suspected (175,176). Additional investigations are needed if inherited
renal tubular disorders or disorders of vitamin D and PTH metabolism are suspected.
Treatment and Prevention

Rickets and fractures from nutritional deficiencies respond well to adequate nutrient intake.
The best treatment for nutritional osteopenia, fractures, and rickets is prevention. For
preterm infants, the use of high Ca and high P parenteral nutrition, until establishment of
enteral feeding with human milk containing commercial fortifiers, or formulae designed
specifically for preterm infants is appropriate (2,177,178). Human milk alone is likely to be
inadequate in a number of nutrients including protein, sodium, Ca, P, and possibly other
nutrients for the needs of the very small preterm infant. All human-milk-fed small preterm
infants, particularly those with birth weights less than 1,500 grams, should receive
commercial fortifier containing multiple nutrients in human milk during their hospital stay
and probably posthospital discharge.
The use of Ca and P supplementation alone is inappropriate for the treatment or prevention
of osteopenia with or without fractures or rickets, since bone growth requires protein and
multiple other nutrients for matrix formation and mineralization. In addition, further large
increases in Ca and P intake beyond the current recommended intake is probably not
advisable because of the risks of bezoar and even intestinal obstruction with excessive oral
intake, and hyperphosphatemia with intravenous intake. In preterm infants, most fractures
have significant callus formation at diagnosis and only require splinting support; short-term
analgesia is needed if the fracture is recent and without callus formation.
Human milk or standard infant formulae should provide adequate amounts of Ca and P for
healthy term infants. Other appropriate nutrients also should be introduced during the
latter half of infancy to maintain adequate nutritional status for all infants.
In preterm infants receiving infant formulae with lower Ca and P content (compared to the
preterm infant formulae currently available in the United States), normal vitamin D status
as indicated by serum 25-OHD concentrations has been reported in infants who received
daily supplements of 400 to 2,000 IU vitamin D. However, for enterally fed preterm infants
given adequate volumes of high Ca-, high P-, and vitamin D-fortified preterm infant
formula, or human milk fortified with commercial human milk fortifiers, a daily total intake
of 400 IU vitamin D should be adequate; additional vitamin D supplementation may be
excessive.
Prophylactic vitamin D supplementation has been recommended for all breastfed term
infants (179) since adequate endogenous production of vitamin D cannot be assured with
infants living in varied geographic regions and in families with varied ethnocultural
practices.
For infants who require parenteral nutrition as the major source of nutritional support, 25
to 40 IU vitamin D/dL of parenteral nutrition solution with a maximum total daily intake of
400 IU is sufficient to maintain vitamin D status regardless of the gestational age of the
infant (180).
For infants with established nutritional vitamin D deficiency, a daily supplementation of 400
IU vitamin D (181) in addition to adequate overall nutritional support is also adequate.
Provision of pharmacologic doses of vitamin D to the infants is associated with
hypercalcemia, nephrocalcinosis, and hypertension.
Specific therapies are required for inherited renal tubular disorders and disorders of
vitamin D and PTH metabolism,
P.871
and usually include one or more of the following: Ca, phosphate, and 1,25(OH)2D.
Monitoring and Followup

The goal is for the affected infants to grow normally without residual defect. Regular
clinical assessment and growth measurements are essential. On followup during infancy
(172,173,181), there were no major residual physical deformities with the use of 400 IU
vitamin D for the treatment of rickets for infants born at term or preterm. Skeletal
maturation as assessed by ossification centers of the wrists for preterm infants is similar to
term infants at 1 year of age (172). However, long-term linear growth in the extremely
low-birth-weight infants may remain delayed, suggesting that bone mineral status in the
smallest preterm infants still may be suboptimal (182) despite the relatively uncommon
occurrence of radiographic rickets and fractures on followup. Current data on DXAmeasured total body bone mineral content in the small preterm infant are difficult to
interpret because of inconsistencies in the techniques used among different studies
(183,184).
Biochemical monitoring of nutritional rickets includes measurement of serum Ca, P, and
alkaline phosphatase and avoiding hypercalciuria (less than 0.15 mmol [6 mg/kg/day],
especially in human-milk-fed preterm infants) at 1- to 2-week intervals until stable and
then at 1- to 2-month intervals. Bone turnover markers, IPTH, vitamin D metabolites, and
any other abnormal biochemical parameters should be monitored at 1- to 2-month
intervals. All biochemical monitoring should be continued until standard radiographs show
completion of healing and remodeling of skeletal defects. Radiographs of the wrists and
fracture site(s) should be taken at 2- to 4-month intervals. Standardization of the DXA
measurement in infants with serial measurements at 2- to 3-month intervals should
provide an added means to better understand bone mineral status in the developing
skeleton (185,186). Screening for other affected family members and molecular studies
may be warranted in heritable conditions. Other specific monitoring depends on the
underlying cause of the skeletal defect.
REFERENCES
1. Ziegler EE, O'Donnell AM, Nelson SE, et al. Body composition of the reference fetus.
Growth 1976;40:329-341.
2. Koo WWK, Steichen JJ. Osteopenia and rickets of prematurity. In: Polin R, Fox W, eds.
Fetal and neonatal physiology, 2nd ed. Philadelphia: WB Saunders, 1998:2335-2349.
3. Steichen JJ, Tsang RC, Gratton TL, et al. Vitamin D homeostasis in the perinatal
period: 1,25-dihydroxyvitamin D in maternal, cord and neonatal blood. N Engl J Med
1980;302:315-319.
4. Saggese G, Baroncelli GI, Bertelloni S, et al. Intact parathyroid hormone levels during
pregnancy, in healthy term neonates and in hypocalcemic preterm infants. Acta Paediatr
Scand 1991;80: 36-41.
5. Thiebaud D, Janisch S, Koelbl H, et al. Direct evidence of a parathyroid related protein

gradient between the mother and the newborn in humans. Bone Miner 1993;23:213-221.
6. Seki K, Wada S, Nagata N, et al. Parathyroid hormone-related protein during

pregnancy and the perinatal period. Gynecol Obstet Invest 1994;37:83-86.
7. Tsang RC, Oh W. Neonatal hypocalcemia in low birth weight infants. Pediatrics

1970;45:773-781.
8. Snodgrass GJ, Stemmler L, Went J, et al. Interrelations of plasma calcium, inorganic

phosphate, magnesium and protein over the first week of life. Arch Dis Child
1973;48:279-285.
9. Venkataraman PS, Tsang RC, Chen IW, et al. Pathogenesis of early neonatal
hypocalcemia: studies of serum calcitonin, gastrin, and plasma glucagon. J Pediatr
1987;110:599-603.
10. David L, Salle BL, Putet G, et al. Serum immunoreactive calcitonin in low birth
weight infants. Description of early changes; effect of intravenous calcium infusion;
relationships with early changes in serum calcium, phosphorus, magnesium, parathyroid
hormone and gastrin levels. Pediatr Res 1981;15:803-814.
11. Venkataraman PS, Tsang RC, Steichen JJ, et al. Early neonatal hypocalcemia in
extremely preterm infants: high incidence, early onset, and refractoriness to
supraphysiologic dose of calcitriol. Am J Dis Child 1986;140:1004-1008.
12. Koo WWK, Tsang RC, Poser JW, et al. Elevated serum calcium and osteocalcin levels
from calcitriol in preterm infants. A prospective randomized study. Am J Dis Child
1986;140:1152- 1158.
13. Nelson N, Finnstrom O, Larsson L. Plasma ionized calcium, phosphate and

magnesium in preterm and small for gestational age infants. Acta Paediatr Scand
1989;78:351-357.
14. Soldin SJ, Hicks JM. Calcium and ionized calcium. In: Soldin SJ, Hicks JM, eds.
Pediatric reference ranges. Washington, DC: American Association for Clinical Chemistry
Press, 1995: 38-39.
15. Murthy JN, Hicks JM, Soldin SJ. Evaluation of i-STAT portable clinical analyzer in a
neonatal and pediatric intensive care unit. Clin Biochem 1997;30:385-389.
16. Dollberg S, Bauer R, Lubetzky R, et al. A reappraisal of neonatal blood chemistry

reference ranges using the Nova M electrodes. Am J Perinatol 2001;18:433-440.
17. Loughead JL, Mimouni F, Tsang RC. Serum ionized calcium concentrations in normal
neonates. Am J Dis Child 1988;142:516- 518.
18. Lowenstein FW, Stanton MF. Serum magnesium levels by age, sex and two racial
groups in the United States, First National Health and Nutrition Examination Survey
(NHANES I), 1971-1974. J Am Coll Nutr 1986;5:399-414.
19. Handwerker SM, Altura BT, Jones KY, et al. Maternal-fetal transfer of ionized serum
magnesium during the stress of labor and delivery: a human study. J Am Coll Nutr
1995;14:376-381.
20. Koo B, Sauser K, Hammami M, et al. Neonatal magnesium homeostasis with and
without maternal magnesium treatment. Clin Chem 1996;42:S309.
21. Marcus JC, Valencia GB, Altura BT, et al. Serum ionized magnesium in premature
and term infants. Pediatr Neurol 1998;18:311- 314.
22. Sanders GT, Huijgen HJ, Sanders R. Magnesium in disease: a review with special
emphasis on the serum ionized magnesium. Clin Chem Lab Med 1999;37:1011-1033.
23. Kronenberg HM, Igarashi T, Freeman MW, et al. Structure and expression of the
human parathyroid hormone gene. Recent Prog Horm Res 1986;42:641-663.
24. Rubin LP, Posillico JT, Anast CS, et al. Circulating levels of biologically active and
immunoreactive intact parathyroid hormone in human newborns. Pediatr Res
1991;29:201-207.
25. Venkataraman PS, Blick KE, Fry HD, et al. Postnatal changes in calcium-regulating
hormones in very-low-birth-weight infants. Effect of early neonatal hypocalcemia and
intravenous calcium infusion on serum parathyroid hormone and calcitonin homeostasis.
Am J Dis Child 1985;139:913-916.
26. Mimouni F, Loughead J, Tsang R, et al. Postnatal surge in serum calcitonin

concentrations: no contribution to neonatal hypocalcemia in infants of diabetic mothers.
Pediatr Res 1990;28:493-495.
27. Dilena BA, White GH. The responses of plasma ionised calcium and intact parathyrin
to calcium supplementation in preterm infants. Acta Paediatr Scand 1991;80:1098-1100.
28. Gelbert L, Schipani EA, Juppner H, et al. Chromosomal localization of the parathyroid
hormone/parathyroid hormone-related protein receptor gene to human chromosome
3p21.1-p24.2. J Clin Endocrinol Metab 1994;79:1046-1048.
P.872
29. Brown EM. Four-parameter model of the sigmoidal relationship between parathyroid
hormone release and extracellular calcium concentration in normal and abnormal
parathyroid tissue. J Clin Endocrinol Metab 1983;56:572-581.
30. Mathias RS, Nguyen HT, Zhang MY, et al. Reduced expression of the renal calciumsensing receptor in rats with experimental chronic renal insufficiency. J Am Soc Nephrol
1998;9:2067- 2074.
31. Drueke TB. Abnormal skeletal response to parathyroid hormone and the expression
of its receptor in chronic uremia. Pediatr Nephrol 1996;10:348-350.
32. Chattopadhyay N, Cheng I, Rogers K, et al. Identification and localization of

extracellular Ca2+-sensing receptor in rat intestine. Am J Physiol 1998;274:G122-G130.
33. Eckert RW, Scherubl H, Petzelt C, et al. Rhythmic oscillations of cytosolic calcium in
rat C-cells. Mol Cell Endocrinol 1989;64:267- 270.
34. Zofkova I, Kancheva RL. The relationship between magnesium and calciotropic
hormones. Magnes Res 1995;8:77-84.
35. Toffaletti J, Cooper DL, Lobaugh B. The response of parathyroid hormone to specific
changes in either ionized calcium, ionized magnesium, or protein-bound calcium in
humans. Metabolism 1991;40:814-818.
36. Fatemi S, Ryzen E, Flores J, et al. Effect of experimental human magnesium

depletion on parathyroid hormone secretion and 1,25-dihydroxyvitamin D metabolism. J
Clin Endocrinol Metab 1991;73:1067-1072.
37. Garel JM, Besnard P, Rebut-Bonneton C. C cell activity during the prenatal and
postnatal periods in the rat. Endocrinology 1981;109:1573-1577.
38. Wolfe HJ, DeLellis RA, Volkel EF, et al. Distribution of calcitonin-containing cells in
the normal neonatal human thyroid gland: a correlation of morphology with peptide
content. J Clin Endocrinol Metab 1975;41:1076-1081.
39. Huwyler R, Born W, Ohnhaus EE, et al. Plasma kinetics and urinary excretion of
exogenous human and salmon calcitonin in man. Am J Physiol 1979;236:E15-E19.
40. Samaan NA, Anderson GD, Adam-Mayne ME. Immunoreactive calcitonin in the
mother, neonate, child and adult. Am J Obstet Gynecol 1975;121:622-625.
41. Purdue BW, Tilakaratne N, Sexton PM. Molecular pharmacology of the calcitonin
receptor. Receptors Channels 2002;8:243-255.
42. Juaneda C, Dumont Y, Quirion R. The molecular pharmacology of CGRP and related
peptide receptor subtypes. Trends Pharmacol Sci 2000;21:432-438.
43. Shinki T, Ueno Y, DeLuca HF, et al. Calcitonin is a major regulator for the expression
of renal 25-hydroxyvitamin D3-1-hydroxylase gene in normocalcemic rats. Proc Natl
Acad Sci USA 1999;96:8253-8258.
44. Besnard P, el M'Selmi A, Jousset U, et al. Effects of 1,25-dihydroxycholecalciferol and

calcium on calcitonin mRNA levels in suckling rats. Mol Cell Endocrinol 1991;79:45-52.
45. Wimalawansa SJ, Gunasekera RD, Datta HK. Hypocalcemic actions of amylin amide
in humans. J Bone Miner Res 1992;7: 1113-1116.
46. Holick MF. Photosynthesis of vitamin D in the skin: effect of environmental and lifestyle variables. Fed Proc 1987;46:1876-1882.
47. Mizwicki MT, Norman AW. Two key proteins of the vitamin D endocrine system come
into crystal clear focus: comparison of the x-ray structures of the nuclear receptor for
1, 24(OH)2 vitamin D3, the plasma vitamin D binding protein, and their ligands. J Bone
Miner Res 2003;18:795-806.
48. Bell NH, Shaw S, Turner RT. Evidence that 1,25-dihydroxyvitamin D3 inhibits the
hepatic production of 25-hydroxyvitamin D in man. J Clin Invest 1984;74:1540-1544.
49. Clements MR, Johnson L, Fraser DR. A new mechanism for induced vitamin D
deficiency in calcium deprivation. Nature 1987;325:62-65.
50. Yoshida T, Yoshida N, Monkawa T, et al. Dietary phosphorus deprivation induces 25hydroxyvitamin D3 1-hydroxylase gene expression. Endocrinology 2001;142:17201726.
51. Portale AA, Halloran BP, Morris RC Jr. Physiologic regulation of the serum
concentration of 1,25-dihydroxyvitamin D by phosphorus in normal man. J Clin Invest
1989;83:1494-1499.
52. The HYP Consortium. A gene (PEX) with homologies to endopeptidases is mutated in
patients with X-linked hypophosphatemic rickets. Nat Genet 1995;11:130-136.
53. Shimada T, Kakitani M, Yamazaki Y, et al. Targeted ablation of Fgf23 demonstrates

an essential physiological role of FGF23 in phosphate and vitamin D metabolism. J Clin
Invest 2004;113: 561-568.
54. Weaver VM, Welsh J. 1,25-Dihydroxycholecalciferol and the genesis of

hypocalcaemia in magnesium-deficient chicks. Magnes Res 1990;3:171-177.
55. Adams JS, Sharma OP, Gacad MA, et al. Metabolism of 25-hydroxyvitamin D3 by
cultured pulmonary alveolar macrophages in sarcoidosis. J Clin Invest 1983;72:18561860.
56. Shinki T, Jin CH, Nishimura A, et al. Parathyroid hormone inhibits 25 hydroxyvitamin
D3-24-hydroxylase mRNA expression stimulated by 1,25-dihydroxyvitamin D3 in rat
kidney but not in intestine. J Biol Chem 1992;267:13757-13762.
57. Miyamoto K, Kesterson RA, Yamamoto H, et al. Structural organization of the human
vitamin D receptor chromosomal gene and its promoter. Mol Endocrinol 1997;11:11651179.
58. Yamada S, Yamamoto K, Masuno H, et al. Three-dimensional structure-function

relationship of vitamin D and vitamin D receptor model. Steroids 2001;66:177-187.
59. Norman AW, Bishop JE, Bula CM, et al. Molecular tools for study of genomic and
rapid signal transduction responses initiated by 1,25(OH)2-vitamin D3. Steroids
2002;67:457-466.
60. Verity CM, Burman D, Beadle PC, et al. Seasonal changes in perinatal vitamin D
metabolism: maternal and cord blood biochemistry in normal pregnancies. Arch Dis Child
1981;56:943-948.
61. Hollis BW, Pittard WB III. Evaluation of the total fetomaternal vitamin D relationship
at term: evidence for racial differences. J Clin Endocrinol Metab 1984;59:652-657.
62. Nehama H, Weintroub S, Eisenberg Z, et al. Seasonal variation in paired maternal

newborn serum 25 hydroxyvitamin D and 24,25 dihydroxyvitamin D concentrations in
Israel. Isr J Med Sci 1987;23:274-277.
63. Overbergh L, Decallonne B, Valckx D, et al. Identification and immune regulation of

25-hydroxyvitamin D-1--hydroxylase in murine macrophages. Clin Exp Immunol
2000;120:139-146.
64. Martin TJ, Moseley JM, Williams ED. Parathyroid hormone-related protein: hormone
and cytokine. J Endocrinol 1997;154: S23-S37.
65. MacIsaac RJ, Caple JW, Danks JA, et al. Ontogeny of parathyroid hormone-related
protein in the ovine parathyroid gland. Endocrinology 1991;129:757-764.
66. Abbas SK, Ratcliff WA, Moniz C, et al. The role of parathyroid hormone-related
protein in calcium homeostasis in the fetal pig. Exp Physiol 1994;79:527-536.
67. Bilezikian JP. Clinical utility of assays for parathyroid hormone-related protein. Clin
Chem 1992;38:179-181.
68. Dvir R, Golander A, Jaccard N, et al. Amniotic fluid and plasma levels of parathyroid
hormone-related protein and hormonal modulation of its secretion by amniotic fluid cells.
Eur J Endocrinol 1995;133:277-282.
69. Law F, Moate PJ, Leaver DD, et al. Parathyroid hormone-related protein in milk and
its correlation with bovine milk calcium. J Endocrinol 1991;128:21-26.
70. Salsburey DJ, Brown DR. Effect of parenteral calcium treatment on blood pressure
and heart rate in neonatal hypocalcemia. Pediatrics 1982;69:605-609.
71. Mirro R, Brown DR. Parenteral calcium treatment shortens the left ventricular systolic
time intervals of hypocalcemic neonates. Pediatr Res 1984;18:71-73.
72. Venkataraman PS, Wilson DA, Sheldon RE, et al. Effect of hypocalcemia on cardiac
function in very-low-birth-weight preterm neonates: studies of blood ionized calcium,
echocardiography and cardiac effect of intravenous calcium therapy. Pediatrics
1985;76:543-550.
73. Broner CW, Stidham GL, Westenkirchner DF, et al. Hypermagnesemia and
hypocalcemia as predictors of high mortality in critically ill pediatric patients. Crit Care
Med 1990;18:921-928.
74. Hanukoglu A, Chalen S, Kowardski AA. Late onset hypocalcemia, rickets and
hypoparathyroidism in an infant of a mother with hyperparathyroidism. J Pediatr
1988;112:751-754.
75. Thomas AK, McVie R, Levine SN. Disorders of maternal calcium metabolism
implicated by abnormal calcium metabolism in the neonate. Am J Perinatol 1999;16:515520.
76. Robertson WC Jr. Calcium carbonate consumption during pregnancy: an unusual

cause of neonatal hypocalcemia. J Child Neurol 2002;17:853-855.
P.873
77. Ravid A, Koren R, Rotem C, et al. Mononuclear cells from human neonates are
partially resistant to the action of 1,25 dihydroxyvitamin D. J Clin Endocrinol Metab
1988;67:755-759.
78. Dincsoy MY, Tsang RC, Laskarzewski P, et al. Serum calcitonin response to
administration of calcium in newborn infants during exchange blood transfusion. J
Pediatr 1982;100:782-786.
79. Mimouni F, Tsang, RC, Hertzberg VS, et al. Polycythemia, hypomagnesemia and
hypocalcemia in infants of diabetic mothers. Am J Dis Child 1986;140:798-800.
80. Banerjee S, Mimouni FB, Mehta R, et al. Lower whole blood ionized magnesium
concentrations in hypocalcemic infants of gestational diabetic mothers. Magnes Res
2003;16:127-130.
81. Sarkar S, Watman J, Seigel WM, et al. A prospective controlled study of neonatal
morbidities in infants born at 36 weeks or more gestation to women with diet-controlled
gestational diabetes (GDM-class Al). J Perinatol 2003;23:223-228.
82. Schwartz R, Teramo KA. Effects of diabetic pregnancy on the fetus and newborn.
83. Specker B, Tsang R, Ho M, et al. Low serum calcium and high parathyroid hormone
levels in neonates fed humanized cow's milk-based formula. Am J Dis Child
1991;145:941-945.
84. Venkataraman PS, Tsang RC, Greer FR, et al. Late infantile tetany and secondary
hyperparathyroidism in infants fed humanized cow milk formula. Longitudinal follow-up.
Am J Dis Child 1985;139:664-668.
85. Perlman JM. Fatal hyperphosphatemia after oral phosphate overdose in a premature
infant. Am J Health Syst Pharm 1997;54: 2488-2490.
86. Davis RF, Eichner JM, Bleyuer WA, et al. Hypocalcemia, hyperphosphatemia and
dehydration following a single hypertonic phosphate enema. J Pediatr 1977;90:484-485.
87. Dincsoy MY, Tsang RC, Laskarzewski P, et al. The role of postnatal age and
magnesium on parathyroid hormone responses during exchange blood transfusion in
the newborn period. J Pediatr 1982;100:277-283.
88. Bainbridge R, Mughal Z, Mimouni F, et al. Transient congenital hypoparathyroidism:

how transient is it? J Pediatr 1988;111: 866-868.
89. Kooh SW, Binet A. Partial hypoparathyroidism: a variant of transient congenital

hypoparathyroidism. Am J Dis Child 1991;145: 877-880.
90. Greig F, Paul E, DiMartino-Nardi J, et al. Transient congenital hypoparathyroidism:

resolution and recurrence in chromosome 22q11 deletion. J Pediatr 1996;128:563-567.
91. Arnold A, Horst SA, Gardella TJ, et al. Mutation of the signal peptide-encoding region
of the preproparathyroid hormone gene in familial isolated hypoparathyroidism. J Clin
Invest 1990;86: 1084-1087.
92. Bilous RW, Murty G, Parkinson DB, et al. Autosomal dominant familial
hypoparathyroidism, sensorineural deafness, and renal dysplasia. N Engl J Med
1992;327:1069-1074.
93. Sunthornthepvarakul T, Churesigaew S, Ngowngarmratana S. A novel mutation of

the signal peptide of the preproparathyroid hormone gene associated with autosomal
recessive familial isolated hypoparathyroidism. J Clin Endocrinol Metab 1999;84: 37923796.
94. Taylor SC, Morris G, Wilson D, et al. Hypoparathyroidism and 22q11 deletion
syndrome. Arch Dis Child 2003;88:520-522.
95. Swillen A, Devriendt K, Legius E, et al. The behavioural phenotype in velo-cardiofacial syndrome (VCFS): from infancy to adolescence. Genet Couns 1999;10:79-88.
96. Gerdes M, Solot C, Wang PP, et al. Taking advantage of early diagnosis: preschool
children with the 22q11. 2 deletion. Genet Med 2001;3:40-44.
97. Greenhalgh KL, Aligianis IA, Bromilow G, et al. 22q11 deletion: a multisystem
disorder requiring multidisciplinary input. Arch Dis Child 2003;88:523-524.
98. Keppen LD, Fasules JW, Burks AW, et al. Confirmation of autosomal dominant
transmission of the DiGeorge malformation complex. J Pediatr 1988;113:506-508.
99. Baron J, Winer KK, Yanovski JA, et al. Mutations in the Ca+2-sensing receptor gene
cause autosomal dominant and sporadic hypoparathyroidism. Hum Mol Genet
1996;5:601-606.
100. Pearce SH, Williamson C, Kifor O, et al. A familial syndrome of hypocalcemia with
hypercalciuria due to mutations in the calcium sensing receptor. N Engl J Med
1996;335:1115-1122.
101. Zhang P, Jobert AS, Couvineau A, et al. A homozygous inactivating mutation in the
parathyroid hormone/parathyroid hormone-related peptide receptor causing Blomstrand
chondrodysplasia. J Clin Endocrinol Metab 1998;83:3365-3368.
102. Ringel MD, Schwindinger WF, Levine MA. Clinical implications of genetic defects in
G proteins: The molecular basis of McCune Albright syndrome and Albright hereditary
osteodystrophy. Medicine 1996;75:171-184.
103. Linglart A, Carel JC, Garabedian M, et al. GNAS1 lesions in

pseudohypoparathyroidism Ia and Ic: genotype phenotype relationship and evidence of
the maternal transmission of the hormonal resistance. J Clin Endocrinol Metab
2002;87:189-197.
104. Bastepe M, Lane AH, Juppner H. Paternal uniparental isodisomy of chromosome

20q and the resulting changes in GNAS1 methylation as a plausible cause of
pseudohypoparathyroidism. Am J Hum Genet 2001;68:1283-1289.
105. Minagawa M, Yasuda T, Kobayashi Y, et al. Transient pseudohypoparathyroidism of

the neonate. Eur J Endocrinol 1995;133: 151-155.
106. Mimouni F, Mimouni CP, Loughead JL, et al. A case control study of hypocalcemia in
high risk neonates: racial, but no seasonal differences. J Am Coll Nutr 1991;10:196-199.
107. Srinivasan M, Abinun M, Cant AJ, et al. Malignant infantile osteopetrosis presenting
with neonatal hypocalcemia. Arch Dis Child Fetal Neonatal Ed 2000;83:F21-F23.
108. Crook MA, Hally V, Panteli JV. The importance of the refeeding syndrome. Nutrition
2001;17:632-637.
109. Weinsier RL, Krumdieck CL. Death resulting from overzealous total parenteral
nutrition: the refeeding syndrome. Am J Clin Nutr 1980;34:393-399.
110. Foldenauer A, Vossbeck S, Pohlandt F. Neonatal hypocalcaemia associated with

rotavirus diarrhoea. Eur J Pediatr 1998;157: 838-842.
111. Wasant P, Matsumoto I, Naylor E, et al. Mitochondrial fatty acid oxidation disorders
in Thai infants: a report of 3 cases. J Med Assoc Thai 2002;85(Suppl 2):S710-S719.
112. Sarici SU, Serdar MA, Erdem G, et al. Evaluation of plasma ionized magnesium
levels in neonatal hyperbilirubinemia. Pediatr Res 2004;55:243-247.
113. Romagnoli C, Polidori G, Cataldi L, et al. Phototherapy induced hypocalcemia. J

Pediatr 1979;94:815-816.
114. Oleske JM. Experience with 118 infants born to narcotic-using mothers: does a
lower serum ionized calcium level contribute to the symptoms of withdrawal? Clin Pediatr
1977;16:418-423.
115. Broner CW, Stidham GL, Westenkirchner DF, et al. A prospective, randomized,
double-blind comparison of calcium chloride and calcium gluconate therapies for
hypocalcemia in critically ill children. J Pediatr 1990;117:986-989.
116. Koo WWK, Kaplan LA. Aluminum and bone disorders: with specific reference to
aluminum contamination of infant nutrients. J Am Coll Nutr 1988;7:199-214.
117. Koo WWK, Antony G, Stevens LH. Continuous nasogastric phosphorus infusion in
hypophosphatemic rickets of prematurity. Am J Dis Child 1984;138:172-175.
118. Koo WWK, Tsang RC, Steichen JJ, et al. Parenteral nutrition for infants: effect of
high versus low calcium and phosphorus content. J Pediatr Gastroenterol Nutr
1987;6:96-104.
119. Koo WWK, Tsang RC, Succop P, et al. Minimal vitamin D and high calcium and
phosphorus needs of preterm infants receiving parenteral nutrition. J Pediatr
120. Lyon AJ, McIntosh N, Wheeler K, et al. Hypercalcemia in extremely low birthweight
infants. Arch Dis Child 1984;59: 1141-1144.
121. Kimura S, Nose O, Seino Y, et al. Effects of alternate and simultaneous

administrations of calcium and phosphorus on calcium metabolism in children receiving
total parenteral nutrition. J Parenter Enteral Nutr 1986;10:513-516.
122. Pearce SH, Trump D, Wooding C, et al. Calcium-sensing receptor mutations in

familial benign hypercalcaemia and neonatal hyperparathyroidism. J Clin Invest
1995;96:2683-2692.
123. Wilkinson H, James J. Self limiting neonatal primary hyperparathyroidism

associated with familial hypocalciuric hypercalcemia. Arch Dis Child 1993;69:319-321.
124. Ross AJ, Cooper A, Attie MF, et al. Primary hyperparathyroidism in infancy. J
Pediatr Surg 1986;21:493-499.
125. Pollak M, Chou Y, Marx S, et al. Familial hypocalciuric hypercalcemia and neonatal
severe hyperparathyroidism. Effects of mutant gene dosage on phenotype. J Clin Invest
1994;93: 1108-1112.
P.874
126. Schipani E, Langman C, Parfitt AM, et al. Constitutively activated receptors for
parathyroid hormone and parathyroid hormone-related peptides in Jansen's metaphyseal
chondrodysplasia. N Engl J Med 1996;335:708-714.
127. Loughead J, Mughal F, Mimouni F, et al. Spectrum and natural history of congenital
hyperparathyroidism secondary to maternal hypocalcemia. Am J Perinatol 1990;7:350355.
128. Savani R, Mimouni F, Tsang R. Maternal and neonatal hyperparathyroidism as a

consequence of maternal renal tubular acidosis. Pediatrics 1993;91:661-663.
129. Rodriguez-Soriano J, Garcia-Fuentes M, Vallo A, et al. Hypercal-cemia in neonatal

distal renal tubular acidosis. Pediatr Nephrol 2000;14:354-355.
130. Lakhdir F, Lawson D, Schatz D. Fatal parathyroid hormone-related protein-induced

humoral hypercalcemia of malignancy in a 3-month old infant. Eur J Pediatr
1994;153:718-720.
131. Michigami T, Yamato H, Mushiake S, et al. Hypercalcemia associated with infantile

fibrosarcoma producing parathyroid hormone-related protein. J Clin Endocrinol Metab
1996;81:1090-1095.
132. Mahoney C, Cassady C, Weinberger E, et al. Humoral hypercalcemia due to an

occult renal adenoma. Pediatr Nephrol 1997;11: 339-342.
133. Amar AM, Tomlinson G, Green DM, et al. Clinical presentation of rhabdoid tumors of
the kidney. J Pediatr Hematol Oncol 2001;23:105-108.
134. Markestad T, Hesse V, Siebenhuner M, et al. Intermittent high-dose vitamin D

prophylaxis during infancy: effect on vitamin D metabolites, calcium, and phosphorus.
Am J Clin Nutr 1987;46: 652-658.
135. Hicks MJ, Levy ML, Alexander J, et al. Subcutaneous fat necrosis of the newborn
and hypercalcemia: case report and review of the literature. Pediatr Dermatol
1993;10:271-276.
136. Bush ME, Dahms BB. Fatal hypervitaminosis A in a neonate. Arch Pathol Lab Med
1984;108:838-842.
137. Tau C, Garabedian M, Farriaux JP, et al. Hypercalcemia in infants with congenital
hypothyroidism and its relation to vitamin D and thyroid hormones. J Pediatr
1986;109:808-814.
138. Telvi L, Pinard J, Ion R, et al. De novo t(X; 21) (q28; q11) in a girl with phenotypic
features of Williams-Beuren syndrome. J Med Genet 1992;29:747-749.
139. Nickerson E, Greenberg F, Keating MT, et al. Deletions of the elastin gene at
7q11.23 occur in 90% of patients with Williams syndrome. Am J Hum Genet
1995;56:1156-1161.
140. Whyte MP. Hypophosphatasia and the role of alkaline phosphatase in skeletal
mineralization. Endocr Rev 1994;15:439-461.
141. Strehle EM, Ahmed OA, Hameed M, et al. The 4q- syndrome. Genet Couns
2001;12:327-339.
142. Zwischenberger JB, Nguyen TT, Upp JR Jr, et al. Complications of neonatal
extracorporeal membrane oxygenation. Collective experience from the Extracorporeal
Life Support Organizations. J Thorac Cardiovas Surg 1994;107:838-848.
143. Fridriksson JH, Helmrath MA, Wessel JJ, et al. Hypercalcemia associated with
extracorporeal life support in neonates. J Pediatr Surg 2001;36:493-497.
144. Rauch F, Plotkin H, Travers R, et al. Osteogenesis imperfecta types I, III, and IV:
effect of pamidronate therapy on bone and mineral metabolism. J Clin Endocrinol Metab
2003;88:986-992.
145. Mathias RS. Rickets in an infant with Williams syndrome. Pediatr Nephrol
2000;14:489-492.
146. Mimouni F, Tsang RC. Perinatal magnesium metabolism: personal data and
challenges for the 1990s. Magnes Res 1991;4: 109-117.
147. Monteleone JA, Lee JB, Tashjian AH Jr, et al. Transient neonatal hypocalcemia,
hypomagnesemia and high serum parathyroid hormone with maternal
hyperparathyroidism. Ann Intern Med 1975;82:670-672.
148. Schlingmann KP, Weber S, Peters M, et al. Hypomagnesemia with secondary

hypocalcemia is caused by mutations in TRPM6, a new member of the TRPM gene family.
Nat Genet 2002;31: 166-170.
149. Walder RY, Shalev H, Brennan TM, et al. Familial hypomagnesemia maps to
chromosome 9q, not to the X chromosome: genetic linkage mapping and analysis of a
balanced translocation breakpoint. Hum Mol Genet 1997;6:1491-1497.
150. Weber S, Schneider L, Peters M, et al. Novel paracellin-1 mutations in 25 families

with familial hypomagnesemia with hypercalciuria and nephrocalcinosis. J Am Soc
Nephrol 2001;12:1872-1881.
151. Ghazali S, Hallett RJ, Barratt TM. Hypomagnesemia in uremic infants. J Pediatr
1972;81:747-750.
152. Bajpai PC, Sugden D, Stern L, et al. Serum ionic magnesium in exchange
transfusion. J Pediatr 1967;70:193-199.
153. Stone SR, Pritchard JA. Effect of maternally administered magnesium sulfate on the
neonate. Obstet Gynecol 1970;35:574-577.
154. Lipsitz PJ. The clinical and biochemical effects of excess magnesium in the newborn.
Pediatrics 1971;47:501-509.
155. Koo WWK, Fong T, Gupta JM. Parenteral nutrition in infants. Aust Paediatr J
1980;16:169-174.
156. Ali A, Walentik C, Mantych GJ, et al. Iatrogenic acute hypermagnesemia after total
parenteral nutrition infusion mimicking septic shock syndrome: two case reports.
Pediatrics 2003;112: e70-e72.
157. Cholst IN, Steinberg SF, Tropper PJ, et al. The influence of hypermagnesemia on
serum calcium and parathyroid hormone levels in human subjects. N Engl J Med
1984;310:1221-1225.
158. Donovan EF, Tsang RC, Steichen JJ, et al. Neonatal hypermagnesemia: effect on
parathyroid hormone and calcium homeostasis. J Pediatr 1980;96:305-310.
159. Lamm CI, Norton KI, Murphy RJ, et al. Congenital rickets associated with
magnesium sulfate infusion for tocolysis. J Pediatr 1988;113:1078-1082.
160. Russell JGB, Hill LF. True fetal rickets. Br J Radiol 1974;47: 732-734.
161. Park W, Paust H, Kaufmann HJ, et al. Osteomalacia of the motherrickets of the
newborn. Eur J Pediatr 1987;146:292- 293.
162. Anatoliotaki M, Tsilimigaki A, Tsekoura T, et al. Congenital rickets due to maternal

vitamin D deficiency in a sunny island of Greece. Acta Paediatr 2003;92:389-391.
163. Rimensberger P, Schubiger G, Willi U. Congenital rickets following repeated

administration of phosphate enemas in pregnancy: a case report. Eur J Pediatr
1992;151:54-56.
164. Steichen JJ, Gratton T, Tsang RC. Osteopenia of prematurity: the cause and
possible treatment. J Pediatr 1980;96:528-534.
165. Nozza JM, Rodda CP. Vitamin D deficiency in mothers of infants with rickets. Med J
Aust 2001;175:253-255.
166. Thacher TD, Fischer PR, Pettifor JM, et al. A comparison of calcium, vitamin D, or
both for nutritional rickets in Nigerian children. N Engl J Med 1999;341:563-568.
167. Koo WWK, Gupta JM, Nayanar VV, et al. Skeletal changes in premature infants.
Arch Dis Child 1982;57:447-452.
168. James JR, Congdon PJ, Truscott J, et al. Osteopenia of prematurity. Arch Dis Child
1986;61:871-876.
169. Lyon AJ, McIntosh N, Wheeler K, et al. Radiological rickets in extremely low
birthweight infants. Pediatr Radiol 1987;17:56-58.
170. Koo WWK, Hammami M, Hockman EM. Use of fan beam dual energy x-ray
absorptiometry to measure body composition of piglets. J Nutr 2002;132:1380-1383.
171. Chauhan S, Koo WWK, Hammami M, et al. Fan beam dual energy x-ray
absorptiometry body composition measurements in piglets. J Am Coll Nutr 2003;22:408414.
172. Koo WWK, Sherman R, Succop P, et al. Fractures and rickets in very low birth
weight infants: conservative management and outcome. J Pediatr Orthop 1989;9:326330.
173. Koo WWK, Sherman R, Succop P, et al. Serum vitamin D metabolites in very low
birth weight infants with and without rickets and fractures. J Pediatr 1989;114:10171022.
174. Koo WWK. Laboratory assessment of nutritional metabolic bone disease in infants.
Clin Biochem 1996;29:429-438.
175. Koo WWK, Succop P, Hambidge KM. Serum alkaline phosphatase and serum zinc
concentrations in preterm infants with rickets and fractures. Am J Dis Child
1989;143:1342-1345.
176. Koo WWK, Succop P, Hambidge KM. Sequential concentrations of copper and
ceruloplasmin in serum from preterm infants with rickets and fractures. Clin Chem
1991;37:556-559.
177. Koo WWK, Warren L. Calcium and bone health in infants. Neonatal Netw
2003;22:23-37.
178. Koo WWK, McLaughlin K, Saba M. Nutrition support for the neonatal intensive care
patients. In: The ASPEN nutrition support practice manual. Washington DC: American
Society for Parenteral and Enteral Nutrition (in press).
179. Gartner LM, Greer FR, for the Section on Breastfeeding and Committee on
Nutrition, American Academy of Pediatrics: prevention of rickets and vitamin D
deficiency: new guidelines for vitamin D intake. Pediatrics 2003;111:908-910.
P.875
180. Koo WWK, Tsang RC, Steichen JJ, et al. Vitamin D requirement in infants receiving
parenteral nutrition. J Parenter Enteral Nutr 1987;11:172-176.
181. Venkataraman PS, Tsang RC, Buckley DD, et al. Elevation of serum 1,25dihydroxyvitamin D in response to physiologic doses of vitamin D in vitamin D deficient
182. Hack M, Taylor HG, Klein N, et al. School-age outcomes in children with birth
weights under 750 g. N Engl J Med 1994;331: 753-759.
183. Koo WWK, Hockman EM, Hammami M. Dual energy X ray absorptiometry
measurements in small subjects: conditions affecting clinical measurements. J Amer Coll
Nutr 2004;23: 212-219.
184. Koo WWK, Hammami M, Shypailo RJ, et al. Bone and body composition
measurements of small subjects: discrepancies from software for fan-beam dual energy
X-ray absorptiometry. J Amer Coll Nutr (in press).
185. Koo WWK, Walters J, Bush AJ, et al. Dual energy x-ray absorptiometry studies of
bone mineral status in newborn infants. J Bone Miner Res 1996;11:997-1002.
186. Koo WWK, Bush AJ, Walters J, et al. Postnatal development of bone mineral status
during infancy. J Amer Coll Nutr 1998;17: 65-70.
Chapter 37
Carbohydrate Homeostasis
Edward S. Ogata
Glucose homeostasis results from the net balance between systemic organ requirements and the
production and regulation of glucose. The neonate's ability to maintain glucose homeostasis is less
well developed than the older child or adult because it is in a metabolic transition period. The
abrupt switch from intrauterine life, in which glucose and metabolic fuels are provided in a wellregulated manner, to a situation in which meals are intermittent and which necessitates regulation
of exogenous glucose and production of endogenous glucose. As the capability to perform these
functions continues to develop in the neonate, clinical disorders that can afflict the neonate may
perturb this balance, resulting in hypoglycemia or hyperglycemia. In addition, antecedent
intrauterine events can alter the development of glucoregulatory capabilities in the fetus, resulting
in altered neonatal glucose homeostasis.
To understand the processes responsible for glucose homeostasis in the normal neonate, an
appreciation of the development of glucoregulatory capabilities in the fetus is necessary. This
chapter reviews this information and its relation to the clinical disorders associated with altered
neonatal glucose homeostasis. Information on the perinatal aspects of diabetes in pregnancy also
is presented.
MATERNAL METABOLISM DURING PREGNANCY

Although the metabolic alterations that develop in the pregnant woman throughout gestation favor
the growth and development of the fetus, the first half of gestation is also a critical period for
maternal anabolism. The increased calories ingested by the woman during early gestation sustain
fetal growth and facilitate maternal fat deposition. This is important preparation for the second half
of gestation, a period of exponential fetal growth during which these maternal stores are mobilized
to meet fetal needs. The storage of maternal energy stores is facilitated by the increased secretion
of insulin that occurs in women with normal carbohydrate metabolism (1,2,3,4,5).
Changes in maternal insulin secretion and action are dramatic during the second half of pregnancy.
During the first half, basal insulin secretion is normal, and insulin secretion mediates the laying
down of metabolic fuels (6). In the second half of gestation, insulin secretion is greatly
exaggerated (7,8). Intravenous glucose tolerance testing indicates that by the third trimester,
normal pregnant women secrete three times the insulin than they do in the nonpregnant state (1).
The development of insulin resistance during the second half of pregnancy is responsible for this
exaggerated insulin secretion. Responsiveness to insulin develops 40% to 60% during the second
half of gestation (7,8). The conceptus causes this insulin resistance in part by the generation of
human placental lactogen, progesterone, and estrogen which become available in ever increasing
amounts as gestation progresses (9,10,11). These directly antagonize the effect of maternal
insulin. Insulin resistance increases in liver, adipose tissue, and skeletal muscle (12). Insulin
binding appears to be normal during late gestation, indicating that insulin resistance is mediated
by events distal to receptor/insulin binding. It is possible that tyrosine kinase inhibition may also
contribute to insulin resistance (12,13).
Preexisting maternal diabetes potentiates the effects of the antiinsulin factors, causing an
excessive provision of glucose and other metabolic fuels to the fetus. This is the basic perturbation
responsible for the problems of the infant of the diabetic mother (Fig. 37-1) (1,14).
Metabolic fuel availability is guaranteed to the fetus even during brief maternal fasting. After an
overnight fast, pregnant women have significantly lower plasma glucose concentrations than
fasted nongravid women (8,15); however, glucose production in the mother is significantly
increased (16,17). This increased production ensures the provision of glucose to the fetus.
Prolonged maternal fasting does alter fuel provision to the fetus. As the fast progresses, maternal
ketogenesis progressively increases (15,18). The human fetal brain at early gestation can use
ketones (Fig. 37-2) (19). However, this ability to use an alternative fuel may be harmful rather
than beneficial to the fetus. Offspring of mothers who were ketotic during pregnancy appear to
have an increased incidence of cognitive and psychomotor delay at 3 and
P.877
5 years of age (20). The mechanisms responsible for this potential association between ketone
body oxidation and impaired brain function are unknown. Because of this uncertainty, maternal
fasting, even foregoing breakfast, should be avoided during pregnancy.
Figure 37-1 Plasma glucose changes after glucose challenge in nongravid, gravid, and gravid
diabetic women. Compared to the nongravid woman, the pregnant woman with normal
carbohydrate metabolism demonstrates delayed glucose clearance from midgestation onward as a
result of antiinsulin factors that develop during pregnancy. The delay in glucose clearance from
the maternal circulation assures glucose provision to the fetus, particularly during the
postprandial period. The blunting of maternal glucose clearance is exaggerated by the
counterinsulin factors in the woman with diabetes mellitus. The decreased clearance of glucose
and other metabolic fuels stimulates fetal insulin production and is responsible for many of the
problems of the infant of the diabetic mother.
DEVELOPMENT OFGLUCOSE-PRODUCING AND

GLUCOREGULATORYCAPABILITIES IN THE FETUS
To understand the problems of neonatal glucose homeostasis, the development of glucose
production and regulatory capabilities in the fetus must be understood.
Glycogen
The third trimester of the human pregnancy is the first period in gestation during which some of
the energy and substrate available to the fetus can be channeled from meeting needs for ongoing
growth and development to energy storage. As the third trimester progresses, fat deposition and
hepatic glycogen storage increase (21). The human fetus can synthesize and mobilize glycogen
and respond to the signals that regulate these processes as early as the ninth week of gestation
(22). However, only minute quantities of hepatic glycogen are present in early gestation as the
great bulk of hepatic glycogen accumulates during the third trimester (22,23).
Several types of infants are at risk for neonatal hypoglycemia as a result of limited hepatic
glycogen stores. Infants delivered prematurely have an abbreviated or no third trimester and thus
have limited glycogen stores. Fetuses who are growth-retarded (i.e., small-for-gestational-age
[SGA]) on the basis of limited metabolic fuel availability and diminished gaseous exchange (i.e.,
uteroplacental insufficiency) will use these fuels for growth and not for glycogen synthesis.
Perinatal stress causes neonatal hypoglycemia in part because of catecholamine-stimulated
mobilization of hepatic glycogen stores. This can occur at birth or during the antepartum period. In
the latter situation, fetuses might recover from stress and be delivered without difficulty. As
newborns, such infants have depleted glycogen stores and are at risk for hypoglycemia.
Figure 37-2 Enzymatic activity of the three key enzymes necessary for the oxidation of ketones
(i.e., [A5]-hydroxybutyrate and acetoacetate). The activities of these enzymes are present in
substantial quantities in the human fetal brain during early gestation. (From Felig P, Lynch V.
[Starvation in human pregnancy: hypoglycemia, hypoinsulinemia, and hyperketonemia.] Science
1970; 170:990-992, with permission.)
Gluconeogenesis
For many years, it was believed that maternally derived glucose was the sole metabolic fuel for the
fetus and that the fetus could not produce glucose. As indicated, the fetus can use other fuels such
as the ketones and can, under special circumstances, mobilize hepatic glycogen. The fetus alsocan
carry out gluconeogenesis to a limited degree, although it is likely that under normal
circumstances it does not need to call on this function. Data from human abortuses have
demonstrated that the four key gluconeogenic enzymes are present in fetal liver by 2 to 3 months
of gestation (24,25). The activities of these
P.878
enzymes are believed to increase throughout gestation and the neonatal period. Thus, all
appropriately grown newborns, including the very premature, probably have some degree of
gluconeogenic capability. However, the growth-retarded neonate may have impaired
gluconeogenic capability.
Endocrine Regulation
Insulin and glucagon, important hormones for regulating glucose, can be measured in fetal plasma
as early as 12 weeks of gestation (26). Although plasma concentrations of these hormones are
low, the relative content of these hormones in the fetal pancreas is quite high (27,28). These high
concentrations may result from the limited ability of the fetal islets to secrete these hormones.
Both premature and term infants have limited capacity to secrete these hormones in response to a
glucose challenge in the newborn period. This indicates that the fetus also has limited secretory
capability (Fig. 37-3) (28,29,30). Of note, amino acids have a greater effect in stimulating insulin
and limiting glucagon secretion than glucose in the fetus (31,32).
Insulin may be more important for enhancing growth than for regulating metabolic fuels during
fetal life. Insulin stimulates the growth of specific tissues (e.g., adipose, hepatic, connective,
skeletal, cardiac muscle) (33,34). Excessive insulin secretion during fetal life resulting from such
conditions as maternal diabetes causes the disproportionate growth of insulin-sensitive tissues,
resulting in macrosomia (1,14,35,36). A lack of insulin, as in infants with transient neonatal
diabetes mellitus, always is accompanied by fetal growth retardation.
Figure 37-3 Insulin secretion after glucose challenge in premature infants. Whereas normal
adults secrete insulin briskly in response to glucagon, premature infants in the neonatal period
secrete insulin only sluggishly. (From Assan R, Buillet J. Pancreatic glucagon and glucagon-like
material in tissues and plasma from human fetuses 6 to 26 weeks old. In: Jonxis JHP, Visser HKA,
Troelstra JA, eds. Metabolic processes in the fetus and newborn infant, nutrition symposium.
Baltimore: Williams & Wilkins, 1971: 210-212, with permission.)
Glucagon or a critical glucagon/insulin ratio is important for inducing gluconeogenic enzymes.

Glucagon stimulates the induction of gluconeogenic enzymes in vitro and in vivo (37,38). Fetal
plasma glucagon concentrations increase progressively during fetal life, and this is associated with
a concomitant increase in gluconeogenic enzyme activity. At birth, plasma glucagon concentrations
surge, coinciding with the rapid postnatal increase in gluconeogenic activity (39). Insulin may
modulate glucagon's effect because it can inhibit gluconeogenic enzyme induction (37). Thus, a
balance between these two hormones controls gluconeogenic enzyme induction during perinatal
life.
Adrenergic mechanisms can stimulate hepatic glycogenolysis during fetal life, much as in the
adult. As labor progresses, fetal sympathoadrenal activity increases, resulting in a considerable
increase in circulating catecholamine levels (40,41). Cord clamping triggers an increase in
glucagon secretion (42). As plasma glucose concentrations plummet with cord clamping, insulin
secretion slowly decreases. These adjustments, particularly the remarkable increase in
catecholamine secretion, stimulate glycogenolysis and gluconeogenesis in the neonate (Fig. 37-4).
Islet cell function remains unresponsive for several weeks of neonatal life in term infants. Newborn
infants increase insulin and limit glucagon secretion sluggishly in response to glucose challenge.
Limited data indicate that these responses become adult-like between the first and second weeks
of life (43). This adaptation is critically important because the neonate, unlike the fetus, must
P.879
regulate glucose production and storage through feeding and fasting cycles. Little is known about
the premature infant's ability to regulate glucose in the neonatal period. The ability to modulate
insulin and glucagon secretion probably develops as the neonatal period progresses.
Figure 37-4 Levels of hormones and metabolic fuels change after birth. At birth, the
counterregulatory hormones (i.e., catecholamines and glucagon) increase greatly, whereas insulin
secretion decreases. Neonatal plasma glucose concentrations plummet as a result of cord
clamping. The changes in counter regulatory hormones and insulin favor mobilization of glucose
and fat and stimulate gluconeogenesis. These changes assure adequate neonatal glucose
production. FAA, free fatty acids. (From Kalhan SC, Bier DM, Savin SM, et al. Estimation of
glucose turnover and 13C recycling in the human newborn by simultaneous [1-13C]glucose and
[6,6-1H2]glucose tracers. J Clin Endocrinol Metab 1980;50:456-460, with permission.)
Glucose Transporters
Glucose transporters (Glut) are a family of structurally similar proteins encoded by a family of
genes and expressed in a tissue-specific manner in most mammalian tissues (44). Several
isoforms have been described in the human fetus and placenta. Because Gluts facilitate transfer of
glucose from the maternal to fetal circulation and also uptake of glucose by most fetal and
neonatal tissues, they are critically important for growth and development.
Glut1 is the dominant isoform in most fetal tissues and the placenta (45,46). Insulin, insulin-like
growth factors, and other hormones and peptides regulate its activity and expression. The Gluts
are developmentally regulated. The appearance of Glut isoforms varies in time of appearance and
type (Glut2 in liver, Glut4 in muscle) during intrauterine and neonatal life (47). Glut1 and Glut3 are
expressed in various tissues, including the placenta. Ambient glucose can affect both glucose
transport activity and expression of Glut genes. Glut1 expression appears to increase with
gestation. In placentas of women with diabetes, Glut3 decreases with gestation, suggesting an
attempt by the placenta to limit glucose transport in the face of maternal hyperglycemia (48).
NEONATAL GLUCOSE REQUIREMENTS

The clinician makes a leap of faith in using the chemical or paper strip determination of plasma or
blood glucose concentration to judge the adequacy of tissue glucose provision in the neonate. A
normal plasma concentration is interpreted to mean that glucose supply to the brain and other
organs is adequate for ongoing metabolic needs. To appreciate glucose requirements, glucose
kinetics must be understood.

Glucose turnover represents the rate of production of glucose by the liver and other organs and
the simultaneous use or uptake of glucose by the brain and other organs. Turnover is usually
expressed as milligrams of glucose per kilogram of body weight per minute. Although stable
isotope technology has allowed quantification of glucose turnover in neonates, this methodology
cannot be directly applied for clinical purposes. Thus, the clinician must rely on the measurements
of glucose concentrations, which are static, as representative of the dynamic rate of glucose
production and use in a neonate. In general, plasma glucose concentrations roughly correlate with
glucose turnover. Diminished plasma glucose concentrations suggest that glucose production is
limited or that glucose use is increased (i.e., need is outstripping production). Elevated plasma
glucose concentrations suggest that either production is excessive or, more likely, organ uptake
and use are diminished. These are the dynamic physiologic conditions that define hypoglycemia
and hyperglycemia.
Figure 37-5 Glucose turnover in premature and term neonates and the adult. When related to
body weight, glucose turnover is greatest in the premature infant and least in the adult. The
increased turnover in neonates results in part from their relatively increased brain-to-body mass
ratio. (From Kalhan SC, Bier DM, Savin SM, et al. Estimation of glucose turnover and 13C
recycling in the human newborn by simultaneous [1-13C]glucose and [6,6-1H2]glucose tracers. J
Clin Endocrinol Metab 1980;50:456-460, with permission.)
In the neonate, glucose production correlates directly with brain and body mass, confirming the
critical role of glucose as a metabolic fuel (49,50,51,52). This holds even in the most immature of
premature infants (53). Glucose turnover for newborn infants, when related to body mass,
significantly exceeds that of adults (Fig. 37-5). Premature infants have even greater turnover
values than term neonates. This in part reflects the ratio of brain to body mass, which is greatest
in premature infants and least in adults. These relations emphasize the importance of glucose as
the primary fuel for the brain (Fig. 37-6).
Neonatal Hypoglycemia
A variety of blood and plasma glucose concentration values based on screening of neonates or
clinical experience have been recommended as values defining hypoglycemia (54,55). All of these
are somewhat arbitrary because they cannot be correlated directly with glucose use rate or
severity of symptoms. Because plasma or blood glucose concentrations only roughly reflect
glucose turnover, a plasma glucose concentration less than 40 mg/dL should be used to define
hypoglycemia. When glucose turnover is sufficient to meet the needs of the organism,
concentrations usually
P.880
exceed this value. It also is important to note that values somewhat less than 40 mg/dL still can
be associated with adequate glucose provision.
Figure 37-6 Linear and curvilinear regression analyses indicate strong relations between glucose
turnover and body mass (A), and between glucose turnover and brain mass (B), in newborn
human infants. (From Devaskar SU, Mueckler MM. The mammalian glucose transporters. Pediatr
Res 1992;31:1-13, with permission.
The chemical definition of hypoglycemia must take into account the methodology of glucose
determination. Glucose concentration in whole blood is approximately 10% to 15% lower than that
in plasma. Delay in determination after blood sampling may result in glucose oxidation by
erythrocytes, causing falsely low values. Although the use of paper strip methods to estimate
glucose concentrations quickly is acceptable, their results should be corroborated by true chemical
determinations.
The clinical manifestations of inadequate glucose provision to the neonatal brain range from no
symptoms to lethargy or mild tremors to frank convulsions (Table 37-1). The degree of glucose
limitation necessary to cause brain damage is unknown. The lack of clearly defined data on this
problem and the prevailing opinion concerning the potentially damaging effects of hypoglycemia
mandate that infants at risk be monitored and that asymptomatic and symptomatic infants be
appropriately treated.
TABLE 37-1 SYMPTOMS OF HYPOGLYCEMIA
Jitteriness
Tremors
Apnea
Cyanosis
Limpness/lethargy
Seizures
All conditions associated with the development of hypoglycemia in the neonate result from one or
a combination of two basic mechanisms: inadequate production or excessive tissue use.
Inadequate glucose production results from a lack of glycogen stores, an inability to synthesize
glucose, or both (Fig. 37-7). Excessive tissue use results from increased insulin secretion. Table 372 categorizes infants at risk for hypoglycemia in relation to these basic mechanisms.
Inadequate Glucose Production as a Result of Limited

Glycogen Stores
Premature Infants
As indicated, the third trimester of pregnancy is an important period for hepatic glycogen
deposition. An infant delivered prematurely without the benefit of part of or the entire third
trimester will have limited hepatic glycogen stores. The greater the degree of prematurity, the less
glycogen will be present. SGA premature infants are at extremely
P.881
high risk for development of hypoglycemia because available nutrients during intrauterine life are
channeled toward growth, with little set aside for glycogen storage. For this reason, SGA
premature infants have extremely limited glycogen stores (55,56,57).
Figure 37-7 The rates of glucose production and utilization are represented by the faucet and
drain of the sink. The level in the sink is equivalent to plasma or blood glucose concentrations. If
production from glycogenolysis and gluconeogenesis is adequate, and use is not excessive,
normoglycemia exists and the plasma or blood glucose concentration (i.e., the level in the sink) is
normal. Hypoglycemia develops if production is inadequate to meet body needs or if use outstrips
production. This results in decreased glucose concentrations (i.e., diminished level in sink). (From
Kalhan SC, Bier DM, Savin SM, et al. Estimation of glucose turnover and 13C recycling in the
human newborn by simultaneous [1-13C]glucose and [6,6-1H2]glucose tracers. J Clin Endocrinol
Metab 1980; 50:456-460, with permission.)
TABLE 37-2 INFANTS AT RISK FOR HYPOGLYCEMIA
Diminished production
Limited glycogen
SGA
Prematurity
Birth stress
Glycogen storage disorders
Limited gluconeogenesis
SGA
Inborn errors
Increased utilization
Hyperinsulinism
IDM
Beckwith-Wiedemann syndrome
Nesidioblastosis or pancreatic adenoma
Erythroblastosis fetalis
Exchange transfusion, chlorpropamide, benzothiazides,
-sympathomimetics, malpositioned UA catheter
Unknown
LGA infants who are not IDM
Sepsis
Polycythemia or hyperviscosity syndrome
Congenital hypopituitarism
a
IDM, infant of diabetic mother; LGA, large for gestational age; SGA,
small for gestational age; UA, umbilical artery.
Infants Who Have Suffered Perinatal Disease

Infants who are stressed in utero are at increased risk for development of hypoglycemia as
neonates. Hypoxia, acidosis, and alterations in fetal blood pressure and flow can stimulate
catecholamine secretion in utero, which in turn will mobilize hepatic glycogen stores. In addition,
hypoxia increases the rate of anaerobic glycolysis, thereby accelerating glucose use. These events
deplete fetal glycogen stores and place the infant at risk for hypoglycemia after delivery.
Glycogen Storage Disease

Intrinsic defects in glycogen synthesis, storage, or breakdown result in complex metabolic
problems (see Chapter 41). Several types of glycogen storage disease (e.g., Ia, I, VI, 0) include
hypoglycemia as one of many associated complications (54,58).
Inadequate Glucose Production as a Result of Limited

Gluconeogenesis
Small-for-Gestational-Age Infants
Full-term and premature SGA neonates are at great risk for development of hypoglycemia as a
result of inadequate hepatic glycogen stores. With treatment, the hypoglycemia is usually short
lived. Approximately 1% of SGA infants in whom hypoglycemia develops have a prolonged course
requiring intravenous therapy for days. A delay in the induction of gluconeogenic capability
probably is responsible for this prolonged hypoglycemia. These SGA neonates have elevated
plasma concentrations of gluconeogenic precursors, suggesting an inability to convert exogenous
gluconeogenic precursors, such as alanine, to glucose (59,60).
In animal models of intrauterine growth retardation, the induction of one gluconeogenic enzyme,
phosphoenolpyruvate carboxykinase, is delayed (61,62). This occurs despite appropriate increases
in glucagon and decreases in insulin; these relations should favor enzyme induction. Why some
SGA infants fail to induce gluconeogenic capability is unclear. Corticosteroid therapy can be used
to treat prolonged hypoglycemia. The success of this practice probably results in part from the
ability of corticosteroids to induce hepatic gluconeogenic enzymes.
Many SGA infants have heightened metabolic requirements during the neonatal period. The
mechanisms for this are not understood, but they may represent an attempt to compensate for the
preceding intrauterine deprivation (63). This may explain the increased glucose requirements
demonstrated by some hypoglycemic SGA infants.
Congenital Absence of Gluconeogenic Capability

Unlike SGA infants in whom gluconeogenic capability eventually develops, a few infants have been
reported to have permanent congenital lack of gluconeogenic enzymes (64). A single infant who
was unable to secrete glucagon also has been reported (65).
Excessive Tissue Use or Hyperinsulinism

A variety of disorders are associated with fetal and neonatal hyperinsulinism. In some disorders,
the mechanisms for heightened beta-cell function are well understood, whereas in others, the
pathogenesis is unclear. In the former category are infants of diabetic mothers (IDM) and infants
with altered pancreatic islets caused by conditions such as nesidioblastosis and pancreatic
adenoma. Those for whom an etiology is not clear include infants with erythroblastosis and
Beckwith-Wiedemann syndrome. The finding of a hypoglycemic infant who is macrosomic and
requires high rates of glucose infusion (10 to 20 mg/kg body weight per minute) suggests a
hyperinsulinemic state.

IDM are at great risk for development of hypoglycemia as a result of the carryover of the fetal
hyperinsulinemic state into neonatal life. They have elevated plasma insulin concentrations and
release insulin briskly in response to glucose challenge. The problems of the IDM are presented in
the following sections.
P.882
NesidioblastosisPersistent Hyperinsulinemic
Hypoglycemia of Infancy
This condition is characterized by macrosomia and profound prolonged neonatal hypoglycemia due
to hyperinsulinism. Several different accidents of development at the cell morphologic and
molecular level cause persistent fetal and neonatal hyperinsulinism. Classically, in nesidioblastosis,
pancreatic ductular cells are found in acinar tissue. In persistent hyperinsulinemic hypoglycemia,
there can be focal hyperplasia of pancreatic islet cells or diffuse lesions of the entire pancreas
(66,67,68). Focal lesions correspond to somatic defects and are, in some cases, related to
mutations of sulfonylurea receptor 1 (69,70,71). Other forms of persistent hyperinsulinemic
hypoglycemia of infancy include mutations of the potassium-linked ATP channel in beta cells (72).
Nesidioblastosis or persistent hyperinsulinemic hypoglycemia of infancy should be considered
whenever a macrosomic infant has hypoglycemia with elevated plasma insulin concentrations over
several days. Rebound hypoglycemia in response to excessive glucose administration is another
characteristic. Increased insulin/glucose ratios and glucose requirements exceeding 10 mg/kg/min
support the possibility of nesidioblastosis.
Surgical excision of a portion of the pancreas can provide definitive diagnosis and therapy.
However, over the long term this may result in the development of diabetes mellitus in the patient.
Both somatostatin and diazoxide have been used successfully to limit insulin secretion for as long
as several months and may produce remission (73,74)
Unexplained Neonatal Hyperinsulinemia and Hypoglycemia

Infants with Beckwith-Wiedemann syndrome, erythroblastosis fetalis, and those whose mothers
took chlorpropamide or benzothiazides are at risk for development of hypoglycemia as a result of
hyperinsulinism. In 1964 Beckwith and colleagues (75) and Lazarus and associates (76)
independently reported exophthalmos-macroglossia-gigantism syndrome. Such infants frequently
have omphalocele, muscular macroglossia, macrosomia, and neonatal hypoglycemia.The
hypoglycemia and macrosomia are caused by hyperinsulinism resulting from beta-cell
hypertrophy. Some of the morbidity and mortality originally associated with this syndrome was
related to unrecognized hypoglycemia. Thus, early recognition and prevention of hypoglycemia is
mandatory. The gene responsible for many of the features of this syndrome is found on
chromosome 11p15.5 (77). Because the insulin and insulin-like growth factor genes are closely
located, combined gene dosage imbalance probably also contributes to the development of
hyperinsulinism and hypoglycemia (78).
Infants with erythroblastosis fetalis caused by Rh incompatibility were reported in the past to be at
risk for hypoglycemia from hyperinsulinism secondary to beta-cell hyperplasia (79). The
mechanisms responsible for islet cell hyperplasia are unknown, although it was proposed that
elevated plasma glutathione concentrations may stimulate the fetal beta cell to increase insulin
secretion (80). The advent of direct intravascular transfusion of Rh-affected fetuses may reduce
the risk of hypoglycemia. With this therapy, severely affected Rh fetuses that receive serial
intravascular transfusions by the percutaneous umbilic technique are normoinsulinemic despite
originally having elevated glutathione concentrations. Hypoglycemia does not develop in these
fetuses after birth (81).
It is important to note that infants undergoing exchange transfusion are at risk for development of
hypoglycemia because of stimulation of insulin secretion by glucose instored erythrocytes (74).
Checking for hypoglycemia during and after an exchange transfusion is therefore important.
Maternal use of chlorpropamide and benzothiazide can directly increase insulin secretion in the
neonate (82,83). -sympathomimetic agents used to stop premature labor have been reported to
cause neonatal hypoglycemia (84). These drugs stimulate glycogen breakdown and
gluconeogenesis in the mother and fetus (85). Both the increased availability of maternal glucose
and the -sympathomimetic agent that crosses the placenta stimulate fetal insulin secretion,
resulting in neonatal hyperinsulinism and hypoglycemia. For these reasons, infants whose mothers
received tocolytic therapy shortly before delivery should be monitored for hypoglycemia.
Some debate exists as to whether glucose administered to the mother during labor and delivery
stimulates fetal beta-cell secretion and causes neonatal hyperinsulinism and hypoglycemia. Acute
maternal glucose loading may stimulate fetal insulin secretion and increase the risk of neonatal
hypoglycemia (86). If glucose infusion is well controlled, the likelihood of this is minimized. Control
of maternal glucose administration is particularly important in situations in which the fetus is
suspected of having heightened beta-cell sensitivity (e.g., maternal diabetes, Rh incompatibility)
because under these circumstances even moderate excursions of glucose may stimulate fetal
insulin secretion.
Malposition of the tip of an umbilic artery catheter at a level between the tenth thoracic and
second lumbar vertebrae may result in glucose-stimulated hyperinsulinism. Several infants have
been reported in whom hypoglycemia was relieved only when the tip of the umbilic artery catheter
was repositioned. It has been proposed that glucose from the malpositioned catheter flows into the
celiac axis, thereby stimulating insulin secretion (87). Animal studies have confirmed this
possibility (88), which should be considered in unexplained cases of hypoglycemia.
Large-for-gestational-age (LGA) infants whose mothers do not have diabetes mellitus are at risk
for transient hypoglycemia. This is particularly true of LGA infants of obese women (89). The
mechanisms responsible for hypoglycemia are unknown, although limited data suggest that
hyperinsulinism is not a major factor.
P.883
Sepsis in a neonate is often heralded by hypoglycemia or hyperglycemia. The mechanisms for this
are not understood. Several studies have indicated rapid glucose disposal rates after intravenous
challenge in septic term neonates. Although this suggests a hyperinsulinemic state, insulin
secretion in these neonates was normal (90,91). The hyperglycemia and hypoglycemia that often
precede the other signs of sepsis in premature infants may be catecholamine mediated.
Hypoglycemia is a well-acknowledged complication of neonatal polycythemia-hyperviscosity
syndrome (92). Although polycythemia is more likely to occur in SGA and LGA infants who are at
risk for hypoglycemia for other reasons, hypoglycemia also occurs at an increased rate in
polycythemic appropriately grown infants. Animal studies have documented diminished cerebral
glucose uptake with polycythemia; however, the mechanisms responsible for decreasing glucose
provision are unknown (93). The increased erythrocyte mass is not sufficient to reduce glucose
availability. The diminished plasma volume resulting from polycythemia may limit glucose
provision. These possibilities remain to be confirmed.
Congenital hypopituitarism is a rare disorder in the neonate resulting from a spectrum of
developmental accidents (94,95). Congenital absence of the anterior pituitary is the common
cause of this disorder, although holoprosencephaly and optic disk dysplasia also have been
associated. Affected male neonates have microphallus, whereas girls have normal external
genitalia (96). Neonatal hypoglycemia often develops and can be severe. The endocrine alterations
resulting from congenital hypopituitarism are complex, and the mechanisms by which they cause
hypoglycemia are not understood. Congenital syphilis has been reported to cause hypopituitarism
and this syndrome (97). Growth hormone is important in this regard because it can reverse
hypoglycemia. Because hypoglycemia can develop later in the postnatal period, infants should
have growth hormone therapy initiated for the long term.
Infants who have suffered hypothermia are at increased risk for development of hypoglycemia
(98). This may result from increased availability of catecholamines (99), which would deplete
glycogen reserves. Tissue use of glucose might also be increased under these conditions.
Other unusual clinical conditions reported in association with hypoglycemia include salicylate
administration (100), congenital adrenal hyperplasia (101), and trisomy 13 mosaicism (102). The
mechanisms for these phenomena are not known.
Hypoglycemia has been noted to sometimes occur in infants who suffer poor calorie intake for
prolonged periods as a consequence of inadequate maternal breast milk production. The
mechanisms for this are not known but probably involve glycogen depletion and diminished release
of gluconeogenic precursors by striated muscle. Severe hepatic damage, most likely a result of
impaired substrate transport and gluconeogenic capability, can also cause hypoglycemia.
Disruption of intravenous glucose administration or rapid reduction in the rate of a glucose infusion
can be associated with rebound hypoglycemia as a result of sluggish beta-cell responsiveness;
insulin secretion may not decrease with appropriate rapidity and in response to cessation of
exogenous glucose.
Neonatal Hyperglycemia
Hyperglycemia occurs primarily in three major groups of infants: those who are very premature,
those who have neonatal diabetes mellitus, and those who are septic. Altered glucoregulation in
response to sepsis was discussed in the preceding section (see Unexplained Neonatal
Hyperinsulinemia and Hypoglycemia) (103,104).
The Very Premature Infant

Advances in perinatal care have improved the survival rate of the very premature infant. With this,
the problem of glucose intolerance has greatly increased because the risk of hyperglycemia is at
least 18 times greater in infants weighing under 1,000 grams than in those weighing more than
2,000 grams. Depending on the definition, age at screening, and type of intravenous solution
administered, the incidence of hyperglycemia in very premature infants has been reported to
range from 20% to 86% (103,104). In general, the smaller and more premature an infant, the
greater the likelihood that he or she will not tolerate exogenous glucose at maintenance rates (4 to
6 mg/kg/minute). It is not uncommon to administer glucose at 1 to 2 mg/kg/minute or less and
still observe plasma glucose concentrations exceeding 200 mg/dL.
Several basic mechanisms are probably responsible for glucose intolerance in very premature
infants. Many probably do not secrete glucoregulatory hormones appropriately. In addition, endorgan response to these hormones may be blunted. Thus, endogenous glucose production may
continue while tissue glucose uptake is limited despite intravenous glucose therapy. Limited data
indicate that the premature infant will only slowly increase insulin secretion in response to glucose
challenge (105,106,107,108,109). The amount secreted may not be sufficient to regulate glucose.
Such infants may (108) or may not (110) decrease glucagon in response to glucose. In addition,
very premature infants can be resistant to insulin (111). This resistance is accentuated by
catecholamines, which are often quite elevated. These factors contribute to the limited ability of
the premature infant to reduce glucose production in the same manner as the adult when
exogenous glucose is provided (Fig. 37-8).
Neonatal Diabetes Mellitus

This is a rare disorder characterized by hypoinsulinism, progressive wasting, polyuria, and
glycosuria during the neonatal period. Such infants usually are not ketotic. Of note, they are
always SGA (112,113). The intrauterine growth
P.884
retardation results from limited fetal insulin secretion and exemplifies the importance of insulin as
a fetal growth-stimulating hormone. This disorder may be heterogeneous with respect to etiology.
Synthesis of an abnormal, poorly functioning insulin molecule or receptor deficiencies are other
possible causes. Neonatal diabetes mellitus usually is transient; the early acute period calls for
standard diabetic therapy (i.e., monitoring caloric intake and administering exogenous insulin).
Figure 37-8 Hepatic glucose production rates in human premature neonates, full-term neonates,
and adults during either saline (open symbols) or glucose (closed symbols) infusion. With glucose
infusion, adults and full-term infants but not preterm infants reduce endogenous glucose
production. Many preterm infants are apparently unable to regulate glucose production; this
contributes to the development of hyperglycemia in these infants. (From Bower BD, Jones LF,
Weeks MM. Cold injury in the newborn. A study of 70 cases. Br Med J 1960;5169:303-309, with
permission.)
DIAGNOSIS AND TREATMENT

Hypoglycemia
All infants at risk for development of hypoglycemia should undergo frequent plasma glucose
determinations. The commercially available paper strip indicators are widely used. Because their
accuracy is fair, it is desirable to confirm values close to hypoglycemia or hyperglycemia with
laboratory chemical determinations. Infants at risk for hypoglycemia should be checked frequently
during the first 4 hours of life and then at 4-hour intervals until the risk period has passed. If an
infant is feeding, blood sampling should be done before feeding. For IDM and SGA infants, the
screening should continue for at least 24hours.
Infants who have borderline asymptomatic hypoglycemia, who do not have respiratory distress
syndrome (RDS) or other serious disorders, and who are capable of enteral feedings may receive
either 5% dextrose solution or formula as their initial treatment. In general, this approach can be
used for infants at term who are LGA or SGA. However, because this mode of therapy is not
always successful, plasma or blood glucose concentrations must be checked shortly after feeding.
Intravenous administration of glucose in a quantity sufficient to meet tissue requirements is the
treatment of choice for hypoglycemia. The administration of 10% or 15% dextrose solution at 5 to
10 mL/kg body weight, followed by a continuous infusion at 5 to 6 mg/kg body weight/minute, will
increase plasma glucose concentrations to 40 mg/dL or greater and acutely meet tissue
requirements. The maintenance rates may require adjustment depending on the etiology of
hypoglycemia.
Glucagon and epinephrine increase glucose production. Because both mobilize hepatic glycogen
stores, their efficacy in treating hypoglycemia is variable, particularly in infants with limited hepatic
stores. The numerous cardiovascular effects of epinephrine also limit its usefulness in infants.
Infants who are hypoglycemic for prolonged periods as a result of an inability to produce glucose
can be treated with corticosteroids (hydrocortisone 5 mg/kg/day every 12 hours; prednisone 2 mg/
kg/day orally). Steroids exert some of their effects by inducing gluconeogenic enzyme activity.
Hyperglycemia
Hyperglycemia in low-birth-weight infants traditionally has been treated by reducing the rate of
administration of exogenous glucose. This is something of a clinical paradox because such a
reduction can theoretically limit glucose availability to the brain. Attempts in the past to provide
exogenous insulin as a means to regulate glucose met with variable success, primarily because of
technical difficulties in providing insulin (111,114,115). New insulin delivery systems designed for
children and adults provide the means to deliver minute amounts of insulin under controlled
conditions. These systems have finely tuned programmable pumps and tubing that does not bind
insulin (116). Application of this technology has met with preliminary success (117). Individualized
continuous insulin infusion to infants at 26 weeks of gestation weighing 700 to 800 grams
enhanced glucose infusion and parenteral energy intake. Weight gain was significantly greater over
7 to 21 days compared with infants managed conventionally (Fig. 37-9). If further studies confirm
these observations and clarify the potential metabolic consequences of this therapy, this method
may prove beneficial in treating hyperglycemia in low-birth-weight infants.
Figure 37-9 Mean glucose infusion rates per day in insulin-treated (solid line) and control
(dashed line) very-low-birth-weight infants. The controlled administration of insulin significantly
increased glucose infusion from day 2 to 16. (From Johansson EDB. Plasma levels of progesterone
in pregnancy measured by a rapid competitive binding technique. Acta Endocrinol (Copenh)
1969;61: 607-617, with permission.)
P.885
CONSEQUENCES
Hypoglycemia
The clinical manifestations of inadequate glucose provision to the brain range from no symptoms
to mild tremors to seizures. The issue of potential long-term sequelae of hypoglycemia remains
unclear. This issue is complicated by the fact that hypoglycemia often occurs in infants who have
coexisting conditions that also can cause brain damage. The observations suggesting that
asymptomatic hypoglycemia poses less of a risk compared to symptomatic hypoglycemia are
limited, as are the observations that associate prolonged hypoglycemia with a greater risk of brain
damage than brief hypoglycemia. Limited data suggest that seizures associated with hypoglycemia
worsen prognosis (118,119,120,121,122,123,124,125,126,127). Because of the uncertainty in this
area, all hypoglycemic infants, symptomatic or not, should be appropriately treated.
Hyperglycemia
Hyperglycemia has been associated with an increased incidence of mortality, intracranial
hemorrhage, and developmental delay in very premature infants. Whether hyperglycemia causes
these problems or is merely associated with their development is unclear. Increased serum
osmolarity resulting from hyperglycemia might disrupt cell-serum balance and cause cell injury
(103). In addition, hyperglycemia can alter cell glucose transport, which may perturb cell
metabolic functions.

Table 37-3 lists the frequently occurring problems of IDM categorized according to proposed
mechanisms. The fetal and neonatal hyperinsulinism central to many of these problems results
from exaggerated provision of maternal metabolic fuels to the fetus, resulting in pancreatic betacell hypertrophy, hyperplasia, and hyperfunction (1,14). The alterations in maternal metabolism
resulting from diabetes mellitus that are responsible for these and other effects on the fetus have
been reviewed.
TABLE 37-3 PROBLEMS OF THE INFANT OF A DIABETIC MOTHER
Macrosomia or birth stress

Hypoglycemia
Intrauterine growth retardation
Hypocalcemia
Hyperbilirubinemia
Polycythemia or hyperviscosity
Cardiomyopathy
Congenital anomalies
Hyperinsulinism
Altered metabolic fuels, uteroplacental insufficiency
Decreased parathormone
Increased erythrocyte destruction, bruising
Altered Fetal Growth

Macrosomia, a clinical term suggesting excessive weight for gestational age, is a well-known
characteristic of IDM (Fig. 37-10). Macrosomia in IDM results primarily from increased adiposity
because IDM have both adipocyte hyperplasia and adipocyte hypertrophy (128). IDM also have
excess nonfatty tissue. The liver and heart often are enlarged, and skeletal muscle increased.
Much of this excess tissue is located in the shoulders and intrascapular area. Because IDM have
normal brain growth, this results in a disproportionality between head and shoulder size and
greatly increases the risk of shoulder dystocia (Fig. 37-11). This
P.886
birth complication occurs far more frequently in macrosomic IDM than in large infants of mothers
who do not have diabetes. Macrosomia is responsible for the great risk of birth trauma, meconium
aspiration syndrome, persistent pulmonary hypertension, and high incidence of cesarean section
delivery of IDM.
Figure 37-10 A macrosomic infant of a diabetic mother (IDM) has head circumference and length
that are at the 90th percentile; the IDM's body weight greatly exceeds the 90th percentile. The
IDM has considerable fat deposition in the shoulder and intrascapular area.
Figure 37-11 Lateral skull and neck radiograph of an infant of a diabetic mother after a difficult
vaginal delivery. Infants of diabetic mothers are at extreme risk for shoulder dystocia, which can
result in severe complications, such as separation of the C1-C2 cervical spine.
Because insulin has both mitogenic and anabolic effects in the fetus, the fetal hyperinsulinemic
state is central to the development of macrosomia. The augmented production of insulin by the
fetus stimulates the growth of insulin-sensitive tissues (e.g., adipose, muscle, connective) to cause
macrosomia. Hepatic glycogen storage is exaggerated. The effect of insulin probably is mediated to
some extent through stimulation of insulin-like growth factors (129). It is not surprising that head
growth is normal in IDM during intrauterine life because insulin does not stimulate brain growth to
any great extent.
The excess fat in IDM develops during the third trimester; IDM delivered before 30 weeks of
gestation rarely are LGA. Serial fetal ultrasound measurements confirm that the fetal IDM does not
exceed normal growth limits until 28 to 30 weeks of gestation (30). Despite improved maternal
therapy, 20% to 30% of insulin-dependent diabetic women continue to bear macrosomic infants
(130). Women with gestational diabetes, the mildest form of carbohydrate intolerance, have as
great an incidence of macrosomia as women with preexisting diabetes.
Intrauterine growth retardation is another well-known complication of diabetes in pregnancy. The
development of growth retardation has been attributed to maternal vascular disease, causing
uteroplacental insufficiency. More recent data suggest that growth retardation may result from
alterations in maternal metabolic fuel availability during early gestation (131).
Figure 37-12 Beta-cell response to intravenous glucose challenge in infants of diabetic mothers
(IDM) and infants of mothers with normal carbohydrate metabolism (INM). C-peptide is cleared
from the proinsulin molecule when the beta cell is stimulated to secrete insulin. The measurement
of C-peptide represents insulin on an equimolar basis and is a more accurate measure of beta-cell
secretion than insulin in IDM. IDM exuberantly secrete insulin in response to glucose challenge.
This adult-like response differs greatly from the normally expected sluggish insulin response of
the INM. The significantly increased insulin concentration in IDM before glucose challenge (i.e., 0
minutes) indicates that basal insulin secretion also is elevated. The increased beta-cell function in
IDM is responsible for their high incidence of hypoglycemia.
Hypoglycemia
Approximately 25% to 50% of all IDM who manifest hypoglycemia will do so within the first 24
hours of life. Hypoglycemia is particularly likely to occur in macrosomic IDM because
hyperinsulinism is responsible for both fetal overgrowth and hypoglycemia (Fig. 37-12) (132).
Several studies also suggest that these IDM may fail to release glucagon or catecholamines in
response to hypoglycemia (133). These hormonal alterations result in both increased glucose
clearance and diminished glucose production.
Hypocalcemia and Hypomagnesemia

Hypocalcemia develops in 10% to 20% of IDM during the neonatal period. This usually occurs in
association with hyperphosphatemia and occasionally with hypomagnesemia. Parathormone
concentrations are significantly lower in IDM than in infants of nondiabetic mothers during the first
4 days of life (Fig. 37-13). This may be a result of hypomagnesemia, which limits parathormone
secretion even in the presence of hypocalcemia. Maternal hypomagnesemia, which may occur from
increased renal loss secondary to diabetes, is believed responsible for fetal and neonatal
hypomagnesemia (134,135). Birth asphyxia, which frequently occurs in IDM, also causes
hypocalcemia.
Figure 37-13 Parathormone (PTH) and calcium concentration in infants of diabetic mothers
(circles) and full-term control (triangles) infants. Infants of diabetic mothers have lower plasma
calcium and PTH concentrations during the first 6 days of life. (From Lucas A, Morley R, Cole TJ.
Adverse neurodevelopmental outcome of moderate neonatal hypoglycaemia. BMJ 1988;297:13041308, with permission.)
P.887
Hyperbilirubinemia
Indirect hyperbilirubinemia develops in 20% to 25% of IDM. Their carbon monoxide production is
increased as a result of increased hemoglobin breakdown and bilirubin production (136). The
increased rate of erythrocyte breakdown in IDM is probably linked to altered erythrocyte
membrane composition resulting from changes in maternal fuel availability. Polycythemia
frequently occurs in IDM, and the normal breakdown of this increased erythrocyte mass also
causes hyperbilirubinemia. Macrosomic IDM are often bruised at birth; the resultant resorption of
blood also contributes to hyperbilirubinemia.
Hyperviscosity
IDM have a 10% to 20% risk of being polycythemic and developing neonatal hyperviscosity
syndrome. Several factors are responsible for this. The hematocrit of umbilic-cord blood at birth
tends to be elevated, probably as a result of increased erythropoiesis (137).

The increased incidence of renal vein thrombosis reported in IDM may be related to hyperviscosity,
although this disorder does occur in IDM with normal hematocrits.
Unexpected Fetal Death

In the past, a high incidence of unexpected fetal death occurred in late gestation. Improvement in
metabolic control of maternal diabetes throughout pregnancy and new methods to assess fetal
status have decreased the incidence of this tragic complication. The mechanisms of unexpected
fetal death are not completely understood. In the fetal sheep, sustained hyperglycemia is
associated with increased insulin secretion, elevated fetal oxygen consumption, acidosis, and death
(138). This could explain the association between poor maternal metabolic control and the
increased risk of fetal death.
Respiratory Distress Syndrome

IDM are at increased risk of developing RDS. In the past, IDM were at a four- to six-fold greater
risk for RDS than infants of nondiabetic mothers (139). This incidence has been reduced
substantially by the recent emphasis on tight control of maternal metabolism. The increased risk of
RDS in poorly regulated diabetic women is due in great part to fetal hyperinsulinism. Insulin
adversely affects fetal lung maturation by inhibiting the development of enzymes necessary for the
synthesis of the phospholipid components of surfactant (140). Standard methods to assess fetal
lung maturity antenatally may not be applicable to a diabetic pregnancy. The measurement of
phosphotidylglycerol has greatly improved this capability.
Cardiomyopathy
IDM are at increased risk for various cardiomyopathies (141). Many have thickening of the
interventricular septum and left or right ventricular wall. The increased cardiac muscle mass
results from the fetal hyperinsulinemic state. Most of these infants are asymptomatic, and the
thickening is detected by electrocardiogram or echocardiogram. In a small fraction of infants,
outflow obstruction severe enough to cause left ventricular failure may occur. These abnormalities
generally regress over 3 to 6 months, and the condition appears to have no permanent effect on
the myocardium. Those infants with congestive heart failure who survive the initial period with
medical management also improve spontaneously.
Occasionally, IDM have severe congestive heart failure at birth. Frequently, these infants have
suffered intrapartum asphyxia and are hypoglycemic and hypocalcemic. Such infants generally
respond to assisted ventilation and correction of their metabolic abnormalities and usually recover
completely. It is unclear whether heart failure results from the combined effects of hypoglycemia,
hypocalcemia, and asphyxia on an inherently normal myocardium, or whether the myocardium is
abnormal and therefore more susceptible to failure.
Figure 37-14 Magnetic resonance image of an infant of a diabetic mother. The infant has spinal
agenesis-caudal regression syndrome. The spinal cord is interrupted, and hip-femur relationships
are malformed.
P.888
Congenital Abnormalities
Major congenital malformations occur two to four times more frequently in IDM than in infants
born to nondiabetic women. Although many abnormalities occur in IDM, ventricular septal defects,
transposition of the great arteries, and spinal agenesis-caudal regression syndrome occur with
particular frequency (Fig. 37-14). Neural tube defects, gastrointestinal atresia, and urinary tract
malformations also are relatively common. A transient anomaly unique to IDM is known as the
neonatal small left colon, microcolon, or lazy colon syndrome. This condition presents as
gastrointestinal obstruction, and barium contrast studies suggest congenital aganglionic
megacolon. Unlike infants with Hirschsprung disease, these infants have normal innervation of the
bowel and ultimately have normal bowel function.
Poor control of maternal diabetes during the first trimester, a critical period of organogenesis, has
been proposed as the mechanism for the increased incidence of malformations (142,143). IDM
with congenital abnormalities generally have normal karyotypes (144). In vitro studies using
embryos of laboratory animals have demonstrated that altering metabolic fuels can produce
profound malformations (145). Clinical studies, however, have not confirmed a relation between
birth defects and alterations in maternal metabolic variables (146).
Postnatal Problems
The postnatal problems of IDM are consistent in the Barker hypothesis which links intrauterine
disorders with adverse adult outcomes (147). IDM are at increased risk for development of obesity
in later life, compared to infants of mothers with normal carbohydrate metabolism (148). Studies
suggest that in utero hyperinsulinism may be responsible for this postnatal phenomenon. IDM who
become obese during childhood have the severest hyperinsulinism in utero (149).
Whether maternal diabetes adversely affects the long-term cognitive development of the offspring
remains unanswered. In the past, the increased incidence of birth trauma and neonatal disorders
probably contributed to an increased risk of poor outcome. Studies suggest that altered maternal
metabolic fuel availability also may have an effect. An inverse correlation has been reported
between childhood IQ and degree of abnormality of second- and third-trimester maternal lipid
metabolism (150). This is consistent with the potential detrimental effect of ketones on fetal brain
development.
A yet unanswered question is whether IDM develop diabetes mellitus during postnatal life. Children
and adults who were IDM have an increased incidence of diabetes mellitus. Limited data suggest
that offspring of fathers with insulin-dependent diabetes have a five-fold greater risk for
development of diabetes mellitus than offspring ofinsulin-dependent diabetic mothers. Although
diabetes is in part a genetic disorder, it has not been possible to delineate precisely the mode of
inheritance in IDM. It is possible that the altered metabolic state of the diabetic pregnancy may
modulate this genetic predisposition. Noninsulin-dependent diabetes mellitus occurs by age 20
years in 45.5% of infants of insulin-dependent diabetic mothers but in only 8.6% and 1.4% of
prediabetic and nondiabetic mothers, respectively. The mechanisms by which alterations in
maternal glucose and other metabolic fuels alter fetal beta-cell function are unknown.
REFERENCES
1. Freinkel N. Banting lecture 1980. Of pregnancy and progeny. Diabetes 1980;29:1023-1035.
2. Hytten ET, Leitch I. The physiology of human pregnancy, 2nd ed. Oxford: Blackwell Scientific,
1971.
3. Kalkhoff R, Schalch DS, Walker JL, et al. Diabetogenic factors associated with pregnancy.
Trans Assoc Am Physicians 1964;77: 270-280.
4. Lind T, Billewicz WZ, Brown G. A serial study of changes occurring in the oral glucose
tolerance test during pregnancy. J Obstet Gynaecol Br Commonw 1973;80:1033-1039.
5. Spellacy WN, Goetz FC. Plasma insulin in normal late pregnancy. N Engl J Med 1963;268:988991.
6. Catalano PM, Tyzbir ED, Roman NM, et al. Longitudinal changes in insulin release and insulin
resistance in nonobese pregnant women. Am J Obstet Gynecol 1991;165:1667-1672.
7. Spellacy WN, Goetz FC, Greenberg BZ, et al. Plasma insulin in normal early pregnancy.
8. Bleicher SJ, O'Sullivan JB, Freinkel N. Carbohydrate metabolism in pregnancy. V. The

interrelations of glucose, insulin and free fatty acids in late pregnancy and post-partum. N Engl J
Med 1964;271:866-872.
9. Grumbach M, Kaplan SL, Sciarra JJ, et al. Chronic growthhormone- prolactin (CGP) secretion;
disposition, biologic activity in man, and postulated function as the growth hormone of the
second half of pregnancy. Ann N Y Acad Sci 1968;148: 501-531.
10. Johansson EDB. Plasma levels of progesterone in pregnancy measured by a rapid

competitive binding technique. Acta Endocrinol (Copenh) 1969;61:607-617.
11. Buchanan TA, Catalano PM. The pathogenesis of GDM: implications for diabetes after
pregnancy. Diabetes Rev 1995;3:584-601.
12. Tsibris JCM, Raynor LO, Buhi WC, et al. Insulin receptors in circulating erythrocytes and
monocytes from women on oral contraceptives or pregnant women near term. J Clin Endocrinol
Metab 1980;51:711-717.
13. Moore P, Kolterman O, Weyant J, et al. Insulin binding in human pregnancy: comparisons to
the postpartum, luteal, and follicular states. J Clin Endocrinol Metab 1981;52:937-941.
P.889
14. Pedersen J. The pregnant diabetic and her newborn, 2nd ed. Baltimore: Williams & Wilkins,
1977.
15. Felig P, Lynch V. [Starvation in human pregnancy: hypoglycemia, hypoinsulinemia, and

hyperketonemia.] Science 1970; 170:990-992.
16. Kalhan SC, D'Angelo LJ, Savin SM, et al. Glucose production in pregnant women at term
gestation. Sources of glucose for human fetus. J Clin Invest 1979;63:388-394.
17. Ogata ES, Metzger BE, Freinkel N. Carbohydrate metabolism in pregnancy. XVI. Longitudinal
estimates of the effects of pregnancy on D-(6-3H) glucose and D-(6-14C) glucose turnovers
during fasting in the rat. Metabolism 1981;30:487-492.
18. Scow RO, Chernick SS, Brinley MS. Hyperlipemia and ketosis in the pregnant rat. Am J
Physiol 1964;206:796-804.
19. Patel MS, Johnson CA, Rajan R, et al. The metabolism of ketone bodies in developing human
brain: development of ketone-body-utilizing enzymes and ketone bodies as precursors for lipid
synthesis. J Neurochem 1975;25:905-908.
20. Churchill JA, Berendes HW, Nemore J. Neuropsychological deficits in children of diabetic
mothers. Am J Obstet Gynecol 1969;105:257-268.
21. Shelley HJ. Glycogen reserves and their changes at birth and in anoxia. Br Med Bull
1961;17:137-143.
22. Schwartz AL, Rall TW. Hormonal regulation of incorporation of alanine-U-14C into glucose in
human fetal liver explants. Effect of dibutyryl cyclic AMP, glucagon, insulin, and triamcinolone.
Diabetes 1975;24:650-657.
23. Schwartz AL, Rall TW. Hormonal regulation of glycogen metabolism in human fetal liver. II.
Regulation of glycogen synthase activity. Diabetes 1975;24:1113-1122.
24. Greengard O. Enzymatic differentiation of human liver: comparison with the rat model.
Pediatr Res 1977;11:669-676.
25. Raiha NC, Lindros KO. Development of some enzymes involved in gluconeogenesis in human
liver. Ann Med Exp Biol Fenn 1969;47:146-150.
26. Kaplan SL, Grumbach MM, Shepard TH. The ontogenesis of human fetal hormones. I.
Growth hormone and insulin. J Clin Invest 1972;51:3080-3093.
27. Assan R, Buillet J. Pancreatic glucagon and glucagon-like material in tissues and plasma
from human fetuses 6 to 26 weeks old. In: Jonxis JHP, Visser HKA, Troelstra JA, eds. Metabolic
processes in the fetus and newborn infant, nutrition symposium. Baltimore: Williams & Wilkins,
1971:210-.
28. Schaeffer LD, Wilder ML, Williams RH. Secretion and content of insulin and glucagon in
human fetal pancreas slices in vitro. Proc Soc Exp Biol Med 1973;143:314-319.
29. Grasso S, Distefano G, Messina A, et al. Effect of glucose priming on insulin response in the
premature infant. Diabetes 1975;24:291-294.
30. Milner RDG. The development of insulin secretion in man. In: Jonxis JHP, Visser HKA,
Troelstra JA, eds. Metabolic processes in the fetus and newborn infant, nutrition symposium.
Baltimore: Williams & Wilkins, 1971:310.
31. Grasso S, Messina A, Distefano G, et al. Insulin secretion in the premature infant. Response
to glucose and amino acids. Diabetes 1973;22:349-353.
32. Wise JK, Lyall SS, Hendler R, et al. Evidence of stimulation of glucagon secretion by alanine
in the human fetus at term. J Clin Endocrinol Metab 1973;37:341-344.
33. Hill DE. Effect of insulin on fetal growth. Semin Perinatol 1978;2:319-328.
34. Susa JB, McCormick KL, Widness JA, et al. Chronic hyperinsulinemia in the fetal rhesus
monkey: effects on fetal growth and composition. Diabetes 1979;28:1058-1063.
35. Ogata ES, Sabbagha R, Metzger BE, et al. Serial ultrasonography to assess evolving fetal
macrosomia. Studies in 23 pregnant diabetic women. JAMA 1980;243:2405-2408.
36. Pedersen J. Weight and length at birth of infants of diabetic mothers. Acta Endocrinol
(Copenh) 1954;16:330-342.
37. Girard JR, Caquet D, Bal D. Control of rat liver phosphorylase, glucose-6-phosphatase and
phosphoenolpyruvate carboxykinase activities by insulin and glucagon during the perinatal
period. Enzyme 1973;15:272-285.
38. Girard JR, Ferr A, Kervran A, et al. Role of the insulin/glucagon ratio in the changes of
hepatic metabolism during development of the rat. In: Foa PP, Bajaj JS, Foa NL, eds. Glucagon:
its role in physiology and clinical medicine. New York: Springer-Verlag, 1977:563-581.
39. Sperling MA, DeLamater PV, Phelps D, et al. Spontaneous and amino acid-stimulated
glucagon secretion in the immediate postnatal period. Relation to glucose and insulin. J Clin
Invest 1974;53:1159-1166.
40. Padbury J, Agata Y, Ludlow J, et al. Effect of fetal adrenalectomy on catecholamine release
and physiologic adaptation at birth in sheep. J Clin Invest 1987;80:1096-1103.
41. Agata Y, Padbury JF, Ludlow JK, et al. The effect of chemical sympathectomy on
catecholamine release at birth. Pediatr Res 1986; 20:1338-1344.
42. Grajwer LA, Sperling MA, Sack J, et al. Possible mechanisms and significance of the neonatal
surge in glucagon secretion: studies in newborn lambs. Pediatr Res 1977;11:833-836.
43. Molsted-Pedersen L. Aspects of carbohydrate metabolism in newborn infants of diabetic

mothers. II. Neonatal changes in K values. Acta Endocrinol (Copenh) 1972;69:189-194.
44. Bell GI, Kayano T, Buse JB, et al. Molecular biology of mammalian glucose transporters.
Diabetes Care 1990;13:198-208.
45. Devaskar SU, Mueckler MM. The mammalian glucose transporters. Pediatr Res 1992;31:1-13.
46. Takata K, Kasahara T, Kasahara M, et al. Localization of erythrocyte/ Hep G2 type glucose
transport (GLUT1) in human placental villi. Cell Tissue Res 1992;267:407-412.
47. Simmons RA, Flozak A, Ogata ES. The effect of insulin and insulin-like growth factor-I on
glucose transport in normal and small for gestational age fetal rats. Endocrinology
1993;133:1361-1368.
48. Sciullo E, Cardellini G, Baroni MG, et al. Glucose transporter (Glut1, Glut3) mRNA in human
placenta of diabetic and non-diabetic pregnancies. Early Pregnancy 1997;3:172-182.
49. Bier DM, Arnold KJ, Sherman WR, et al. In-vivo measurement of glucose and alanine
metabolism with stable isotopic tracers. Diabetes 1977;26:1005-1015.
50. Bier DM, Leake RD, Haymond MW, et al. Measurement of true glucose production rates in
infancy and childhood with 6,6-dideuteroglucose. Diabetes 1977;26:1016-1023.
51. Kalhan SC, Bier DM, Savin SM, et al. Estimation of glucose turnover and 13C recycling in the
human newborn by simultaneous [1-13C]glucose and [6,6-1H2]glucose tracers. J Clin Endocrinol

Metab 1980;50:456-460.
52. Kalhan SC, Savin SM, Adam PAJ. Measurement of glucose turnover in the human newborn
with glucose-1-13C. J Clin Endocrinol Metab 1976;43:704-707.
53. Sunebag A, Ewald U, Larsson A, et al. Glucose production rate in extremely immature
neonate (<28 weeks) studied by use of deuterated glucose. Pediatr Res 1993;33:97-100.
54. Cornblath M, Schwartz R. Disorders of carbohydrate metabolism in infancy, 3rd ed. Boston:
Blackwell Scientific, 1991.
55. Pagliara AS, Karl IE, Haymond M, et al. Hypoglycemia in infancy and childhood: parts I and
II. J Pediatr 1973;82:365- 379,558-577.
56. Ogata ES. Carbohydrate metabolism in the fetus and neonate and altered neonatal
glucoregulation. Pediatr Clin North Am 1986;33:25-45.
57. Lubchenco LO, Bard H. Incidence of hypoglycemia in newborn infants classified by birth
weight and gestational age. Pediatrics 1971;47:831-838.
58. Greene HL. Glycogen storage disease. Semin Liver Dis 1982;2: 291-301.
59. Haymond MW, Karl IE, Pagliara AS. Increased gluconeogenic substrates in the small-forgestational-age infant. N Engl J Med 1974;291:322-328.
60. Mestyan J, Soltesz G, Schultz K, et al. Hyperaminoacidemia due to the accumulation of

gluconeogenic amino acid precursors in hypoglycemic small-for-gestational age infants. J Pediatr
1975; 87:409-414.
61. Bussey ME, Finley S, LaBarbera A, et al. Hypoglycemia in the newborn growth-retarded rat:
delayed phosphoenolpyruvate carboxykinase induction despite increased glucagon availability.
Pediatr Res 1985;19:363-367.
62. Pollak A, Susa JB, Stonestreet BS, et al. Phosphoenolpyruvate carboxykinase in

experimental intrauterine growth retardation in rats. Pediatr Res 1979;13:175-177.
63. Sinclair JC, Silverman WA. Intrauterine growth in active tissue mass of the human fetus,
with particular reference to the undergrown baby. Pediatrics 1966;38:48-62.
P.890
64. Vidnes J, Sovik O. Gluconeogenesis in infancy and childhood. II. Studies on the glucose
production from alanine in three cases of persistent neonatal hypoglycaemia. Acta Paediatr
Scand 1976; 65:297-305.
65. Vidnes J, Oyasaeter S. Glucagon deficiency causing severe neonatal hypoglycemia in a

patient with normal insulin secretion. Pediatr Res 1977;11:943-949.
66. Garces LY, Drash A, Kenny FM. Islet cell tumor in the neonate. Studies in carbohydrate
metabolism and therapeutic response. Pediatrics 1968;41:789-796.
67. Heitz PU, Kloppel G, Hacki WH, et al. Nesidioblastosis: the pathologic basis of persistent
hyperinsulinemic hypoglycemia in infants. Morphologic and quantitative analysis of seven cases
based on specific immunostaining and electron microscopy. Diabetes 1977;26:632-642.
68. Salinas ED Jr, Mangurten HH, Roberts SS, et al. Functioning islet cell adenoma in the
newborn. Report of a case with failure of diazoxide. Pediatrics 1968;41:646-653.
69. Otonkoski T, Ammala C, Huopio H, et al. A point mutation inactivating the sulfonylurea
receptor causes the severe form of persistent hyperinsulinemic hypoglycemia of infants in
Finland. Diabetes 1999;48:408-415.
70. Fournet JC, Verkarre V, De Longlay P, et al. Loss of imprinted genes and paternal SUR1
mutations lead to hyperinsulinism in focal adenomatous hyperplasia. Ann Endocrinol (Paris)
1998;59: 485-491.
71. de Lonlay-Debeney P, Fournet JC, Martin D, et al. [Persistent hyperinsulinemic hypoglycemia

in the newborn and infants.] Arch Pediatr 1998;5:1347-1352.
72. Aguilar-Bryan L, Bryan J. Molecular biology of adenosine triphosphate-sensitive potassium

channels. Endocr Rev 1999;20: 101-135.
73. Otonkoski T, Andersson S, Simell O. Somatostatin regulation of beta-cell function in the

normal human fetuses and in neonates with persistent hyperinsulinemic hypoglycemia. J Clin
Endocrinol Metab 1993;76:184-188.
74. Leibowitz G, Glaser B, Higazi AA, et al. Hyperinsulinemic hypoglycemia of infancy

(nesidioblastosis) in clinical remission: high incidence of diabetes mellitus and persistent betacell dysfunction at long-term follow-up. J Clin Endocrinol Metab 1995;80:386-392.
75. Beckwith JB. Macroglossia, omphalocele, adrenal cytomegaly, gigantism, and hyperplastic
visceromegaly. Birth Defects 1969;5:188-196.
76. Lazarus L, Young JD, Friend JC. EMG syndrome and carbohydrate metabolism. Lancet
1968;2:1347-1348.
77. Normal AM, Read AP, Clayton-Smith J, et al. Recurrent Wiedemann-Beckwith syndrome with
inversion of chromosome (11) (p11.2p15.5). Am J Med Genet 1992;42:638-641.
78. Weksburg R, Shen DR, Fei YL, et al. Disruption of insulin-like growth factor 2 imprinting in
Beckwith-Wiedemann syndrome. Nat Genet 1993;5:143-150.
79. Barrett CT, Oliver TK Jr. Hypoglycemia and hyperinsulinism in infants with erythroblastosis
fetalis. N Engl J Med 1968;278: 1260-1262.
80. Steinke J, Gries FA, Driscoll SG. In vitro studies of insulin inactivation with reference to
erythroblastosis fetalis. Blood 1967;30: 359-363.
81. Socol ML, Dooley SL, Ney JA, et al. Absence of hyperinsulinemia in isoimmunized fetuses
treated with intravascular transfusion. Am J Obstet Gynecol 1991;165:1737-1740.
82. Senior B, Slone D, Shapiro S, et al. Benzothiadiazides and neonatal hypoglycaemia. Lancet
1976;2:377.
83. Zucker P, Simon G. Prolonged symptomatic neonatal hypoglycemia associated with maternal
chlorpropamide therapy. Pediatrics 1968;42:824-825.
84. Brazy JE, Pupkin MJ. Effects of maternal isoxsuprine administration on preterm infants. J
Pediatr 1979;94:444-448.
85. Ogata ES. Isoxsuprine infusion in the rat: alterations in maternal, fetal and neonatal glucose
homeostasis. J Perinat Med 1981;9:293-301.
86. Kenepp NB, Kumar S, Shelley WC, et al. Fetal and neonatal hazards of maternal hydration
with 5% dextrose before caesarean section. Lancet 1982;1:1150-1152.
87. Nagel JW, Sims S, Aplin CE 2nd, et al. Refractory hypoglycemia associated with a
malpositioned umbilical artery catheter. Pediatrics 1979;64:315-317.
88. Cowett RM, Tenenbaum DG, Fatoba O, et al. The effects of arterial glucose infusion above
the celiac axis in the neonatal lamb. Biol Neonate 1985;47:179-185.
89. Kliegman R, Gross T, Morton S, et al. Intrauterine growth and postnatal fasting metabolism
in infants of obese mothers. J Pediatr 1984;104:601-607.
90. Leake RD, Fiser RH Jr, Oh W. Rapid glucose disappearance in infants with infection. Clin
Pediatr (Phila) 1981;20:397-401.
91. Yeung CY, Lee VWY, Yeung CM. Glucose disappearance rate in neonatal infection. J Pediatr
1973;82:486-489.
92. Wiswell TE, Cornish JD, Northam RS. Neonatal polycythemia: frequency of clinical
manifestations and other associated findings. Pediatrics 1986;78:26-30.
93. Rosenkrantz TS, Philipps AF, Skrzypczak PS, et al. Cerebral metabolism in the newborn lamb
with polycythemia. Pediatr Res 1988;23:329-333.
94. Lovinger RD, Kaplan SL, Grumbach MM. Congenital hypopituitarism associated with neonatal
hypoglycemia and microphallus: four cases secondary to hypothalamic hormone deficiencies. J
Pediatr 1975;87:1171-1181.
95. Johnson JD, Hansen RC, Albritton WL, et al. Hypoplasia of the anterior pituitary and neonatal
hypoglycemia. J Pediatr 1973; 82:634-641.
96. Kauschansky A, Genel M, Smith GJ. Congenital hypopituitarism in female infants. Its
association with hypoglycemia and hypothyroidism. Am J Dis Child 1979;133:165-169.
97. Daaboul JJ, Kartchner W, Jones KL. Neonatal hypoglycemia caused by hypopituitarism in
infants with congenital syphilis. J Pediatr 1993;123:983-985.
98. Bower BD, Jones LF, Weeks MM. Cold injury in the newborn. A study of 70 cases. Br Med J
1960;5169:303-309.
99. Schiff D, Stern L, Leduc J. Chemical thermogenesis in newborn infants: catecholamine

excretion and the plasma non-esterified fatty acid response to cold exposure. Pediatrics
1966;37:577- 582.
100. Pickering D, Ellis H. Neonatal hypoglycaemia due to salicylate poisoning. Proc R Soc Med
1968;61:1256.
101. Gemelli M, De Luca F, Barberio G. Hypoglycaemia and congenital adrenal hyperplasia. Acta
Paediatr Scand 1979;68:285-286.
102. Smith VS, Giacoia GP. Hyperinsulinaemic hypoglycaemia in an infant with mosaic trisomy
13. J Med Genet 1985;22:228-230.
103. Pildes RS. Neonatal hyperglycemia. J Pediatr 1986;109:905-907.
104. Dweck HS, Cassady G. Glucose intolerance in infants of very low birth weight. I. Incidence
of hyperglycemia in infants of birth weights 1,100 grams or less. Pediatrics 1974;53:189-195.
105. Cowett RM, Oh W, Schwartz R. Persistent glucose production during glucose infusion in the
neonate. J Clin Invest 1983;71: 467-475.
106. Hertz DG, Karn CA, Liu YM, et al. Intravenous glucose suppresses glucose production but
not proteolysis in extremely premature newborns. J Clin Invest 1993;92:1752-1758.
107. Lilien LD, Rosenfield RL, Baccaro MM, et al. Hyperglycemia in stressed small premature
neonates. J Pediatr 1979;94:454-459.
108. Massi-Benedetti F, Falorni A, Luyckx A, et al. Inhibition of glucagon secretion in the human
newborn by simultaneous administration of glucose and insulin. Horm Metab Res 1974;6: 392396.
109. Zarif M, Pildes RS, Vidyasagar D. Insulin and growth-hormone responses in neonatal
hyperglycemia. Diabetes 1976;25:428-433.
110. Grasso S, Fallucca F, Mazzone D, et al. Inhibition of glucagon secretion in the human
newborn by glucose infusion. Diabetes 1983;32:489-492.
111. Goldman SL, Hirata T. Attenuated response to insulin in very low birthweight infants.
Pediatr Res 1980;14:50-53.
112. Gentz JCH, Cornblath M. Transient diabetes of the newborn. Adv Pediatr 1969;16:345-363.
113. Hutchison JH, Keay AJ, Kerr MM. Congenital temporary diabetes mellitus. Br Med J
1962;5302:436-440.
114. Vaucher YE, Walson PD, Morrow G 3rd. Continuous insulin infusion in hyperglycemic, very
low birth weight infants. J Pediatr Gastroenterol Nutr 1982;1:211-217.
115. Binder ND, Raschko PK, Benda GI, et al. Insulin infusion with parenteral nutrition in
extremely low birth weight infants with hyperglycemia. J Pediatr 1989;114:273-280.
116. Ostertag SG, Jovanovic L, Lewis B, et al. Insulin pump therapy in the very low birth weight
infant. Pediatrics 1986;78:625-630.
117. Collins JW Jr, Hoppe M, Brown K, et al. A controlled trial of insulin infusion and parenteral
nutrition in extremely low birth weight infants with glucose intolerance. J Pediatr 1991;118: 921927.
P.891
118. Creery RDG. Hypoglycaemia in the newborn: diagnosis, treatment and prognosis. Dev Med
Child Neurol 1966;8:746-754.
119. Haworth JC. Neonatal hypoglycemia: how much does it damage the brain? Pediatrics
1974;54:3-4.
120. Haworth JC, McRae KN. Neonatal hypoglycemia: a six-year experience. J Lancet
1967;87:41-45.
121. Haworth JC, Vidyasagar D. Hypoglycemia in the newborn. Clin Obstet Gynecol
1971;14:821-839.
122. Koivisto M, Blanco-Sequeiros M, Krause U. Neonatal symptomatic and asymptomatic

hypoglycaemia: a follow-up study of 151 children. Dev Med Child Neurol 1972;14:603-614.
123. Pildes RS, Cornblath M, Warren I, et al. A prospective controlled study of neonatal
hypoglycemia. Pediatrics 1974;54:5-14.
124. Pildes R, Forbes AE, O'Connor SM, et al. The incidence of neonatal hypoglycemiaa
completed survey. J Pediatr 1967;70: 76-80.
125. Raivio KO. Neonatal hypoglycemia. II. A clinical study of 44 idiopathic cases with special
reference to corticosteroid treatment. Acta Paediatr Scand 1968;57:540-546.
126. Griffiths AD, Bryant GM. Assessment of effects of neonatal hypoglycaemia. A study of 41
cases with matched controls. Arch Dis Child 1971;46:819-827.
127. Lucas A, Morley R, Cole TJ. Adverse neurodevelopmental outcome of moderate neonatal
hypoglycaemia. BMJ 1988;297: 1304-1308.
128. Fee BA, Weil WB Jr. Body composition of infants of diabetic mothers by direct analysis. Ann
N Y Acad Sci 1963;110:869-897.
129. Roth S, Abernathy MP, Lee WH, et al. Insulin-like growth factors I and II peptide and
messenger RNA levels in macrosomic infants of diabetic pregnancies. J Soc Gynecol Investig
1996;3: 78-84.
130. Ogata ES, Freinkel N, Metzger BE, et al. Perinatal islet function in gestational diabetes:
assessment by cord plasma C-peptide and amniotic fluid insulin. Diabetes Care 1980;3:425-429.
131. Eriksson UJ, Lewis NJ, Freinkel N. Growth retardation during early organogenesis in
embryos of experimentally diabetic rats. Diabetes 1984;33:281-284.
132. Sosenko IR, Kitzmiller JL, Loo SW, et al. The infant of the diabetic mother: correlation of
increased cord C-peptide levels with macrosomia and hypoglycemia. N Engl J Med 1979;301:
859-862.
133. Stern L, Ramos A, Leduc J. Urinary catecholamine excretion in infants of diabetic mothers.
Pediatrics 1968;42:598-605.
134. Schedewie HK, Odell WD, Fisher DA, et al. Parathormone and perinatal calcium
homeostasis. Pediatr Res 1979;13:1-6.
135. Noguchi A, Eren M, Tsang RC. Parathyroid hormone in hypocalcemic and normocalcemic
infants of diabetic mothers. J Pediatr 1980;97:112-114.
136. Stevenson DK, Bartoletti AL, Ostrander CR, et al. Pulmonary excretion of carbon monoxide
in the human infant as an index of bilirubin production. II. Infants of diabetic mothers. J Pediatr
1979;94:956-958.
137. Widness JA, Susa JB, Garcia JF, et al. Increased erythropoiesis and elevated erythropoietin
in infants born to diabetic mothers and in hyperinsulinemic rhesus fetuses. J Clin Invest
1981;67: 637-642.
138. Philips AF, Dubin JW, Matty PJ, et al. Arterial hypoxemia and hyperinsulinemia in the
chronically hyperglycemic fetal lamb. Pediatr Res 1982;16:653-658.
139. Robert MF, Neff RK, Hubbell JP, et al. Association between maternal diabetes and the
respiratory-distress syndrome in the newborn. N Engl J Med 1976;294:357-360.
140. Bourbon JR, Farrell PM. Fetal lung development in the diabetic pregnancy. Pediatr Res
1985;19:253-267.
141. Gutgesell HP, Speer ME, Rosenberg HS. Characterization of the cardiomyopathy in infants
of diabetic mothers. Circulation 1980;61:441-450.
142. Miller E, Hare JW, Cloherty JP, et al. Elevated maternal hemoglobin A1c in early pregnancy
and major congenital anomalies in infants of diabetic mothers. N Engl J Med 1981;304:13311334.
143. Fuhrmann K, Reiher H, Semmler K, et al. Prevention of congenital malformations in infants

of insulin-dependent diabetic mothers. Diabetes Care 1983;6:219-223.
144. Simpson JL, Elias S, Martin AO, et al. Diabetes in pregnancy, Northwestern University
series (1977-1981). I. Prospective study of anomalies in offspring of mothers with diabetes
mellitus. Am J Obstet Gynecol 1983;146:263-270.
145. Freinkel N, Lewis NJ, Akazawa S, et al. The honeybee syndrome implications of the
teratogenicity of mannose in rat-embryo culture. N Engl J Med 1984;310:223-230.
146. Mills JL, Knopp RH, Simpson JL, et al. Lack of relation of increased malformation rates in
infants of diabetic mothers to glycemic control during organogenesis. N Engl J Med 1988;318:
671-676.
147. Barker DJ, Eriksson JG, Forsen T, et al. Fetal origins of adult disease: strength of effects
and biological basis. Int J Epidemiol 2002;31:1235-1239.
148. Vohr BR, Lipsitt LP, Oh W. Somatic growth of children of diabetic mothers with reference to
birth size. J Pediatr 1980;97: 196-199.
149. Metzger BE, Silverman BL, Freinkel N, et al. Amniotic fluid insulin concentration as a
predictor of obesity. Arch Dis Child 1990;65:1050-1052.
150. Rizzo T, Metzger BE, Burns WJ, et al. Correlations between antepartum maternal
metabolism and child intelligence. N Engl J Med 1991;325:911-916
Chapter 38
Congenital Anomalies
Scott Douglas McLean
Much of our practice of pediatrics and neonatology is geared toward the provision of acute care, such as
resuscitation, antibiotics, and intravenous fluids. When we encounter a newborn with one or more birth defects,
this acute-care paradigm remainsand appropriately soin the forefront of our clinical perspective. However,
the child with a congenital anomaly must also be simultaneously considered from a different perspective, one
that looks far back and far forward in time. When did this structuralnormality have its origin? Was there a
teratogenic exposure during a critical period of organogenesis? Has a new mutation occurred in one of the
parents' germ cells? Are some of the DNA alterations in this newborn ancient relics that only now reveal
themselves in the living world? Is this congenital anomaly a clue to the presence of other pathology? Will our
best efforts have a reasonable chance of ensuring long-term health? Will siblings or offspring be similarly affected
Over the past several decades, research has dramatically improved our insights into the genetic and
environmental causes of many isolated birth defects, multiple congenital anomaly syndromes, and other genetic
conditions. In some instances, the molecular pathology has been dissected in impressive detail, such as for
cystic fibrosis (CF) and the CF transmembrane regulator gene on chromosome 7q31. For other conditions, such
as the VACTERL association, we have yet to understand the genomic origins but have accumulated much clinical
and epidemiologic data that allow us to formulate helpful diagnostic criteria, predict prognosis, and estimate
recurrence risks. For any individual newborn with a congenital anomaly, establishing an accurate diagnosis is the
key to intelligent clinical management in both the short and long term.
In a busy nursery, caring for a newborn with a congenital malformation is nearly a daily activity. Although the
consultative services of a clinical geneticist or dysmorphologist often are quite helpful, this resource may not be
available, and the attending pediatrician or neonatologist must assume that role, at least temporarily. Naturally,
such a task brings many challengescognitive, managerial, and emotionalfor which the practitioner often feels
underprepared. Useful tools include a careful prenatal history, three-generation pedigree, meticulous physical
examination, and sensitivity for the emotional and social impact of the birth of a malformed child. A dramatically
visible departure from the norm multiplies a parent's fear, concern, worry, and guilt and might also become very
stressful for nurses and physicians. Consequently, to optimize care for the infant and to instill confidence and
control in an intrinsically unsettling situation, the clinical leaders of the health care team should prepare
themselves by becoming familiar with a general strategy for diagnosis and management of an infant with
congenital malformations.
In aggregate, birth defects are ubiquitously common and have likely been so throughout human history. Epidemiologic studies consistently place the incidence of major malformations in newborns at 2% to 3% (1,2).
Many other neonates harbor occult anomalies that are eventually detected later in childhood, giving a
cumulative rate of major birth defects of approximately 4%, or one of every 25 newborn children. In the United
States, 150,000 children with birth defects are born each year. Congenital anomalies are the leading cause of
infant mortality and the second leading cause of death in children between the ages of 1 and 4 years (3).
Excluding the intangible costs of pain and suffering, the lifetime economic cost per child in 1992 ranged from
$75,000 to $503,000 (4). The Centers for Disease Control and the National Birth Defects Prevention Network
conduct and facilitate a number of research and surveillance activities, central to which is the commitment of
individual practitioners to diagnose and report all congenital defects accurately and reliably.
DEFINITIONS AND CLASSIFICATIONS

A congenital anomaly is any alteration, present at birth, of normal anatomic structure. It may be major or
minor, isolated or part of a larger constellation of defects, of clear or uncertain cause. Several genetic and
environmental etiologies are well delineated (Table 38-1), but the fundamental etiology of nearly half of all birth
defects is unknown (5).
The nomenclature for various birth defects and malformation syndromes can be a source of confusion for both
the novice and veteran clinician alike. Categories of anomalies
P.893
may seem arbitrary and capricious, but a developmental or embryologic perspective often will cast light on the
rationale behind a particular designation. The term birth defect, enjoys wide usage and conveys immediate
meaning for parents. Congenital anomaly is fundamentally equivalent, indicating an abnormality of anatomic
structure present at birth and may be further refined in terms of severity (major and minor), pathogenesis
(malformation, deformation, disruption, dysplasia), or pattern (isolated, syndromic). These terms are
defined in Table 38-2.
TABLE 38-1 CAUSES OF MALFORMATIONS IN NEWBORNSa
Percent
Chromosome abnormalities
Single mutant genes
Familial
Multifactorial inheritance
Teratogens
10.1
3.1
14.5
23.0
3.2
Uterine factors
Twinning
Unknown cause
2.5
0.4
43.2
Modified from Nelson K, Holmes LB. Malformations due to presumed spontaneous mutations in
newborn infants. N Engl J Med 1989; 320:19, with permission.
The great majority of congenital anomalies occur in isolation, as a single phenomenon, and are postulated to
arise because of a primary, intrinsic malformation of a fetal structure that occurs at 10 weeks of gestation or
earlier. On occasion, other family members are similarly affected, implying the presence of a single gene
mutation that behaves as an autosomal dominant, autosomal recessive, or X-linked trait. More commonly,
however, the family history is entirely bereft of other affected individuals. Classic mendelian genetics does not
adequately explain this situation, but another modelmultifactoral inheritancehas proven quite useful.
TABLE 38-2 Terminology
Major anomaly
Minor anomaly
Minor variant
Malformation
Deformation
An anatomic abnormality severe enough to reduce normal life expectancy

or compromise normal function; e.g., neural tube defect, cleft lip.
A structural alteration that either requires no treatment or can be
corrected in a straightforward manner, with no permanent consequences,
and present in less than 4% of the normal population; e.g., preauricular
skin tag, small ventricular septal defect.
A physical feature, often familial, that is present in only a small
proportion (1%-5%) of normal individuals, e.g., simian crease of the
palm, epicanthal folds.
A morphologic defect of an organ, part of an organ, or region due to an
intrinsically abnormal developmental process; e.g., microphthalmia,
ectrodactyly.
An abnormal form, shape, or position of a part of the body caused by
unusual mechanical forces on normal tissue; e.g., club foot,
plagiocephaly.
Disruption
Sequence
Developmental field defect
A morphologic defect due to extrinsic interference with a normal

developmental process resulting in breakdown of normal tissue; e.g.,
amniotic band sequence, fetal alcohol syndrome.
Several anomalies that occur due to a cascade of events, caused by a
single initiating event or anomaly; e.g., Potter sequence, Pierre Robin
sequence.
A set of morphologic defects that share a common or contiguous region
during embryogenesis;e.g., hemifacial microsomia.
Association
Syndrome
A nonrandom set of anomalies in which the specific components occur

together more frequently than would be expected by chance and for
which cause is not established; e.g., VACTERL association, CHARGE
association.
A recognizable pattern of anomalies that run together; e.g., Down
syndrome, Smith-Lemli-Opitz syndrome.

VACTERL: Vertebral defects, imperforate Anus, congenital Cardiac anomalies, Tracheo Esophageal
fistula, Radial defects, Renal dysplasia, and other Limb malformations.
CHARGE: Coloboma, Heart disease, Atresia choanae, Retarded growth and development and/or CNS
anomalies, Genital hypoplasia, and Ear anomalies or deafness.
In the multifactoral model, a constellation of influences, including multiple genes inherited from each parent,
and poorly defined environmental influences, seems to allow a developing fetal structure to cross a threshold of
liability, beyond which morphogenesis proceeds abnormally (6). These birth defects tend to recur at a low
rate, approximately 3% to 5% for each subsequent pregnancy for the parents of one affected child, 10% to 15%
if two siblings have previously been similarly affected.
Multiple congenital malformations, on the other hand, are caused by a very different spectrum of fundamental
problems. For a child with several major anomalies, the underlying cause is more likely to be a malformation
sequence, developmental field defect, association, or syndrome, each of which may in turn may be caused by a
chromosomal anomaly, a single-gene mutation, a teratogen, or unknown factors. Minor anomalies deserve
special attention because, in terms of diagnostic significance, they are frequently the equals of the major
malformations (7), often providing the linchpin of clinical recognition of a rare
P.894
syndrome. A single minor anomaly is present in 13% of neonates, but more than one is decidedly less common
two minor anomalies occur in only 1%, and three minor anomalies in just 0.05% (8). As the numbers of minor
anomalies increase, the search for occult major anomalies and the consideration that a syndromic diagnosis is
possible should increase proportionately.
MANAGEMENT STRATEGY
Since the last edition of this textbook, the American College of Medical Genetics (ACMG), under the sponsorship
of the New York State Department of Health, has published clinical guidelines for health care practitioners who
care for newborn infants with one or more birth defects (9). This document incorporates expert opinion from a
broad array of disciplines and describes practical, detailed components of the history, physical examination,
differential diagnosis, diagnostic stratagem, genetic counseling, and record keeping. As such, the ACMG
guidelines represent a significant step forward in promoting a national standard of care with which
neonatologists, pediatricians, and family practitioners should become familiar.
The fundamental approach to managing an infant with one or more congenital anomalies is much the same as
the management of any other clinical scenario. Effective clinical intervention is organized around an
understanding of the natural history of the condition at hand. History taking begins with conception and includes
a detailed three-generation pedigree. Physical features must be scrutinized, measured, and documented with
precision, and confirmatory studies must be carefully chosen and accurately interpreted. Common pitfalls include
incomplete ascertainment of all relevant information, impetuous diagnosis and prognostication, and failure to
communicate with parents in a straightforward and compassionate manner.
Neonates with congenital malformations present with extremely variable clinical needs, so a one size fits all
strategy for management has limited practical value. However, a number of algorithmic approaches have proven
useful. In the ACMG clinical management algorithm (Fig. 38-1), an early determination of whether the
malformation is isolated or multiple helps organize subsequent clinical activity. Beyond this, the major thrust of
activity is devoted to pursuing and establishing a precise diagnosis, the essential starting point for
understanding natural history, designing effective intervention, and educating the family about recurrence risks
and prenatal management options for subsequent pregnancies. Another, more integrated approach suggested
by Dr. Judith Hall (10) delineates three parallel and simultaneous lines of activity, as outlined in Fig. 38-2.
Urgent interventions for the infant are undertaken immediately. Also, from the outset, the family is provided
with support in the form of information, preliminary interpretation, and acknowledgment of their distress and
concern. The third concurrent activity is data collectionhistory, physical, laboratory studies and imaging, and
preliminary definition of the nature of the problem. All efforts should be made to verify information and to amass
a complete database.
History
Prenatal information, beginning with conception, must be specifically sought from both the mother and her
obstetric records. The nature and timing of maternal illnesses, such as rubella or cytomegalovirus infections,
including febrile episodes, can suggest a direct infectious disruption. Maternal use of alcohol, drugs, medications,
or tobaccodegree, gestational timing, and durationshould be documented. Quickening, the date at which
there is maternal sensation of fetal movement, and the subsequent vigor and frequency of fetal activity, reflects
central nervous system integrity and peripheral muscular function. Prenatal ultrasonography, especially if
performed serially and in detail, may yield critical information regarding the amount of amniotic fluid and
malformations of the kidneys, brain, heart, skeleton, and gastrointestinal system. Recent reports have
suggested an association between in vitro fertilization and several genetic conditions, such as Angelman
syndrome (11), retinoblastoma (12) and Beckwith-Wiedemann syndrome (13). Chorionic villus sampling has
previously been suspected to cause limb-reduction deformities, but recent studies by the World Health
Organization affirm its safety for first-trimester prenatal diagnosis (14). A review of the ultrasound studies and
the obstetric record, and a discussion with the attending obstetrician or perinatologist, can save valu-able time
and effort. Maternal serum and amniotic fluid fetoprotein levels and prenatal chromosome studies should be
verified and documented in the infant's medical record. Late trimester problems with fetal position, such as
transverse or breech lie, potentially indicate neuromuscular or structural abnormalities of the fetus. Fetuses with
significant neuromuscular problems often encounter perinatal distress and have a poor transition to extrauterine
life. Finally, it often is useful to inquire about any prenatal event or factors that the parents suspect, even
remotely, may have caused their infant's problems. Concerns about witnessing a solar or lunar eclipse, for
example, are best made explicit, if only for the purpose of assuaging guilt.
Family History
A systematic and detailed inquiry into the age, health, development, and congenital anomalies of all members of
the immediate family comprises the core of an adequate family history. All second-degree relatives, such as
grandparents, aunts, uncles, nieces, and nephews, should be considered, and more distant relatives may
contribute valuable data. A three-generation pedigree provides a concise picture of patterns of inheritance. One
must specifically clarify the biologic parentage for each individual. Spontaneous abortions, miscarriages,
stillbirths, and infant deaths clearly are germane, but parents typically omit this information unless the
interviewer specifically inquires. Consanguinity also should be directly but tactfully questioned. A quick
pedigree is an
P.895
oxymoron; routinely the entire process requires patience and time. With this in mind, the interviewer will be
sympathetic, on the day of delivery, to parents who are understandably terrified, exhausted, and disoriented.
After rest, time, and a chance to speak with relatives, they may be more prepared to help refine and expand on
their complete family history.
Figure 38-1 Management algorithm for the infant with multiple congenital anomalies. Reprinted from American
College of Medical Genetics Foundation, sponsored by the New York State Department of Health. Evaluation of
the newborn with single or multiple congenital anomalies: a clinical guideline. May 1999. Available at http://
www.health.state.ny.us/nysdoh/dpprd/main.htm. Accessed 1/1/05, with permission.
It is axiomatic that a proper newborn physical examination is detailed and complete. The careful observer,
especially one who repeats the examination several times, will recognize departures from the norm. Several
points are worth keeping in mind when examining infants with a malformation or a generalized dysmorphic
appearance.
Be alert. There is a heightened chance that additional, initially unsuspected anomalies are present.
Although a malformation may be minor in severity, it might represent the most important, critical clue to
the diagnosis
Collect clues systematically, by closely examining each topographical segment of the body and by
scrutinizing progressively more detailed regions. For example, on
P.896
initial observation it may be noted that the fingers are disproportionately short; a closer examination may
reveal that the fourth and fifth fingers are stiff and rigidly extended, with faint flexion creases, especially
at the distal interphalangeal joints, that the nails are short and narrow, almost absent on the fifth finger.
The placenta may be examined with the obstetrician or pathologist to seek evidence of cryptic twinning,
umbilical cord anomalies, or amniotic bands
Document the examination with great care, using appropriate morphologic terms and sufficient detail, and
strongly consider supplementing written findings with clinical photographs. Parental consent for
photographs should be documented.
Measure those features that are obviously or potentially abnormal in size, shape, position, or symmetry.
Normal standards are available for virtually any anatomic structure, encompassing all ages from preterm
infant to adulthood (15,16). Hall (7) admonishes, Never make a clinical judgment on a measurable
parameter without measuring it.
Examine both parents, if possible, seeking any signs of similar anomalies. Dominant conditions frequently
manifest a subtle but distinctive phenotype in adults.
Figure 38-2 Integrated approach to management. Modified from reference 48.
Adjunctive Investigations
The emerging clinical picture will dictate imaging studies and consultations with pediatric subspecialists. For
example, a newborn with Down syndrome, even when a cardiac murmur is absent, merits the attention of a
pediatric cardiologist, because significant structural defects of the heart are present in 50% but may be missed
on clinical examination. A newborn girl with puffy hands and feet, a webbed neck, and coarctation of the aorta
also should receive a renal ultrasound, because kidney malformations commonly are associated with Turner
syndrome. When the diagnosis is unclear and several malformations are present, occult anomalies of the central
nervous system, heart, kidneys, vertebrae, and eyes are reasonable to pursue, especially when the known birth
defects are multiple and severe. Renal ultrasounds are often ordered when the neonate has a single umbilical
artery or an ear malformation, such as a preauricular pit. In the absence of supporting findings, such as other
malformations, a family history of deafness or renal anomalies, or maternal diabetes, this investigation is
unlikely to be useful (17). Skeletal films, to include hands, feet, long bones, pelvis, vertebrae, chest, and
cranium, are helpful when length is less then the 5th percentile for gestational age or when the limbs are
disproportionately short. These studies often require interpretation by a pediatric radiologist skilled in this area.
Chromosomal Analysis
Forty-six chromosomes are present in most normal human cells. Formation of ova and sperm, however, is a
surprisingly error-prone process: nearly two-thirds of all fertilizations result in aneuploidy or abnormal
chromosomal number or structure, with subsequent reproductive loss. Some of this prenatal loss occurs late
enough to be recognized as a miscarriage or spontaneous abortion, but most wastage is occult. Ninety-eight
percent of these chromosomal defects are lethal (13).
An abnormal karyotype occurs in one of 170 liveborn infants. Among chromosomally abnormal neonates, onethird have an extra sex chromosome with mild or no phenotypic manifestations in the newborn period, onefourth have trisomy of an autosome, such as trisomy 21 or trisomy 18, and 40% have a variation of
chromosomal structure, such as a deleted or duplicated segment or a translocation. Of the latter, most (79%)
are balanced and generally do not cause birth defects. Approximately 10% of infants who die in the perinatal
period secondary to multiple congenital malformations have abnormal cytogenetic studies (13).
Which infants deserve chromosomal studies? Truly isolated malformations are very infrequently caused by a
cytogenetic defect. On the other hand, neonates with multiple major malformations or a generalized dysmorphic
appearance, and stillborn infants, with or without malformations, should have cytogenetic testing. Between
these extremes lies a sizable gray area. Factors in favor of chromosomal testing include intrauterine growth
retardation, an abnormal neurologic exam, a major malformation accompanied by several minor malformations,
and a history
P.897
of multiple pregnancy losses for the mother or in close relatives.
Peripheral blood lymphocytes are the tissue of choice for most cytogenetic analysis, but many other tissues can
be used, including skin fibroblasts, bone marrow, placenta, and pericardium, usually harvested postmortem. Two
to three milliliters of venous or arterial blood, collected in a sodium heparin (green top) tube, are generally
sufficient and should be kept at room temperature or refrigerated, never frozen, in transit. The cells usually will
remain viable for 1 or 2 days, but the shortest possible transit time increases the chances for useful results. If
the sample is shipped, an overnight courier is recommended. The lymphocytes are separated, incubated in the
presence of a mitogen to stimulate cell division, which is then abruptly halted with colchicine, and the cell
membranes are disrupted gently while being placed on a glass slide. After enzymatic preparation and staining,
the spread of chromosomes is analyzed in approximately 20 cells. Generally, 550 or more bands are visible
with Giemsa staining of 46 chromosomes at metaphase. Several cells are photographed and arranged in
standard groupings, a karyotype. Turnaround time for cytogenetic analysis is typically three to four days,
although some laboratories are able to provide results in slightly less time. Because of the high percentage of
dividing cells in bone marrow, karyotypes from this tissue may be obtained in a matter of hours, although the
quality of the banding frequently is inadequate for high-resolution analysis of small or subtle abnormalities. To
achieve the latter, a special request for prometaphase analysis should accompany a peripheral blood sample.
In the past several years, an increasing number of syndromic conditions have been found to be caused by very
small chromosomal deletions that are not visible by routine karyotyping, even at pro-metaphase levels of detail.
Molecular probes that will hybridize at these loci with great specificity have been developed. These DNA probes
are complexed with a fluorescent marker and become powerful tools for detecting submicroscopic chromosomal
deletions. This technique, termed fluorescent in situ hybridization (FISH), has also been adapted for whole,
entire chromosomes and can identify the nature of many chromosomal anomalies, such as translocations. A
metaphase spread can, in essence, be painted with several single-locus or whole-chromosome FISH probes
simultaneously, providing a highly specific, and colorful, picture of genomic structure.
Among children with mental retardation of unknown etiology, in whom extensive investigations, including
standard cytogenetic studies, are normal, approximately 5% harbor very small, occult chromosomal deletions or
unbalanced translocations. A recently developed techniquesubtelomeric probe analysisuses FISH and/or
other molecular tools to ascertain whether the regions just proximal to the tips of each of the 46 chromosomes
are present in their normal locations. Although initially designed to investigate subnormal intelligence in older
children, subtelomeric analysis appears to also be useful for some infants with congenital anomalies in whom
standard investigations have not been fruitful. This technology seems to have better diagnostic success for
individuals with intrauterine growth retardation, microcephaly, and a positive family history (19).
Metabolic Studies
Inborn errors of metabolism are often assumed to have a purely biochemical or neurologic phenotype, but
metabolic disease is well recognized as an occasional cause of dysmorphic facial features and congenital
malformations (20). For instance, ambiguous genitalia are seen in some cases of 21-hydroxylase deficiency and
other types of congenital adrenal hyperplasia, and infants with pyruvate dehydrogenase deficiency may have
agenesis of the corpus callosum and facial features that resemble fetal alcohol syndrome. A number of the
congenital disorders of glycosylation feature congenital malformations of the heart, limbs, and central nervous
system (21). Many peroxisomal conditions result in distinctive phenotypes: Zellweger syndrome, also known as
cerebro-hepato-renal syndrome, and rhizomelic chondrodysplasia punctata (RCDP) exemplify this class of
disease. For the Zellweger spectrum of peroxisomal abnormalities, serum levels of very long chain fatty acids
will be elevated; in RCDP the very long chain fatty acids are normal but phytanic acid is elevated.
Synthesis and Analysis of Data

Every clinician develops a unique, individual strategy of diagnosis. Acquiring a deep and broad fund of
knowledge is arguably fundamental and, oftentimes, sufficient in itself. Unfortunately, the sheer numbers of
conditions and syndromes impose some limits on the brute force approach; the London Dysmorphology
Database, for instance, contains over 3,000 multiple congenital anomaly syndromes (22).
Precise diagnosis is neither possible nor necessary in the immediate newborn period for many neonates with a
multiple congenital anomaly syndrome. More than half of all individuals with congenital anomalies never receive
a firm diagnosis, even after the most definitive of workups, and remain an unknown. Categorization of an
anomaly as a malformation, disruption, or deformation, however, is a feasible and useful first step. These terms
have been defined and discussed previously. An isolated major anomaly in the absence of any similarly affected
relative would suggest a multifactoral etiology. If there are multiple malformations, one of the following
strategies may be useful:
Instant recognition, or gestalt diagnosis, which depends on the clinician's previous experience and
strength of visual memory. Certain caveats apply, however: many disorders have a considerable range of
phenotypic variation, and other conditions, or phenocopies, may mimic the one that has instantly come to
mind
Perusal of an atlas or illustrated text, such as Smith's Recognizable Patterns of Human Malformation, to
match a photograph with the patient. This simple strategy often yields excellent results.
P.898
Pattern analysis, in which all phenotypic and clinical problems are enumerated, grouped, combined,
recombined, and weighed to discern developmental relationships, sequences, and influences. Major organ
systems or classes of disease (e.g., skeletal dysplasias) then become entry points for further comparison,
matching the patient's pattern against published descriptions while attempting to take into account
phenotypic variability
Focusing the initial investigation on the anomaly that is most distinctive, rare, or unusual. Clinodactyly of
the fifth finger is very common, but a coloboma of the iris is fairly unusual. A variety of texts or electronic
databases then can be consulted and a relatively short list of diagnostic possibilities generated.
Once a preliminary analysis has generated a differential diagnosis, all reasonable efforts are made to test each
competing hypothesis. Often, a clinical finding can corroborate a possibility. For instance, a lateral radiograph of
the knee may allow confirmation of chondrodysplasia punctata by demonstrating the typical stippled, punctate
mineralization of the epiphyses. Although many diagnoses are purely clinical, molecular tools are becoming
extremely helpful. An electronic literature search or querying an internet resource, such as Online Mendelian
Inheritance in Man, can help determine whether a novel approach using direct deoxyribonucleic acid (DNA)
sequence analysis, FISH, or linkage analysis has been developed for the disease in question.
More often than not, the diagnostician will not be able to establish an etiology. When this is the case, there is a
temptation to force a diagnosis, analogous to hammering a somewhat square peg into a slightly round hole.
This may not be in the best interests of the patient so that, in these situations, there is honor in admitting
ignorance. Certain characteristics of many syndromic conditions, such as the so-called elfin facies of Williams
syndrome, are not apparent in the newborn period. The most useful diagnostic decision may be to wait and start
afresh at a later time, recollecting data, recombining features into new patterns, and researching the literature.
The assistance of a clinical geneticist or dysmorphologist during all phases of evaluation frequently optimizes the
diagnostic process. If such services are not immediately available via direct consultation, telephonic or
telemedicine consultation may be highly effective.
Other Management Issues

Communication with the parents of a child with congenital anomalies requires a compassionate, timely, and
honest presentation of the facts. One's choice of terminology is important but often challenging. Many parents
find it quite helpful for the principal caregiver to examine the infant in their presence, pointing out the features
that seem to be unusual, and delineating those that are normal. Parents naturally feel responsible for the birth
defect, which they might interpret as a reflection of their own shortcomings, real or imagined. Guilt is as
common for the parents of a malformed child as pride is for the parents of a normal newborn. Although it is not
always possible to convince parents to set aside unreasonable guilt, they can at least be reassured that they had
no control over the events causing the abnormality and that they have permission to not feel guilty.
Expect the full spectrum of grief from both parents. They have experienced the loss of a much-anticipated
normal child. Shock, denial, bargaining, and acceptance will all occur, even when their child does not have a
lethal condition. The physicians and nursing personnel involved in the care of the malformed infant work as a
team to keep track of this process and be alert for dysfunctional grief. Social work services, clergy, and support
groups are important adjuncts. The National Organization for Rare Disorders, the Alliance of Genetic Support
Groups, and specific support groups can provide up-to-date information and lay contacts for interested parents.
Not uncommonly, parents and relatives seek and find extensive information about birth defects, syndromes,
chromosomal anomalies, and teratogenic conditions on the internet. In some cases, this information is more
detailed and current than that available to a busy clinician who relies on brief textbook entries. Based on this
information, parents may become firm, even strident and argumentative, demanding specific tests and
management. This situation requires judgment, tact, and an open mind on the part of the attending physician,
because this information may be quite useful, and parents who remain your allies will positively influence the
entire spectrum of medical management. As correct and cutting edge as this information may be, parents will
appreciate your pointing out that the clinical acumen and experience of the neonatal team is essential to assess
the extent to which generic recommendations may be helpful for any particular patient. An imperative
component of this process will be to verify the reliability of the information and ascertain whether it is based on
credible scientific data.
If the malformed infant dies and there is any question regarding the precise diagnosis, a full, unrestricted
autopsy can prove extremely useful. Visceral, central nervous system, and skeletal anomalies frequently come
to light at postmortem examination. Clinical photography and cytogenetic analysis of fibroblasts obtained from
sterile skin biopsy, fascia, or pericardium also may yield important insights. Tissue, cells in culture, and
extracted DNA can be stored in a long-term repository and later reanalyzed in light of new research or
collaboration.
Families often are reluctant to grant permission for an autopsy. In many instances, however, this procedure has
profound implications for the parents' reproductive options and even those of distant relatives. Cultural and
social beliefs and practices must, of course, be carefully respected regarding the care of the child's body after
death. Nevertheless, an autopsy can be recast in the light of a final gift of the child to his family, perhaps even
to the world, if in fact a diagnosis is thereby established or medical science advanced.
P.899
SELECTED EXAMPLES
Teratogenic Conditions
A teratogen, from the Greek root teras, meaning monster or marvel, is any environmental factor that causes a
structural or functional abnormality in the developing fetus or embryo. These environmental agents include
infections, medications, drugs, chemicals, and maternal metabolites, such as phenylalanine (Table 38-3). By
their very nature, teratogens induce a disruption or sequence of disruptions of inherently normal tissue.
Extensive compendia of these agents have been studied in humans and laboratory animals, and several
excellent resources are available for the clinician. Additionally, both professionals and patients can access
regional teratogen hot lines.
Ethanol, the most common human teratogen, is estimated to affect as many as 1 in 300 newborns, primarily as
a neurotoxin, with consequences ranging from cerebral palsy to learning disability. Up to one-fifth of mental
retardation (usually mild) is attributable to fetal alcohol syndrome (FAS), but numerous structural anomalies
have also been reported (Table 38-4). Although the syndrome has been recognized for over 30 years, the
dysmorphisms and other clinical signs of FAS are still often overlooked in the newborn nursery (23).
Multifactoral Disorders
The multifactoral model of inheritance, as previously discussed, provides a conceptual basis for understanding
the pathogenesis and recurrence risks of isolated, nonsyndromic congenital malformations. The central concept
of this model is that multiple genes and environmental factors influence whether a particular anatomic structure
may develop abnormally. The susceptibility, or genetic liability, of a malformation in a population is described in
terms of
P.900
a continuous distribution of susceptibility factors in which there is a point, or threshold, beyond which, in an allor-none fashion, a structural defect will occur. Table 38-5 lists some common multifactoral disorders and their
recurrence risks. This type of multifactoral heritability is postulated to account for most isolated malformations.
TABLE 38-3 Selected teratogens and their effects
Teratogen
Phenytoin (Dilantin)
Warfarin (Coumadin)
Anomalieos
Comments
Growth deficiency
Similar facial features seen with
Wide anterior fontanelle

Hypertelorism
Cleft lip and palate
Hypoplastic distal phalanges
Small nails
Nasal hypoplasia
Stippled epiphyses
Short fingers
Seizures
exposure to carbamazepine, valproate,

mysoline, phenobarbital.
Full spectrum in 10%, milder effects in
one-third.
Critical period between 6 and 9 week's
gestation. One-third of exposed
fetuses are affected.
Retinoic acid (Accutane)
Microtia or anotia
Hypertelorism
Micrognathia
Conotruncal cardiac defects
Hydrocephalus
Microcephaly
Cortical, cerebellar dysplasia
Growth deficiency
Microcephaly
Deafness
Cataracts
Microphthalmia
Chorioretinitis
Rubella
If exposed more than 15 days after

conception, one-third have
embryopathy.
Fifty-percent chance of effects if

exposed in first trimester, but risk
extends into second trimester.
May have late, persistent infectious
sequelae, e.g., diabetes mellitus.
Cardiac septal defects

Peripheral pulmonic stenosis
Mental deficiency
Seizures
Cortical atrophy/microcephaly
Growth deficiency
Limb hypoplasia/club foot
Varicella
Maternal phenyketonuria
Cutaneous scars
Mental retardation (73% to 92%)
Hypertonia
Low birth weight (52%)
Microcephaly (73%)
Cardiac defects (15%)
Spontaneous abortion (30%)
One to two percent with effects when

exposed between 8 and 20 week's
gestation; wide spectrum of severity.
Even when on diet, phenylalanine

levels may rise above 4 to 10 mg/dL,
the apparent threshold for fetal effects.
Percentages cited here are for levels
>.16 mg/dL. Normal is <,2.
TABLE 38-4 Structural anomalies in fetal alcohol syndrome
Growth deficiency of prenatal onset

Mental retardation
Microcephaly
Short palpebral fissures
Maxillary hypoplasia
Cervical vertebral malformations (10%-20%)
Short nose
Smooth philtrum
Thin upper lip
Ventricular septal defect, atrial septal defect
Small distal phalanges
Small fifth fingernails
The neural tube defects (NTDs) result from failed neural tube folding before 28 days' gestational age and
comprise a spectrum of malformations ranging from anencephaly at one extreme to simple meningocele at the
other. A number of genetic and environmental (e.g., ethanol, valproic acid) causes are well delineated. Seventy
percent to 80% of NTDs have traditionally been considered to be isolated phenomena with recurrence risks of
the usual order of magnitude for multifactoral disorders: 3% to 5%. Over the past decade, although,
epidemiologic studies have demonstrated that seventy percent of isolated NTDs can be avoided by adequate
maternal intake of folic acid in the periconceptual period (24). Numerous public health and professional
organizations recommend that all women of childbearing age take 0.4 mg of folic acid per day. Women who
have previously borne an affected child should take folic acid at an increased dosage, 4.0 mg/d, beginning 1
month prior to conception and continuing through at least the third month of gestation (25).
TABLE 38-5 MULTIFACTORAL DISORDERS
Disorder
Empiric Recurrence Risk
Cleft lip with or without cleft palate

Cleft palate
Pyloric stenosis
Hirschsprung anomaly
Clubfoot
4%5%
2%6%
3%4%
3%
3%5%
2%8%
Congenital hip dysplasia

Neural tube defects
Atrial septal defect (secundum)
3%4%
3%5%
2%3%
Since 1956, when Tjio and Levan demonstrated that the normal diploid number of human chromosomes (46),
seemingly innumerable permutations of abnormal chromosome number and structure have been described.
Many of these are unique, but others are recurrent and produce readily recognized phenotypes. Most trisomies,
such as trisomy 16, the most common trisomy in humans, are uniformly lethal in the prenatal period. Other
rearrangements, such as most balanced translocations, have no phenotypic consequences unless a breakpoint
disrupts a gene. Aneuploidy that adds or deletes enough genetic material to be cytogenetically visible commonly
causes multiple congenital anomalies in several developmental fields.
Trisomy 21
Down Syndrome is the most common chromosomal aberration recognized at birth, with an incidence of about 1
per 700 live births. Only one of four conceptions that result in trisomy 21 is viable (27). The phenotype is quite
variable, but distinctive facies (Fig. 38-3) and some degree of mental retardation are always present. Gestalt
recognition of Down syndrome usually is uncomplicated, but diagnosis may be difficult if the infant is seriously ill
or has atypical findings. Conversely, between 13 and 18 percent of children referred to a clinical geneticist
because of suspected Down syndrome actually have another diagnosis (28).
An individual infant with Down syndrome will almost certainly lack one or several classic findings, such as a
wide gap between the first and second toes. Keep in mind that a single feature, such as the simian crease, falls
well short of being pathognomonic, but the overall constellation of major and minor anomalies suggests the
diagnosis. Tables 38-6 and 38-7 list the major and minor anomalies, respectively, that have been observed in
children with Down syndrome. Although major heart defects are present in about half of Down syndrome
infants, a heart murmur and other signs of cardiovascular pathology may be quite subtle in the immediate
newborn period. Atrioventricular canal can be particularly tricky to diagnose by auscultation. An echocardiogram
is recommended for all newborns with suspected or confirmed Down syndrome, ideally prior to discharge,
certainly by 1 month of age. If for any reason the cardiac ultrasound is delayed or unavailable before hospital
discharge, an electrocardiogram should be obtained. Leftward deviation of the QRS axis to between zero and
90 degrees is characteristic of an endocardial cushion defect.
For 90% of newborns with Down syndrome, nondisjunction during maternal meiosis results in the formation of a
gamete with two copies of chromosome 21 (27). The well-recognized relationship between trisomy 21 and
maternal age (Table 38-8) has yet to be well-explained. Although risk increases with age, most Down syndrome
infants are born to younger mothers because their birth rate is much greater than that of older mothers.
Translocation of
P.901
chromosome 21 to another acrocentric chromosome occurs in 1 of 30 Down syndrome neonates, and other
mechanisms of abnormal gametogenesis can also give rise to an extra copy of this chromosome. Accordingly,
these possibilities must be evaluated via karyotyping, to provide accurate genetic counseling. Any chromosomal
pattern other than trisomy 21 mandates analysis of parental karyotypes. For example, if an isochromosome 21
is present in the infant, parental chromosomal analysis may disclose that a parent carries the same
isochromosome. Recurrence risk for that parent is then 100%. Straightforward trisomy 21 in the infant is not
associated with chromosomal abnormalities of either parent and their testing is unnecessary. Recurrence risks
are then 1% plus the age-related maternal risk.
Figure 38-3 A Trisomy 21. B Down syndrome.
Improvement in routine health care and advances in cardiac surgery have improved both the quality and
longevity of life for individuals with Down syndrome. Median life expectancy is now 58.6 years (29). The
American Academy of Pediatrics has published helpful health supervision guidelines that address the needs of
Down syndrome children and their parents from birth to adulthood (30).
TABLE 38-6 Major anomalies at birth indown syndrome
Cardiac: all types

Atrioventricular canal
40%
16%20%
16%
3%5%
Atrial septal defect

Gastrointestinal: all types
4%10%
Duodenal stenosis/atresia
Imperforate anus
Other
Hematologic: leukemoid reaction
Hypothyroidism (congenital)
From refs. 17a-17c.
3%5%
2%
6%
10%18%
Common
1%
Trisomy 18
Described by Edwards and associates (31) in 1960, trisomy 18 affects 1 of 5,000 newborns. Girls outnumber
boys at a 4:1 ratio, and a maternal age effect is well established: de novo nondisjunction during meiosis
accounts for 90%. One-tenth represent mosaicism, and various translocations and isochromosomal anomalies
occasionally are seen. Life expectancy is markedly reduced, with nearly 90% mortality in the first year and
frequent demise in the neonatal period (32). The congenital anomalies in trisomy 18 neonates usually are
multiple, severe, and associated with considerable
P.902
morbidity. Severe psychomotor retardation always is present. Fig. 38-4 demonstrates typical facial features.
TABLE 38-7 Minor anomalies in Down syndrome
Microbrachycephaly
75%
Upslanting palpebral fissures
80%
Epicanthal folds
59%
Speckling of iris (Brushfield spots)
56%
90%
Flat facial profile

Low nasal bridge
68%
100%
Small ears
50%
Mildly dysplastic ears

Short neck
61%
Excess skin at nape of neck ,

Protruding tongue
Narrow palate
80%
Open mouth
Short hands and fingers
58%
a b
47%
76%
60%
Clinodactyly (curving) of the fifth finger
45%
Simian crease
Wide gap between the first and second toes
68%
85%
Poor Moro reflex
Hyperflexibility of joints
80%
80%
Hypotonia
a
a Among the ten cardinal features cited by Bryan Hall from ref. 7.
From ref. 28 as having superior discriminative efficacy and power.

TABLE 38-8 Incidence of Down syndrome as a function of maternal age
Maternal Age (yr)
Incidence
20
25
1:1667
1:1250
30
35
36
37
38
39
40
1:952
1:385
1:295
1:227
1:175
1:137
1:106
41
42
1:82
1:64
43
44
45
46
47
48
49
1:50
1:38
1:30
1:23
1:18
1:14
1:11
From ref. 27.

Considerable overlap with trisomy 13 syndrome may result in frequent diagnostic uncertainty in the newborn
period, pending cytogenetic analysis. Table 38-9 lists some features that are common to both conditions, and
Table 38-10 includes those findings that are more common to trisomy 18. Careful communication with the
cytogenetic laboratory will ensure that chromosomal data are available as soon as possible, a helpful adjunct for
decision-making when the clinical management options include expectant care and minimal intervention. In
selected instances, there may be considerable benefit from rapid analysis via FISH probes for chromosome 13
and 18 markers. Bear in mind, however, that, on occasion, these modalities may not detect trisomic states that
involve translocations or partial duplication and that full chromosomal analysis is superior and definitive. These
results usually are available 48 to 72 hours after the laboratory receives the specimen.
Trisomy 13
This is the third most common autosomal trisomy recognized at birth, affecting 1 of 12,000 newborns, and
initially was delineated in 1960 by Patau and associates (33). Three-fourths are straightforward trisomy 13; a
maternal age effect is apparent. Twenty percent are caused by translocations, mostly Robertsonian, in which the
long arm of the acrocentric chromosome 13 becomes attached via the centromere to another acrocentric
chromosome, commonly chromosome 14. A small percentage of these translocations are familial, so parental
karyotyping is
P.903
essential for adequate recurrence risk counseling. Mosaicism is seen in 5%. For both trisomy 13 and trisomy 18,
the median age of death is 10 days and 91% die within the first year. As with trisomy 18, cognitive and motor
development is profoundly affected. Table 38-11 lists some of the anomalies that tend to be encountered more
commonly in trisomy 13. Figure 38-5 shows typical facies.
Figure 38-4 A:, B: Trisomy 13.
TABLE 38-9 FINDINGS MORE COMMON TO TRISOMY 13
Growth deficiency
Severe developmental retardation
Microcephaly
Ear anomalies
Microphthalmia
Highly arched palate
Micrognathia
Excessive neck skin
Congenital heart defects, various
Ventricular septal defect, patent ductus arteriosus
Umbilical hernia
Renal anomalies
Cystic kidneys
Cryptorchidism
Overlapping, flexed fingers
Prominent heels
The Support Organization for Trisomy 18, 13, and Related Disorders (34) is a consumer-oriented, lay group that
serves as an excellent resource for parents, professionals, and other interested parties regarding trisomies 13
and 18 in particular.
Abnormalities of the Sex Chromosomes

With the exception of Turner syndrome, sex chromosome aneuploidies tend to be phenotypically bland in
newborn children. Klinefelter syndrome, 47, XXY, is associated with a few, occasional congenital anomalies, such
as cryptorchidism and hypospadias. Variants with more than one Y or more than two X chromosomes tend to
have greater mental deficiency and more malformations, both minor and major.
TABLE 38-10 FINDINGS MORE COMMON TO TRISOMY 18
Prominent occiput
Narrow palpebral fissures
Small mouth
Short sternum
Widely spaced nipples
Cardiac: polyvalvular disease
Prominent clitoris
Hypoplastic labia
Hip dislocation
Clubfoot
Hypoplastic nails
Syndactyly between toes 2 and 3
Hammertoes
Seizures
TABLE 38-11 ANOMALIES COMMON TO BOTH TRISOMY 13 AND TRISOMY 18
Scalp defects
Holoprosencephaly
Sloping forehead
Capillary hemangiomas
Ocular hypotelorism
Iris coloboma
Prominent nasal bridge
Cleft lip
Cleft palate
Short neck
Hypoplastic nipples
Cardiac: dextrocardia
Polydactyly (postaxial)
Apnea
Turner syndrome is caused by complete or partial absence of one X chromosome and occurs in 1 of 2,500
newborn girls. One half have a 45, X karyotype; a large
P.904
number of other X chromosome anomalies, ranging from various isochromosome X patterns to simple deletions,
ring chromosomes, and mosaics, account for the remainder. Almost all conceptuses with Turner syndrome are
spontaneously aborted but, of those that are live born, only one-third are recognized in the newborn period.
Clinical findings may include prominent ears, low posterior hairline, webbed neck, broad chest with widely
spaced nipples, and puffiness of the dorsa of the hands and feet (Fig. 38-6). Although short stature is common
in older girls, mean birth length is 47 cm, just within two standard deviations of the population mean. Ovarian
dysgenesis (greater than 90%), renal anomalies (40% to 60%, horseshoe kidney in particular), and cardiac
malformations (10%-20%, especially coarctation of the aorta) will require directed investigations once this
diagnosis is suspected (35). An AAP health supervision guideline (36) summarizes the age-specific evaluations
for girls with Turner syndrome. For neonates, recommendations include careful examination for hip dysplasia,
hearing screening, pediatric cardiology consultation, comparison of systolic blood pressures in the arms and
legs, renal ultrasonography, and consultation with pediatric endocrinology.
Figure 38-5 A, B Trisomy 13.
Partial Aneusomies
Deletions, duplications, and translocations of any segment, small or large, of any chromosome are possible. The
resulting partial monosomy, partial trisomy, or combination of the two often causes multiple major and minor
anomalies, intrauterine growth retardation, and a dysmorphic facial appearance. The deleted or duplicated
segment may be interstitial, involving the midsection of one of the arms of a chromosome, or terminal. As a
general rule, the larger the deletion or duplication, the more severe the somatic and functional effects. For
example, deletion of the short arm of chromosome 5 (sometimes designated 5p- or 5p minus) is one of the most
common autosomal deletion syndromes, with an incidence of 1 in 20,000 births. Also called cri-du-chat
syndrome because of the high-pitched cry of affected infants, 5p- results in a distinctive phenotype:
microcephaly; round facies; hypertelorism; broad nasal bridge, epicanthal folds; posteriorly rotated, malformed
ears; preauricular skin tags, and small chin (Fig. 38-7). Various other malformations may be present, including
cleft lip with or without cleft palate, heart defects, and megacolon. Hypotonia and mental retardation are typical.
The severity of cri-du-chat syndrome can be correlated roughly with the location and extent of the deletion:
absence of band 5p15.2 is associated with severe mental retardation; deletion of band 5p15.3 causes the typical
cat cry (37). Other common partial aneusomies are summarized in Table 38-12.
Figure 38-6 A, B Turner syndrome.
Many chromosomal rearrangements occur de novo as spontaneous events in a single egg or sperm. These are
unlikely to recur. Some, however, are the consequence of a balanced translocation or other rearrangement, such
as pericentric inversion, in one parent. Generally, balanced translocations do not cause problems because most
of the human genome consists of long stretches of DNA between genes. Breaks and rejoinings in these
noncoding regions will have neutral consequences. If the break falls within
P.905
a gene, some phenotypic effects may ensue. Careful analysis of balanced translocations has, in fact, provided
the critical link for mapping and cloning several important genes.
Figure 38-7 Cri-du-chat syndrome (5p-).
TABLE 38-12 Selected chromosomal anomaly syndromes
Karyotype
4p-
Dup 9p
Del 11p13
13q-
Features
Poor growth, microcephaly, prominent glabella, hypertelorism, downturned corners of

mouth, micrognathia, cleft lip/palate, cryptorchidism, heart defects, profound
developmental retardation. May involve only a small region, to include 4p16, requiring
high-resolution banding or FISH for confirmation. Also known as the Wolf-Hirschhorn
syndrome.
Poor growth, macrocephaly, hypertelorism, downslanting palpebral fissures, prominent
nose, downturned corners of mouth, cupped ears, short fingers and toes, developmental
retardation.
Mental deficiency, poor growth and microcephaly, aniridia, cataracts, blindness,
cryptorchidism, hypospadias, Wilms tumor (50%). The interstitial deletion encompasses a
tumor-suppressor gene and provides the de facto first hit in tumorigenesis.
Growth deficiency, mental deficiency, microcephaly, prominent nasal bridge,
hypertelorism, microphthalmia, colobomata, bilateral retinoblastoma, micrognathia,
anomalous auricles, hypoplastic thumbs, congenital heart disease, genitourinary
anomalies. Band q14 is the locus of the RB gene, a tumor-suppressor gene implicated in
18p-
18qDup 22q
retinoblastoma.
Mild growth deficiency and microcephaly, psychomotor retardation, ptosis, epicanthal
folds, hypertelorism, wide mouth, protruding ears, small hands and feet, pectus
excavatum.
Short-stature hypotonia, variable mental deficiency, conductive deafness, auricular
anomalies, long hands, tapering digits, cardiac defects.
Highly variable, also known as cat-eye syndrome, after the bilateral inferotemporal
colobomata of the irides, which also may involve the choroid and retina. Mild
developmental delays, normal growth, slight hypertelorism, downslanting palpebral
fissures, preauricular tags or pits, cardiac defects, anal atresia, renal agenesis. Fortyseven chromosomes are present: the extra chromosome is composed of one or two
chromosome 22 fragments, joined by their acrocentric short arms, creating trisomy or
tetrasomy 22 q. FISH provides confirmation of the nature of the small marker
chromosome. Parental karyotypes are necessary because a mildly affected parent also
may have the marker.
Not uncommonly, an unbalanced translocation in a newborn with multiple congenital anomalies is the
consequence of a balanced translocation in a parent. During formation of eggs or sperm at meiosis, balanced
rearrangements often yield gametes that have significant deletions, duplications, or more complex anomalies,
with serious consequences for the offspring. Unbalanced translocations, in which there is both a duplicated
chromosomal segment and a deletion, produce phenotypes that are unique, because they are a blending of a
partial monosomy and a partial trisomy. Chromosomal studies on both mother and father then are critical for
providing adequate recurrence risk counseling.
Microdeletion Syndromes
An increasing number of dysmorphologic syndromes are proving to be caused by fairly subtle interstitial or
terminal chromosomal deletions. These are sometimes detectable via chromosome studies, especially if highresolution or prometaphase analysis is employed, but more often the deleted region is submicroscopic and must
be specifically investigated via specialized cytogenetic and molecular techniques, most commonly FISH. Although
these microdeletions seem almost trivially small, they amount to a loss of several thousand of base pairs of
DNA, usually resulting in haploinsufficiency of several contiguous genes. In some instances, microdeletions may
be transmitted from parent to child as an autosomal dominant trait, but most occur de novo, with unaffected
parents, as a consequence of imprecise alignment of homologous chromosomes and unequal crossing over
during meiosis. Another technique, comparative genomic hybridization (38), can also delineate small deletions.
Table 38-13 lists several common microdeletion syndromes, their loci, and general features.
22q11.2 Deletion Syndrome has emerged over the last several years as the most commonly diagnosed
microdeletion syndrome. Also known as Shprintzen syndrome, velo-cardio-facial syndrome (VCFS), DiGeorge
syndrome (DGS), or conotruncal anomaly face syndrome, this condition may be manifest by congenital heart
defects, especially those affecting the conotruncal structures (tetralogy of Fallot, conal ventricular septal defect,
persistent truncus arteriosus, or interrupted aortic arch), and palate abnormalities, ranging from complete cleft
palate as a result of the Pierre Robin sequence to submucous clefting or velo-pharyngeal insufficiency.
Hypocalcemia as a result of parathyroid dysplasia, thymic hypoplasia with diminished T-cell production,
developmental delays and mental retardation, renal malformations, subnormal growth, and a characteristic
craniofacial appearance may be apparent in many affected individuals. Minor ear malformations, a bulbous nasal
tip, malar hypoplasia, and long facies are frequently encountered in children and adults; however, these typical
facial features may not be present in newborns, especially African-American infants (39). Consequently, the
concurrence of cardiovascular and palatal anomalies should prompt FISH studies for the 22q11 submicroscopic
deletion. Because other chromosomal defects can resemble
P.906
22q11 deletion syndrome, cytogenetic testing should routinely be submitted simultaneously. In about 15% of
individuals with the VCFS/ DGS phenotype, 22q11.2 FISH is normal; in some, atypical or other cytogenetic
deletions can be identified. Should testing be considered for infants with apparently isolated palatal clefting or
conotruncal anomalies? Many authorities advocate a low threshold for testing. Our appreciation of the incidence
of 22q11.2 deletion syndrome continues to grow, FISH analysis is sensitive and specific, and our considerable
insights into the natural history of this condition allow for practical and beneficial management (40).
TABLE 38-13 Microdeletion syndromes
ChromosomeLocus
Syndrome
1p36
Hypotonia, growth retardation, developmental

delays, deep-set eyes, pointed chin, seizures,
cardiomyopathy, deafness
4p16
Wolf-Hirschhom
5p16
Cri du chat
7p13
7q11.23
Grieg cephalopolysyndactyly
Williams
8q24
Langer-Giedion
11q13
WAGR
13q14.11
Retinoblastoma
15q12
Prader-Willi
15q12
Clinical Features
Angelman
Hypotonia, microcephaly, hypertelorism,

prominent glabella, severe mental deficiency
Low birth weight, cat-like cry, microcephaly,
downslanting palpebral fissures
Polydactyly, syndactyly, frontal bossing.
Supravalvular aortic stenosis, hyperacussis,
stellate irides, prominent lips, mild mental
retardation
Multiple exostoses, bulbous nose, protruding
ears, loose skin in infancy
Wilms tumor, Aniridia, Genital anomalies, growth
Retardation
Retinoblastoma, growth deficiency, microcephaly,
prominent nasal bridge, hypoplastic thumbs,
cardiac defects
Paternal deletion in 70%, maternal 15q disomy in
30%, hypotonia, hypogonadism, poor growth in
infancy, hyperphagia and obesity in childhood,
behavioral difficulties
Maternal deletion in 60%, paternal disomy in 5%,
severe mental retardation, seizures, ataxic gait,
microcephaly, paroxysmal laughter
16p13.3
16p13.3
ATR-16
Rubinstein-Taybi
17p13
Miller-Dieker
Alpha thalassemia, mild mental retardation

CBP gene deletion in 25%, growth deficiency,
mental retardation, small mouth, beaked nose,
broad thumbs and toes
Lissencephaly, high forehead with vertical
Smith-Magenis
furrowing while crying

Brachycephaly, broad nasal bridge, short
I7p11.2
philtrum, prominent jaw, broad hands
20p11.23
Alagile
22q11.2
DiGeorge, Shprintzen,
Deep-set eyes, broad forehead, chronic

cholestasis, vertebral defects, peripheral
pulmonary artery stenosis
Aortic arch anomalies, parathyroid and Velocardio-facial thymic deficiency leading to
hypocalcemia and diminished cellular immunity,
cleft palate, prominent nose with narrow alar
base, conotruncal cardiac defects, tapering fingers
Single Gene Disorders

Achondroplasia
The most common skeletal dysplasia recognized at birth, achondroplasia is an autosomal dominant condition
characterized by short limbs, frontal bossing, a large head, midface
P.907
hypoplasia, and distinctive radiographic findings. In the newborn period, a thoracolumbar gibbus, or sharp
kyphotic angulation of the vertebrae, may be present, followed in early childhood by lumbar lordosis. Midface
hypoplasia, a low nasal root, and relative prognathism (Fig. 38-8) are apparent at birth and become more
pronounced in childhood. The limb shortening is of the rhizomelic type, involving the humeri and femurs more
than the distal extremities. There may be limitation of elbow extension, yet other joints may be relatively
hyperextensible. The fingers often display a trident pattern, with a splay or gap between the third and fourth
digits. Mild hypotonia is common. The enlarged head circumference is associated with mild ventriculomegaly in
some cases, and in older children and adults a narrow foramen magnum has, on occasion, caused symptomatic
cervicomedullary compression. Cognitive development is normal.
Figure 38-8 Achondroplasia.
Achondroplasia is caused by specific mutations of the fibroblast growth factor receptor-3 gene (FGFR3) on
chromosome 4p16.3. Heterozygosity for a point mutation at nucleotide 1138, converting a glycine residue to
either a arginine or cysteine, is responsible for 99% of cases (41).
Smith-Lemli-Opitz Syndrome
For this inborn error of metabolism, a deficiency of 7-dehydrocholesterol reductase results in a distinctive
pattern of multiple congenital anomalies (42). Like most other enzyme deficiencies, this SLOS is inherited in an
autosomal recessive fashion. The reductase, which maps to chromosome 7q32.1, is responsible for the last step
in cholesterol synthesis. Consequently, serum cholesterol tends to be low, although in 10% it falls within normal
limits, and the immediate precursor, 7-dehydrocholesterol, is markedly elevated. A variable spectrum of
anomalies may be seen, including microcephaly, various central nervous system structural defects, hypotonia,
growth deficiency, and distinctive facies with a high, square forehead, ptosis, a short nose, anteverted nares,
and micrognathia (Fig. 38-9). Peripheral and central neural myelinization is reduced. About three-fourths of
genotypic XY males have genital anomaliescryptorchidism, ambiguous genitalia, even complete sex reversal.
Polydactyly and syndactyly between the second and third toes is very common. More severely affected
individuals may have visceral defects, such as renal cysts or agenesis, cardiac anomalies, pancreatic
hyperplasia, hepatic dysfunction, cataracts, severe growth retardation, postaxial polydactyly, and may be
stillborn or die in the neonatal period.
Figure 38-9 A, B Smith-Lemli-Opitz syndrome.
Chondrodysplasia Punctata, Rhizomelic Type

Rhizomelic chondrodysplasia punctata, an autosomal recessive inborn error of peroxisome metabolism as a
result of mutations of the peroxin 7 (PEX7) gene on chromosome 6q22-24, is manifest as a distinctive skeletal
dysplasia with mental retardation and poor long-term survival
P.908
(43). Birth length is reduced and the proximal limbs are shortened. Additionally, there is microcephaly, full
cheeks, a depressed nasal bridge, cataracts, ichthyosis, coronal clefts of the vertebral bodies, and congenital
heart disease (Fig. 38-10). In the neonatal period, punctate calcific stippling of the epiphyses is readily apparent
on lateral radiographs of the knee. Severe psychomotor retardation is the rule, and death usually occurs in the
first year of life. Serum phytanic acid is increased and plasmalogen biosynthesis is deficient. Ultrastructural
studies of hepatocytes demonstrate that peroxisomes are absent or markedly abnormal.
Figure 38-10 A, B Rhizomelic chondrodysplasia punctata.
Treacher Collins Syndrome

Also known as mandibul facial dysostosis, this autosomal dominant condition was first described in the mid1800s. Clinical features vary within and between families but typically include a narrow face with downslanting
palpebral fissures, zygomatic underdevelopment, a small chin, and an extension of scalp hair onto the cheek
(Fig. 38-11). The lower eyelids are notched at the outer third; medially the eyelashes are sparse or absent. The
ears often are crumpled and small. One-third lack the external auditory canal and/or have ossicular anomalies of
the middle ear that cause conductive hearing loss. Ear tags and preauricular fistulae are common. The mouth is
wide, and overt or occult palatal clefts are common. Intelligence is usually normal. In the early 1990s, several
investigators localized the gene to chromosome 5q32-33.2 using linkage analysis and FISH. In 1996, the gene,
designated TCOF1, was positionally cloned (44). The gene product, named treacle (British for molasses), is
truncated in most TCOF1 patients as a result of the introduction of a premature termination codon via various
deletions, insertions, and splicing alterations.
Apert Syndrome
This autosomal dominant craniosynostosis syndrome is readily recognized by the combination of an unusual
head shape and characteristic limb anomalies. Premature fusion of coronal sutures bilaterally leads to acrobrachycephalya short yet tall cranial shapewith a full forehead, flat occiput, flat midface, shallow orbits, and
downslanting palpebral fissures. There is a mitten-like syndactyly of the fingers and toes, both cutaneous and
osseous, and broad thumbs (Fig. 38-12). Other anomalies may involve the cardiac, gastrointestinal, central
nervous, and genitourinary systems. Most infants with Apert syndrome are born to unaffected parents: in a
single germ cell, a new mutation occurs in the fibroblast growth factor receptor-2 (FGFR2) gene on chromosome
10q25-26. Other FGFR2 mutations cause several other dominant craniosynostosis syndromes, such as Crouzon
syndrome and Pfeiffer syndrome (45).
Malformation Sequences
A malformation sequence represents the ultimate consequence of a precise cascade of fetal events. Initiated by
one primary event, which is often a mechanical or vascular disruption of a specific developmental field, a
sequence may stand alone as a recognizable entity or may be a component
P.909
of a larger picture, associated with a chromosomal defect, microdeletion, or single-gene disorder. The Pierre
Robin sequence, for example, is recognized as a cleft of the posterior palate, often in the shape of a U, in a
child with a markedly retruded, small mandible. At 5 to 9 weeks' gestation, the malpositioned jaw allows the
tongue to interfere with the medial apposition of the posterior palatal shelves as they migrate toward the
midline, thereby mechanically prohibiting their fusion. The newborn can quickly develop significant airway
obstruction and will require very close observation and occasional surgical intervention. The Robin sequence may
be isolated but, compared to other orofacial clefts, has a relatively high association with chromosomal anomalies
and syndromic diagnoses, such as trisomy 18, 22q11 deletion Beckwith-Wiedemann syndrome, Miller-Dieker
syndrome, fetal alcohol syndrome, and many others.
Figure 38-11 Treacher Collins-Franceschetti syndrome.
Figure 38-12 Apert syndrome.
Other Mechanisms
Uniparental disomy involves the inheritance of both copies of a particular chromosome or chromosome segment
from only one parent and has been observed in cystic fibrosis, Russell-Silver syndrome, Prader-Willi syndrome,
and Angelman syndrome, among others. Imprinting refers to how a gene may function differently, depending on
whether it is inherited from the mother or the father, and is often mediated via methylation of DNA sequences
that regulate gene expression.
Beckwith-Wiedemann syndrome (BWS) is a complex condition that may be caused by several mechanisms
affecting a multigenic region on the short arm of chromosome 11, including mutations of the CDKN1C gene,
abnormal methylation, paternal uniparental disomy, and chromosome 11p15 duplications. Diagnosis of BWS,
however, remains clinical: frequent features include somatic overgrowth, macroglossia, omphalocele,
visceromegaly, and dysplasia of the renal medulla. Transitory, symptomatic hypoglycemia is present in 30%,
and a number of neoplasms, including Wilms tumor, adrenal cortical carcinoma, and hepatoblastoma, are
common, especially in individuals with hemihypertrophy, which affects about 13% of patients. Glabellar nevus
flammeus, linear grooves of the ear lobes, and posterior helical ear pits are valuable diagnostic signs (Fig. 38-
13). About 85% of cases are sporadic, 15% familial. About 1% to 2% have either a small, interstitial duplication
of chromosome 11 involving band p15.5 or an inversion or translocation in this region; invariably the duplicated
region is paternal in origin. Other individuals derive both 11p15.5 regions from their father. They have the
normal number of chromosomesthat is, they are disomic for this regionbut have inherited both segments
from only one parent, i.e., uniparental disomy. These data also suggest that BWS is caused by a dosage effect
related to a gene or genes that are differentially expressed in maternally and paternally derived alleles, i.e.,
imprinting. Recently, two genes, KCNQ1OT1 and H19, have been found to have abnormal methylation in 2/3
cases of BWS, and direct DNA analysis of the CDKN1C gene, which is available on a research basis, can uncover
mutations many individuals. The recurrence risk depends on the nature of the genetic mutation. For instance, if
an 11p15 inversion is inherited from the mother, the recurrence risk is 50%, but paternal uniparental disomy
has a
P.910
very low chance of recurrence. Some genotype-phenotype associations have been delineated. Individuals with
cytogenetic abnormalities may have some degree of mental retardation; uniparental disomy and H19
hypermethylation are associated with an increased risk for Wilms tumor (46).
Figure 38-13 Beckwith-Wiedemann syndrome.
Disorders of Unknown Etiology

Cornelia de Lange syndrome is a sporadic condition of unknown etiology that can be recognized by the facial
gestalt. The eyebrows are crescent shaped, neatly defined, and often meet in the midline (synophrys); the
anterior hairline is low; the nose is short and the nostrils upturned; and the lips are thin with down turned
corners, like that of a carp (Fig. 38-14). Microcephaly and mental retardation are common but not invariable.
Growth is deficient prenatally and postnatally. About one-fourth have upper limb anomalies, which can range
from proximally placed thumbs to severe reduction defects. Empiric recurrence risk is estimated to be 2% to
5%. Duplication of the long arm of chromosome 3 causes a very similar malformation pattern (phenocopy) and
must be excluded.
GENETIC COUNSELING
Throughout human history, various astrologers, wise men, shamans, village elders, high priestesses, court
jesters, and mothers in law have paid close attention to birth defects. Curiosity has naturally inspired us to
theorize, to attempt to find meaning and significance, and then to communicate the truth about a congenital
anomaly to parents, relatives, communities, and rulers. As an omen or prophecy, the malformation of a newborn
has been inferred to be the visitation of a god, of celestial forces, of bad seed or bad morals, sure to portend
either a bright future or ruin to the bearer of bad fruit. From these basic impulses to understand and to predict
has grown the modern discipline of genetic counseling.
Figure 38-14 A, B: Cornelia de Lange syndrome.
The term counseling has the unfortunate connotation of giving authoritative advice, a fairly formal process in
which information and directive statements flow in one direction, from counselor to consultand. Genetic
counseling seeks to eschew these misconceptions. It is a process of communication between a genetic counselor
or counseling team and an individual, couple, or family about what exactly has caused a genetic problem, how
they might understand it, and what options they have for the futurein which they can find additional
educational resources, support groups, research studies, what modifications to the usual health maintenance
activities might be able to prevent potential problems, whether a similar problem could affect other family
members, how one could detect this prenatally or after birth, the risks and benefits of doing so, and the options
for prenatal management. Informally, many clinicians perform these activities as an integral component of
routine care giving. Formally, however, genetic
P.911
counseling is a complex and comprehensive activity that has blossomed as a discipline in its own right. It is
learned didactically and practically via postgraduate programs in medical genetics and genetic counseling, with
board certification (and recertification) incumbent on thoroughly documented case workups, close supervision by
experienced preceptors, and successful performance on rigorous, standardized examinations. There are 27
American Board of Genetic Counseling (ABGC) accredited programs in the United States and Canada; nearly
2500 individuals have received ABCG certification since 1993. Additionally, physicians who have completed
formal postgraduate training in clinical genetics, via fellowship or, in the past decade, through residency
training, and who obtain board certification through the ABMG are also qualified to provide genetic counseling.
The goals of genetic counseling have evolved over the past several decades. The provision of advice regarding
reproductive choice and behavior, with an eye toward reducing the amount of genetic disease in future
generations, has been superseded by other aims. The genetic counselor fundamentally seeks to provide
information with clarity, sensitivity, and support to enable the person seeking information, the consultand, to
understand well and decide capably. Emphasis is placed on the psychological and cultural aspects of this
interactive process, and support and follow-up are key elements. Genetic counseling requires that the diagnosis
be established with as much precision as possible, that an accurate and full pedigree be obtained, and that up-todate information regarding the diagnosis be researched. Recurrence risk calculations may be simple and
straightforward, or they may require more sophisticated tools, such as Bayesian analysis. Bayes' theorem,
created more than 200 years ago, allows the laws of probability to be applied to a specific clinical scenario and
quantifies the risk of recurrence by incorporating multiple observations into a complex formula (47). These data,
the natural history of the disorder in question, management options, and the full spectrum of reproductive
options are discussed in detail with the consultand, often with explicit attention to the significance these have for
that individual, the psychological and practical burden perceived in the context of the social structure, finances,
and personal experience. Principles of genetic counseling also include a commitment to strive for a nondirective
approach, truth telling, avoidance of paternalism, respect for autonomy and dignity, and anticipation of ongoing
psychological needs and issues. A concerted effort is made to identify and facilitate outside sources of
information and support, such as from clergy, genetic support groups, and social services.
Difficult Decisions
Some infants are born with anomalies that are irreparable and incompatible with life. For these newborns, the
decision to limit intervention, although deeply saddening, is clear. Bilateral renal agenesis and anencephaly are
salient examplesno amount of intervention will help; solace and support are the only options. Other newborns,
whose malformations are profound, but who may respond to heroic therapy, demand greater courage on the
part of parents and clinicianscourage to analyze carefully in the face of chaos, in the context of a health care
system whose bias is to act, simply because action is possible, and without much time in which to reach a
decision. A clinical genetics consultation that incorporates rapid dysmorphologic analysis and directed,
confirmatory testing will usually clarify the key issues of this decision making process.
The central importance of an accurate diagnosis is difficult to overstate. For a neonate with multiple major
anomalies, the neurologic prognosis cannot always be predicted solely on the basis of the severity of the obvious
malformations. For instance, in the case of VACTERL association, the tracheo-esophageal fistula and anal atresia
will require urgent, life-saving surgery, the vertebral abnormalities can seem grimly bizarre, and the overall
picture may appear dismal. However, these children typically have normal cognitive development, our surgical
and rehabilitation medicine colleagues have much to offer, and prognosis may be quite good. On the other hand,
some infants with trisomy 13 or trisomy 18 have remarkable few major anomaliesparents may even have
strong doubts about the diagnosisbut the neuro-developmental progress will be minimal.
The dignity of the child compels the caregiver to have as clear a vision of the future as possible, to deal with
decisions such as whether or not to intubate, transplant an organ, etc. The infant with thanatophoric dysplasia
may well live, with extraordinary care, to several years of age, as has been reported in a few children, and one
in ten infants with trisomy 18 lives to the first birthday, albeit with significant psychomotor retardation.
However, these facts are a starting point for an individualized analysis, shared between family and physician,
and not the foundation of policy to either routinely offer maximal intervention or insist that care be denied.
The parents of some profoundly disabled children perceive them as interactive and capable of both receiving and
giving affection. For these families, the rationale to limit medical support, because the long-term prognosis is
not good, is specious. They counter that they find richness and reward in their child, which they feel is
reciprocated, and that any judgment regarding quality of life is theirs, and theirs alone, to make. Of course,
the parents of a newborn with, for instance, trisomy 13 have had only the limited experience of the pregnancy
to develop an appreciation for their baby's human worth. If they have had prenatal genetic testing and
counseling, they may have been aware of this diagnosis for more than 5 months, have made a conscious
decision to not terminate the pregnancy, and have become well-read and quite sophisticated regarding options
for intervention and support. These parents will be seeking a neonatal and pediatric team with whom they can
be partners, not adversaries. If there is disagreement regarding prognosis or intervention, medical or surgical,
the clinician may do worse than fall back on basic principlescomplete ascertainment of facts,
P.912
diagnostic precision, and solid understanding of the natural history. One cannot underestimate the power of
careful and frequent communication with the family, with the dual aims of appreciating the family's points of
view and of presenting, with clarity, the medical point of view.
Recommended Resources
Suggested Internet Bookmarks
http://www.health.state.ny.us/nysdoh/dpprd/main.htm Evaluation of the newborn with single or multiple
congenital anomalies: a clinical guideline, American College of Medical Genetics Foundation and the New
York State Department of Health. See reference 9 below and text. Thorough familiarity with this resource
is highly recommended
http://www.kumc.edu/gec/geneinfo.html. Information for Genetic Professionals, University of Kansas

Medical Center. This is one of the original online clearinghouses of practical genetic information.
Thousands of internet links to other genetic web sites are available via this starting point. Particularly
useful is the section on support groups.
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=OMIM Online Mendelian Inheritance in Man (OMIM),

National Center for Biotechnology Information-JohnsHopkins University. Originally published in print by Dr.
Victor McKusick, this catalogue reigns as one of the most definitive and current sources of information for
over 14,000 mendelian disorders, with search capabilities and links to MEDLINE, graphics, and clinical
photographs, and access to DNA sequence data
http://www.geneclinics.org/ GeneTests, funded by the National Institutes of Health, the Health Resources
and Services Administration, and the Department of Energy. Originally a database of laboratories offering
esoteric DNA tests, GeneTests still specializes in connecting clinicians and over 500 molecular testing
laboratories. Additionally, this site incorporates GeneReviews, a compendium of definitive treatises,
written and updated by leading authorities, of about 300 genetic diseases for which diagnostic DNA testing
is available. Also at this site, GeneClinics can locate genetics professionals and services in the United
States
http://www.geneticalliance.org/ Genetic Alliance, The Alliance of Genetic Support Groups. For current
information on support groups and information that can be understood by patients and their families, this
site is the definitive resource. Additionally, the Alliance can provide expert information about the ethical,
social, and legal issues for genetic medicine in general and specific conditions. Families, researchers,
policymakers, health care professionals and industry are invited to participate as partners to make the
promise of genetics real.
http://www.aap.org/visit/gencomp.htm Genetics Compendium, The American Academy of Pediatrics, has

numerous links to information from the AAP, especially the Committee on Genetics, including policy
statements and health supervision guidelines for many conditions and scenarios.
For Your Library
Smith's Recognizable Patterns of Human Malformation, 5th Edition. First published in 1970, the 1997
edition, written by Kenneth Lyon Jones, continues to provide a classic tool for diagnosis, education, and
reference, with hundreds of clinical descriptions, detailed photographs, and definitive chapters on clinical
strategy and dysmorphology. W.B. Saunders Company
Syndromes of the Head and Neck, 4th Edition. Despite the title, this text goes far beyond the head and
neck. Robert J. Gorlin, D.D.S, M.S. D.Sc., M. Michael Cohen, Jr., D.M.D., Ph.D., and Raoul C.M.
Hennekam, M.D., Ph.D. Oxford University Press, 2001
Teratogenic Effects of Drugs, 2nd Edition. J. M. Friedman, M.D., Ph.D. and Janine E. Polifka, Ph.D. Johns
Hopkins University Press, 2000.
REFERENCES
1. Chung CS, Myrianthopoulos NC. Congenital anomalies: mortality and morbidity, burden and classification.
Am J Med Genet 1987; 27:505.
2. Van Regemorter N, Dodion J, Druart C. Congenital malformations in 10,000 consecutive births in a

university hospital: need for genetic counseling and prenatal diagnosis. J Pediatr 1984;104:386.
3. MacDorman MF, Minino AM, Strobino DM, et al. Annual summary of vital statistics-2001. Pediatrics
2002;110:1037.
4. Economic costs of birth defects and cerebral palsyUnited States, 1992. MMWR Morb Mortal Wkly Rep
1995;44:694-699.
5. Nelson K, Holmes LB. Malformations due to presumed spontaneous mutations in newborn infants. N Engl J
Med 1989;320:19.
6. Risch NJ. Genetic Epidemiology. In: Rimoin DL, Connor JM, Pyeritz RE, et al, eds. Principles and Practice of
Medical Genetics London: Churchill Livingstone, 2002: 457-458.
7. Hall BD. The state of the art of dysmorphology. Am J Dis Child 1993;147:1184.
8. Stevenson RE, Hall JG. Terminology. In: Stevenson RE, Hall JG, Goodman RM, eds. Human malformations
and related anomalies, vol 1. New York: Oxford University Press, 1993:21.
9. American College of Medical Genetics Foundation, sponsored by the New York State Department of Health.
Evaluation of the newborn with single or multiple congenital anomalies: a clinical guideline. May 1999.
Available at http://www.health.state.ny.us/nysdoh/dpprd/main.htm. Accessed 1/1/05.
10. Hall JG. An approach to malformation syndromes. In: Berg K, ed. Medical genetics: past, present, future.
New York: Alan R. Liss, 1985:275.
11. Cox GF, Burger J, Lip V, et al. Intracytoplasmic sperm injection may increase the risk of imprinting
defects. Am J Hum Genet 2002;71:162.
12. Moll AC, Imhof SM, Schouten-van Meeteren AY, et al. Retinoblas-toma is associated with in-vitro
fertilization. Lancet 2003;361:309.
13. DeBaun MR, Niemitz EL, Feinberg AP. Assisted Reproductive Technology may increase the risk of Beckwith
Wiedemann syndrome. Am J Hum Genet 2003;72:156.
14. Elias S, Simpson JL, Shulman LP. Techniques for prenatal diagnosis. In: Rimoin DL, Connor JM, Pyeritz RE,
et al, eds. Principles and Practice practice of Medical medical genetics. London: Churchill Livingstone,
2002:810.
15. Hall JG, Froster-Iskenius UG, Allanson JE. Handbook of normal physical measurements. Oxford: Oxford
University Press, 1989.
16. Saul RA, Seaver LH, Sweet KM, et al. Growth references: third trimester to adulthood. Greenwood, SC:
Keys Printing, 1998.
P.913
17. Wang RY, Earl DL, Ruder RO, et al. Syndromic ear anomalies and renal ultrasounds. Pediatrics 2001;108:
E32.
18. Opitz JM. Study of the malformed fetus and infant. Pediatr Rev 1981;3:57.
19. De Vries BBA, Winter R, Schinzel A, et al. Telomeres: a diagnosis at the end of the chromosomes. J Med
Genet 2003;40:385.
20. Burton BK. Inborn errors of metabolism in infancy: a guide to diagnosis. Pediatrics 1998;102. Available at
http://www.pediatrics.org/cgi/content/full/102/6/e69. Accessed 1/1/05.
21. Westphal V, Srikrishna G, Freeze HH. Congenital disorders of glycosylation: have you encountered them?
Genet Med 2000;2: 329.
22. Winter RM, Baraitser M. London dysmorphology database, vol 3.0. Oxford: Oxford University Press, 2001.
23. Thackray HM, Tifft C. Fetal alcohol syndrome. Pediatr Rev 2001; 22:47.
24. Hall JG, Solehdin F. Genetics of neural tube defects. Ment Retard Dev Disabil Res Rev 1998;4:269.
25. American Academy of Pediatrics, Committee on Genetics. Folic acid for the prevention of neural tube
defects. Pediatrics 1999; 104:325.
26. Tjio JH, Levan A. The chromosome number of man. Hereditas 1956;42:1.
27. Hook EB. Chromosome abnormalities: prevalence, risks, and recurrence. In: Brock DH, Rodeck CH,
Ferguson-Smith MA, eds. Prenatal diagnosis and screening. Edinburgh: Churchill Livingstone, 1992:351.
28. Rex AP, Preus M. A diagnostic index for Down syndrome. J Pediatr 1982;100:903.
29. Glasson EJ, Sullivan SG, Hussain R, et al. The changing survival profile of people with Down's syndrome:
implications for genetic counselling. Clin Genet 2002;62:390.
30. American Academy of Pediatrics, Committee on Genetics. Health supervision for children with Down
syndrome. Pediatrics 2001: 107:442.
31. Edwards JH, Harnden DG, Cameron AH. A new trisomic syndrome. Lancet 1960;1:787.
32. Rasmussen SA, Wong L-Y C, Yang Q, et al. Population-based analyses of mortality in trisomy 13 and
trisomy 18. Pediatrics 2003;111:777.
33. Patau K, Smith DW, Therman E, et al. Multiple congenital anomalies caused by an extra autosome. Lancet
1960;1:790.
34. S.O.F.T., 2982S Union St., Rochester, NY 14624; (716) 594-4621. Available at http://www.trisomy.org.
Accessed 1/1/05.
35. Robinson A, de la Chapelle A. Sex chromosome abnormalities. In: Emery AEH, Rimoin DL, eds. Principles
and practice of medical genetics, 2nd ed, vol 1. Edinburgh: Churchill Livingstone, 1990:973.
36. Frias JL, Davenport ML, AAP Committee on Genetics, AAP Section on Endocrinology. Health supervision for
children with Turner syndrome. Pediatrics 2003;111:692.
37. Cerruti Mainardi P, Perfumo C, Cali A, et al. Clinical and molecular characterization of 80 patients with 5p
deletion: genotype-phenotype correlation. J Med Genet 2001;38:151.
38. Levy B, Dunn TM, Kaffe S, et al. Clinical applications of comparative genomic hybridization. Genet Med
1998;1:4.
39. McDonald-McGinn DM, Driscoll DA, Emanuel BS, et al. The 22q11 deletion in African American patients: an
underdiagnosed population. Am J Hum Genet 1996;59:90(A).
40. McDonald-McGinn DM, Emanuel BS, Zackai EH. 22q11 deletion syndrome. GeneReviews 1999;(Sept 23).
Available at http://www.geneclinics.org. Accessed 1/1/05.
41. Francomano CA. Achondroplasia. Gene Reviews 2001;(Mar 8). Available at http://www.geneclinics.org.
Accessed 1/1/05.
42. Kelly RI, Hennekam RCM. The Smith-Lemli-Opitz syndrome. J Med Genet 2000;37:321.
43. White AL, Modaff P, Holland-Morris F, et al. Natural history of rhizomelic chondrodysplasia punctata. Am J
Med Genet 2003; 118A: 332.
44. The Treacher Collins Syndrome Collaborative Group. Positional cloning of a gene involved in the
pathogenesis of Treacher Collins syndrome. Nat Genet 1996;12:130.
45. Robin NH, Falk MJ. Craniosynostosis syndromes (FGFR-related). GeneReviews 2003;(Feb 13). Available at
http://www.geneclinics.org. Accessed 1/1/05.
46. Shuman C, Weksberg R. Beckwith-Wiedemann syndrome. GeneReviews 2003;(Apr 10). Available at http://
www.geneclinics.org. Accessed 1/1/05.
47. Young ID. Risk estimation in genetic counseling. In: Rimoin DL, Connor JM, Pyeritz RE, eds. Principles and
practice of medical genetics, 3rd ed, vol 1. Edinburgh: Churchill Livingstone, 1996:521.
48. Hall JG. When a child is born with congenital abnormalities. Contemp Pediatr 1988; August:78.
Chapter 39
Endocrine Disorders of the Newborn
Mary M. Lee
Thomas Moshang Jr.
From the moment of conception, physiologic endocrine processes are
actively involved in growth and development of the human fetus.
Disturbances in the interplay of these complex hormonal processes
during intrauterine life can cause somatic or biochemical alterations in
the fetus and newborn infant. Clinical disorders of endocrine function in
the neonate therefore, can reflect an altered physiologic state in the
fetus, the mother, or the fetal-maternal unit. Moreover, the occurrence
of these perturbations of endocrine function at varying stages of fetal
development results in diverse clinical manifestations. Knowledge of the
ontogeny of the endocrine glands and their physiologic function during
fetal development facilitates understanding disorders of endocrine
function in the newborn.
DISORDERS OF SEXUAL DIFFERENTIATION

Normal Sexual Differentiation
The normal regulation of sexual differentiation is broadly illustrated in
Fig. 39-1. All embryos are initially undifferentiated with a bipotential
gonad and the anlagen for both male and female reproductive tracts and
genitalia (1). Differentiation of the gonads as testes or ovaries dictates
the subsequent development of the internal and external genitalia. The
gonad forms when germ cells migrate from the dorsal endoderm of the
yolk sac to populate the genital ridges. At the fifth to sixth week of
gestation, these primitive bipotential gonads consist of both cortical
(ovarian) and medullary (testicular) components. The genital ridge is
composed of three cell types: (a) germ cells destined to become
prespermatogonia in the male or oogonia in the female, (b) supporting
epithelial cells destined to become Sertoli cells (male) or granulosa cells
(female) and (c) mesenchymal cells destined to become the steroidproducing Leydig cells (male) or theca cells (female). Development of the
supporting cells as Sertoli or follicular are a critical determinant in

whether the germ cells differentiate as spermatogonia or oogonia.
Of the genes that are critical for gonadal development, mutations in two
have been clearly associated with gonadal dysgenesis. Mutations of the
Wilms tumor gene (2) (WT1), are associated with three related
syndromes (the WAGR contiguous gene syndrome, and Denys-Drash and
Frasier) that affect renal function and gonadal development and
mutations in the transcription factor, steroidogenic factor 1 (3) (SF-1),
cause agenesis of the adrenals and gonads.
Sexually dimorphic differentiation of the gonads and reproductive system
commences when the testis-determining gene is first expressed. In 1959,
Ford and colleagues determined that the Y chromosome was necessary
for male development (4); in 1966, the critical region for testis
determination was localized to the short arm of the Y chromosome (5);
and in 1990, the primary testis-determining gene was definitively
identified at Yp11.3 by positional cloning in patients with sex reversal
(6). This gene, termed SRY (sex-determining region of the Y
chromosome), is a member of the Sox family of transcription factors that
all contain a high mobility group (HMG) DNA-binding motif (6). Activation
of SRY initiates differentiation of the bipotential gonad as a testis. Loss
of function mutations in SRY or a delay in its onset of expression can
cause XY sex reversal with gonadal dysgenesis, although a gain of
function mutation in SRY causes XX sex reversal associated with
incomplete testicular development.
SOX9, a presumptive target for SRY, is a related HMG box gene that
induces the supporting cells of the gonadal ridge to differentiate as
Sertoli cells (7). Inactivating mutations of SOX9 cause the skeletal
anomalies of camptomelic dysplasia that is associated with variable
penetrance of 46, XY sex reversal and testicular dysgenesis (8). A role
for additional downstream genes in testicular determination is supported
by the absence of identifiable SRY or SOX9 mutations in
P.915
some cases of sex reversal and the association of sex reversal and
dysgenetic gonads with other autosomal deletions.
Figure 39-1 Schematic of the sex differentiation pathway. The

urogenital ridge and gonad are initially undifferentiated. In male
embryos, the induction of SRY expression initiates testis determination.
The testicular hormones, MIS/AMH and androgens stimulate male
phenotypic development of the internal and external genitalia. In female
embryos, the absence of SRY in concert with the expression of genes
that inhibit testicular determination enables the gonad to develop as an
ovary. In the absence of androgens and MIS/AMH, the internal and
external genitalia differentiate as female.
At least two genes, Wnt-4, a locally secreted signaling glycoprotein, and

Dax-1, a nuclear hormone receptor in the DSS region of the Xchromosome, are critical for ovarian development, perhaps by
suppressing testicular genes. A duplication of either of these genes
interferes with normal testicular development to cause a dosage
sensitive form of 46, XY sex reversal (3,9).
Sexually dimorphic differentiation of the wolffian (male anlagen) and

mullerian (female anlagen) internal genital tracts depends on the
hormonal milieu established by the somatic cells. If SRY is expressed,
the primary sex cords develop into testes and the somatic cells
differentiate as Sertoli and Leydig cells. The Sertoli cells secrete
Mullerian inhibiting substance (MIS), a 140-kd glycoprotein in the TGF-
family of proteins. MIS causes degeneration of the mullerian ducts by
inducing apoptotic cell death of the ductal epithelial cells.
The Leydig cells secrete testosterone, which stimulates the wolffian ducts
to differentiate into the vas deferens, seminal vesicle, and epididymis
and virilizes the external genitalia.
Differentiation of the external genitalia requires the activation of
testosterone by 5-reductase-2 to its more active metabolite,
dihydrotestosterone (DHT). DHT stimulates fusion of the urethral folds
and the labioscrotal swellings to form the corpus spongiosa and scrotum.
DHT also stimulates growth of the genital tubercle and prostate. Sexually
dimorphic differentiation of the internal ducts and the external genitalia
is complete by 12 weeks of gestation. Subsequently, during the latter
part of gestation, the testes descend into the scrotum and the phallus
enlarges as testosterone production increases under the stimulus of
pituitary gonadotropins.
In 46, XX embryos, the primary sex cords become follicles by 10 weeks
gestation and the germ cells differentiate to oogonia. Both X
chromosomes are needed for maintenance of the ovary. In the absence
of one X chromosome,
P.916
as in Turner syndrome, the ovaries form initially, but regress before
birth. In females, fetal ovaries do not secrete MIS, thus the mullerian
ducts differentiate to form the uterus, fallopian tubes and upper vagina.
The absence of testosterone and DHT production by the fetal ovary cause
the wolffian ducts to degenerate and enables the external genitalia to
differentiate as female.
DISORDERS OF CHROMOSOMAL SEX

A number of sex chromosome aberrations have been reported (see
Chapter 38): some are embryonic lethal (e.g., 45, YO), others cause
minimal somatic or hormonal manifestations in the newborn (e.g., 47,
XXY), and a few perturb gonadal and genital development (45, X/46, XY,
45, X0). In contrast to the autosomes, extra genetic material from the X
chromosome can be tolerated with minor untoward effects as a result of
inactivation of the second and additional X chromosomes. Although
ovarian formation and function are intact in patients with X polyploidy,
early menopause may occur (10). In contrast, a Y chromosome is
generally necessary for testicular development, although rare cases of
sex-reversed 46, XX males with normal testicular function and male
genitalia are reported (11).
The classic sex chromosomal anomalies occur relatively frequently as
determined by newborn screening. In the New Haven Study, 47, XXY
occurred once in 545 males, 47, XYY occurred once in 728 males, 47,
XXX occurred once in 727 females, and 45, XO occurred once in 2181
female newborns (12). The incidence of 45, XO if higher than this, but
leads to increased fetal demise and is found in 1 in 15 spontaneous
abortions (13). The diagnosis of Turner syndrome, however, is made with
greater frequency than the other sex chromosomal aberrations because
of the associated somatic abnormalities.
Turner Syndrome
The classic and most common chromosomal abnormality is total loss of
one X chromosome. Over 50% of girls with Turner syndrome have a 45,
XO karyotype, 17% have mosaicism with an isochromosome 46, X, i(Xq),
8% are chimeras with 45, X0/46, XX, and the remainder have other
forms of mosaicism with loss of X material (14). The presence of a
mosaic 46, XX cell line has little bearing on stature or somatic
abnormalities, but does influence gonadal development. Goldberg and
associates reported spontaneous female sexual development in 3 of 25
patients with mosaic karyotypes, but in none in those with 45, X0 (15).
The Turner phenotype in the newborn is secondary to lymphangiectasia
and lymphedema. The webbed neck is most often seen as redundant
folds about the posterior neck. The lymphedema involves the dorsa of
the hands and feet. A host of associated somatic defects have been
described in this syndrome (14,16), most of which become more evident
with increasing age. The most common are triangular facies with low-set
ears, high-arched palate, low hairline, shield-like chest with widespread
and hypoplastic areolae, and cubitus valgus. Coarctation of the aorta is
the typical cardiovascular abnormality; however the more benign
condition of bicuspid aortic valves occurs more frequently. Miscellaneous
renal malformations are observed. Skin manifestations include
hemangiomas, cutis laxa, pigmented nevi, dysplastic nails, and tendency
to keloid formation. Skeletal abnormalities include beaking of the
medial tibial condyle, drumstick-shaped distal phalanges, vertebral
anomalies, and short metacarpals (17). Dermatoglyphic abnormalities
include palmar simian creases, distal axial triradius, and an increased
number of digital ulnar whorls. Declining growth can manifest in young
children and is the most consistent characteristic in the older child.
After diagnosis, screening for associated disorders such as cardiac and
renal defects is needed (16). Therapy is focused on the specific
developmental anomaly, such as coarctation of the aorta, and on
providing education about potential associated problems, such as
recurrent otitis media, chronic lymphocytic thyroiditis, and idiopathic
hypertension. The incidence of mental retardation is slightly increased
with specific X chromosome rearrangements. In most children with
Turner syndrome however, cognition is normal with good verbal skills
and selected spatial deficits.
A major concern for girls with Turner syndrome is extreme short stature
with a mean adult height of 148 cm. Recombinant growth hormone
increases final height and is approved for treatment of short stature in
Turner syndrome. The combination of early use of growth hormone
(before 5 years of age) and low dose estrogen replacement at an
appropriate age is thought to give the best outcome in terms of height
and pyschosexual development (18). Estrogen/progesterone therapy at
the appropriate age is indicated for the treatment of sexual infantilism.
Questions regarding fertility may arise even in the newborn period
because primary gonadal failure occurs in more than 90% of individuals
with Turner syndrome. Women with Turner syndrome, however, have

successfully carried offspring to term using donor oocytes, at a rate
similar to couples with other causes of infertility (19).
The presence of Y material in the karyotype raises concerns about
testicular elements that are at risk for malignant transformation. In
patients with mosaicism that includes Y material, gonadectomy is
therefore recommended, both to eliminate the risk for gonadoblastoma
and to avoid the virilizing effects of hormonally active residual testicular
elements (20).
DISORDERS OF GONADAL DETERMINATION

Disorders of gonadal determination can occur in association with
autosomal or sex chromosomal anomalies and/or loss of function
mutations or deletions of SRY and
P.917
SOX-97. Mutations have also been identified in other genes essential for
gonadal formation such as WT-1 and SF-1. The clinical phenotype of
these single gene mutations varies from complete gonadal dysgenesis to
lesser degrees of testicular damage. Teratogens such as radiation,
viruses, and drugs have also been implicated in in utero gonadal
damage. Differentiation and development of the internal ducts and
external genitalia in these infants depend on the timing and extent of
insult to the developing gonad.
Complete Gonadal Dysgenesis

Complete dysgenesis of the genital ridges results in normal female
genitalia with no associated somatic findings, thus the diagnosis is often
not suspected during infancy. Infants with 46, XY complete gonadal
dysgenesis may be identified as a result of genitalia that are discordant
with the prenatal karyotype. Affected girls tend to be tall with eunuchoid
proportions and often present with primary amenorrhea and sexual
infantilism. An autosomal recessive form of 46, XX gonadal dysgenesis is
associated with sensory neural deafness (21). The majority of 46, XY
gonadal dysgenesis is sporadic but familial forms can be sex-limited
autosomal recessive, X-linked, or autosomal dominant in inheritance (22)
Partial Gonadal Dysgenesis

Incomplete loss of function of genes essential for testicular
differentiation, or exposure to teratogens that damage the developing
testis cause partial gonadal dysgenesis. Testicular loss after 9-10 weeks
of gestation will not interfere with involution of the mullerian structures
because the critical time for exposure to MIS is at 7 to 10 weeks, but will
perturb mid-line fusion and development of the external genitalia, which
is dependent on on-going testosterone production by the testes. Thus the
external genitalia are female or severely undervirilized, but the gonads,
uterus and fallopian tubes are absent and the wolffian structures are
incompletely developed.
True Hermaphroditism
In true hermaphroditism, both ovarian and testicular elements are
present. Findings may consist of an ovary on one side and a testis on the
contralateral side, an ovary or a testis and a contralateral ovotestis, or
two ovotestes (23). Most patients with true hermaphroditism have
ambiguous external genitalia, although differentiation of the internal
duct structures and external genitalia depend on the amount of
functioning testicular tissue. In those reared as female, the testicular
component of the gonad may secrete androgens at puberty to cause
unwanted virilization, thus gonadectomy should be performed early.
Although some patients have sex chromosome abnormalities, 46, XX is
the most common karyotype, followed by 46, XY. The pathogenesis of
true hermaphroditism is not well understood, but is not consistently
linked to alterations in SRY expression.
DISORDERS OF PHENOTYPIC SEX

Disorders of phenotypic sex result when the anatomic development of the
external genitalia does not correspond to the chromosomal and gonadal
sex. The external genitalia may be truly ambiguousthat is, the sex of
the infant cannot be ascertained by physical examination. Alternatively,
the phenotype may be normal male or female, but inappropriate for the
genotype. These conditions may be secondary to genetic defects
affecting hormonal synthesis or action, problems in timing or regulation
or hormonal secretion, defects in receptor binding or signaling defects,

or teratogens or maternal hormones that perturb reproductive tract
development. A genotypic (46, XY) male with testes and inadequate
virilization falls in the category of male pseudohermaphroditism. A
virilized genotypic (46, XX) female with ovaries is considered to have
female pseudohermaphroditism.
Female Pseudohermaphroditism
The female fetus can be virilized by fetal adrenal androgens or maternal
androgens transferred across the placenta. Exposure to androgens prior
to week 12 of gestation results in fusion of the urogenital sinus and
genital folds. Exposure to androgens after week 12 of gestation or after
birth causes milder manifestations of clitoral enlargement, labial
hyperpigmentation, and posterior labial fusion.

The major cause of virilization in a female is congenital adrenal
hyperplasia (CAH) as a result of steroidogenic enzyme defects in cortisol
biosynthesis. The more common inherited enzymatic deficiencies of
adrenal biosynthesis (21-hydroxylase, 11-hydroxylase, and 3hydroxysteroid dehydrogenase defects) all virilize the female. The
gonads are ovaries, thus internal genital duct development is female
the wolffian ducts regress and the mullerian structures are normally
formed. The excess adrenal androgens, however, promote fusion of the
labia and the urogenital sinus, and phallic enlargement. Rarely, the
virilization is severe enough to cause complete external masculinization.
The 21-hydroxylase and 3-hydroxysteroid dehydrogenase forms of CAH
are complicated by associated mineralocorticoid deficiency that may
present as a salt-losing adrenal crisis. The enzymatic defects causing
CAH are more fully discussed in the section on adrenal disorders.
Drug-Induced Female Pseudohermaphroditism

A number of female newborns have been virilized by progestational
agents used to prevent spontaneous abortion or other medications with
androgenic potency (24). When these drugs are used during the first
trimester, the labioscrotal folds are fused with formation of a urogenital
P.918
sinus and clitoromegaly. When fetuses are exposed after the first
trimester, they develop only clitoral enlargement, without labioscrotal
fusion. In contrast to untreated infants with CAH, there is neither
progressive virilization nor continued acceleration of growth and skeletal
maturation after birth. No medical intervention is needed as androgens
are not elevated but surgical correction might be warranted. These
children will feminize normally at puberty and achieve normal fertility.
Virilizing Disorders in the Mother

Virilization of a female fetus by an androgen-producing maternal tumor is
rare. These tumors almost always are caused by an ovarian lesionarrhenoblastomas, Krukenberg tumors, luteomas, lipoid or stromal cell
tumors, although adrenal adenomas are also reported (25). The tumors
cause clitoromegaly, acne, deepening of the voice, decreased lactation
and hirsutism in the mothers and are associated with elevated serum
androgens and elevated excretion of urinary 17-ketosteroids (26).
Aromatase Deficiency
Rare genetic defects in the fetal or placental aromatase gene impair
aromatization of maternal and placental androgens to estrogens and
cause in utero elevations of androgens (27). Both fetal and maternal
virilization can occur.
Idiopathic Female Pseudohermaphroditism

Idiopathic virilization may be caused by nonhormonal factors as exposure
to androgens cannot be documented. This can occur in isolation or in
conjunction with congenital anomalies of the gastrointestinal (GI) and
urinary tracts that include imperforate anus, renal agenesis, urinary tract
obstructions, urethrovaginal fistulas, and/or defective mullerian duct
formation.
Male Pseudohermaphroditism
Incomplete masculinization of the male fetus can be as a result of a
myriad of causes that disrupt either androgen action or the response of
target tissues to androgens during sexual differentiation. The differential
diagnosis of male pseudohermaphroditism is extensive, including
enzymatic defects of testosterone synthesis, unresponsiveness to testosterone action (androgen-resistance syndromes), hypothalamic or
pituitary dysfunction, and vascular or teratogenic insult to the testis.
Defects in Testosterone Biosynthesis

Male fetuses are undervirilized when in utero testosterone production is
reduced as a result of a genetic mutation in one of the testosterone
synthetic enzymes (28). The phenotype varies according to the specific
gene affected. Defects in 17-hydroxylase/lyase and 17-hydroxysteroid
dehydrogenase primarily affect testosterone production although defects
in enzymes needed for both testosterone and cortisol synthesis cause a
form of CAH associated with incomplete masculinization. Steroidogenic
acute regulatory protein (StAR) and 3-hydroxysteroid dehydrogenase
mediate early steps in cortisol and aldosterone synthesis. Mutations in
either of these genes can result in a severe salt-losing syndrome and
insufficient masculinization of the genitalia. The full details of these
disorders, including diagnosis and treatment, are outlined in the section
on adrenal hyperplasia.
Syndromes of Androgen Resistance

This is a group of disorders characterized by undervirilized genitalia with
normal mullerian duct regression and normal testosterone synthesis
(29). The term androgen resistance encompasses androgen receptor or
postreceptor defects (androgen insensitivity syndrome [AIS]) and 5reductase deficiency in which the conversion of testosterone to its more
active metabolite, DHT is affected. In both conditions, MIS is produced
normally by the fetal testis and causes involution of the mullerian
structures. In AIS, although testosterone is produced, the defect resides
in the receptor or its signaling, thus target tissue response is

compromised. Consequently, all aspects of male development mediated
by androgens, including development of the wolffian structures and
external genitalia are affected.
In contrast, in 5-reductase deficiency, sufficient testosterone is
produced for differentiation of the wolffian structures, but not for
complete virilization of the external genitalia, which requires DHT (30).
DHT is needed for mid-line fusion and phallic growth, thus patients with
5-reductase deficiency typically have a blind vaginal pouch, a small
phallic structure with chordee, a hooded prepuce, and perineoscrotal
hypospadias. At puberty, the increase in secretion of testosterone and
induction in expression of 5-reductase and the androgen receptor in
genital tissues stimulate growth of pubic hair, penile enlargement, and
descent of the testes. 5-reductase deficiency is suspected in 46, XY
patients with perineoscrotal hypospadias and an elevated testosterone to
DHT ratio of more than 35 under basal conditions and more than 74 after
human chorionic gonadotropin (hCG) stimulation. The diagnosis is
confirmed by finding reduced 5-reductase activity in fibroblasts from
genital skin.
Complete AIS (CAIS), an X-linked condition, was previously referred to
as the testicular feminization syndrome (31). Affected patients have a
46, XY karyotype, normal female external genitalia with a blind vaginal
pouch, absent wolffian and mullerian structures, and abdominal or
inguinal testes. Unlike patients with 5-reductase deficiency, virilization
does not occur at puberty but peripheral conversion of the high
testosterone concentrations to estradiol stimulates good breast
development and estrogenization of the vaginal mucosa. Most patients
have little pubic hair and some have total absence of sexual hair. In
P.919
all other respects, including height, habitus, voice, breast development,
and gender identity, these individuals are feminine. The diagnosis is
made in infancy or childhood as a result of female genitals that are
discrepant with an 46, XY karyotype, or to the discovery of testicular
tissue during hernia repair. Adolescent patients frequently present with
primary amenorrhea. Genetic mutations of the androgen receptor are

identified in only two thirds of individuals with suspected AIS. The
gonads in CAIS have a 9% risk for malignant transformation to
neoplasia, thus should be removed surgically.
A wide spectrum of phenotypes is observed in individuals with incomplete
forms of androgen insensitivity. Partial or mild AIS (PAIS) can range
from a female phenotype with clitoromegaly and posterior labial fusion to
a male phenotype with infertility and sparse pubic hair. Sex assignment
may be difficult in patients with PAIS. In some cases, assessment of
responsiveness of the phallus to androgens is helpful.
OTHER CONDITIONS INVOLVING

GENITOURINARY DEVELOPMENT
Male or Female Genital Phenotype
Inconsistent with Genotype
A problem ensues at birth when the phenotype is inconsistent with the
chromosomal sex determined by prenatal cytogenetic analysis. A repeat
karyotype after birth may reveal an error in the prenatal cytogenetic
analysis. A number of intersex conditions, however, can present at birth
with external genitalia that are normal in appearance but discordant with
chromosomal sex (Table 39-1). In general, these children should be
raised according to the phenotypic sex.
Hypospadias and Cryptorchidism

Isolated hypospadias occurs in 0.8% of newborn infants and isolated
cryptorchidism is present in approximately 5% of full-term and up to
15% of premature infants. Generally, neither condition by itself is
associated with an endocrine abnormality. The incidence of intersex
conditions however, is greater if the hypospadias is severe (on the shaft
or perineum), or if the testes are nonpalpable. If cryptorchidism and
hypospadias are both present, 25% of infants have a disorder of
intersex.
TABLE 39-1 ETIOLOGY OF MALE OR FEMALE GENITAL PHENOTYPE

INCONSISTENT WITH THE GENOTYPE
Disorder
Pure gonadal
Genotype
XY
Phenotype
Female
dysgenesis
Etiology
mutations in SRY,
SOX9 WT1, SF-1
46, XX Males
XX
Male
SRY translocation
46, XY Females
XY
Female
SRY deletion
Congenital lipoid
XY
Female
CAH (StAR)
XY
Female
CAH (17-HSD)
XY
Female
CAH (p450c17)
XY
Female
mutation in AR
hyperplasia
17,20-Lyase
deficiency
17-Hydroxylase
deficiency
Androgen
resistance
syndrome
Micropenis
Isolated micropenis with otherwise normally formed genitalia generally is
not considered as ambiguous genitalia. This condition is associated with
insufficient testosterone secretion during the third trimester. The
evaluation of micropenis is discussed under hypopituitarism, the most

common treatable cause of this condition.
EVALUATION
The evaluation of a newborn with ambiguous genitalia should be
managed expediently by a team of experienced providers. Parents should
be reassured that incomplete or excessive differentiation of the genitals
occurred as part of a continuum in the developmental process and that
the appropriate sex will be determined within several days. It is our
general philosophy not to discuss pending studies in detail because there
are occasions for gender assignment that are inconsistent with either
chromosomal or gonadal sex and presentation of all data available
enables a more cohesive explanation. As in any diagnostic problem, the
approach to the child with ambiguous genitalia should begin with a
thorough history, a careful physical examination, and appropriate
laboratory and radiologic testing. Table 39-2 outlines the different
causes of sexual ambiguity.
A history of drug ingestion, particularly in the first trimester, or recent
androgenic changes in the mother might suggest the cause of female
pseudohermaphroditism. First trimester infections or exposure to
teratogens might suggest gonadal dysgenesis. A family history of an
unexplained neonatal death or siblings with virilization or precocious
puberty, might suggest the diagnosis of CAH although a history of female
relatives with sexual infantilism suggests X-linked causes such as AIS.
A thorough physical examination is important, but on no account should
a diagnosis be based on the physical
P.920
findings. The presence or absence of palpable gonads helps to
differentiate the major categories of intersex conditions. In general,
gonads lacking testicular elements will not descend below the inguinal
region. Thus, a palpable gonad excludes the diagnosis of female
pseudohermaphroditism in which the gonads are ovaries by definition.
Measurement of the length and diameter of the penis is valuable both for
prognostic information and also as a baseline if treatment is given to
enlarge the penis. The urethral opening should be identified and the
existence or absence of a vagina should be determined. The degree of
fusion of the labial-scrotal folds and the presence of associated urinary
or GI tract anomalies should be assessed.
TABLE 39-2 ETIOLOGY OF AMBIGUOUS GENITALIA
Disorders of gonadal determination

Complete or partial gonadal dysgenesis
Sex reversal
True hermaphroditism
Virilization of females
Congenital adrenal hyperplasia
21-Hydroxylase deficiency
11-Hydroxylase defiency
3-Hydroxysteroid dehydrogenase deficiency
Chromosomal aberrations
XO/XY
XX/XY
Variants
Maternal virilization
Drug-induced
Excess androgen production by mother
Aromatase deficiency
Idiopathic
Isolated
Associated with midline congenital anomalies
Inadequate masculinization of males
Defects in testosterone biosynthesis
17-hydroxylase deficiency
17--hydroxysteroid dehydrogenase
StAR
3-hydroxysteroid dehydrogenase deficiency
Androgen resistance syndromes
5-Reductase deficiency
Androgen receptor defects
Congenital anorchia/vanishing testis
Teratogenic insult
Idiopathic
Isolated
Associated with midline congenital anomalies
The physical examination can help direct the laboratory and radiologic
investigation. Certain tests are obtained as soon as it is apparent that
there is sexual ambiguity, although others may be required at a later
stage to make an accurate diagnosis (Table 39-3). For example, serum
17-hydroxyprogesterone and electrolytes are useful initial screening
tests for congenital adrenal hyperplasia but other steroid precursors and
genetic studies may help establish the specific diagnosis. In the newborn
period, testosterone, follicle-stimulating hormone (FSH), and luteinizing
hormone (LH) should be drawn to assess the hypothalamic-pituitarygonadal axis. Serum testosterone can be elevated from either gonadal or
adrenal production and should be interpreted in the context of the
examination and other laboratory studies. Measurement of MIS may help
determine the presence of testicular tissue (32). It should be stressed
that sex assignment does not require that all studies leading to a final
diagnosis be completed (e.g., the exact type of congenital adrenal
hyperplasia may be important for genetic counseling and future prenatal
diagnosis, but not for sex assignment). The karyotype can help
determine whether the infant is a virilized female or an inadequately
virilized male. This, however, should not be used as the primary criteria
for sex assignment as other factors such as gonadal function, sensitivity
to androgens, future sexual function and potential for fertility or
pregnancy (even if by in vitro fertilization) are also critical.
Pelvic ultrasound to evaluate the internal genital structures and gonads

should be performed by an experienced radiologist. Ultrasonography may
identify nonpalpable gonads, and may be able to define the echogenic
pattern as either ovarian or testicular tissue (33). The presence of a
uterus indicates a lack of MIS action and probable absence of functioning
testicular tissue in early gestation, and usually indicates that the child
should be raised as a female. Conversely, the absence of mullerian
structures implies the presence of functioning testicular tissue that
secreted MIS at 7 to 9 weeks of gestation. This is consistent with sexdetermining region Y (SRY) expression and suggestive of an XY
karyotype, but is not a major determinant for male sex assignment. The
karyotype, phallic size and degree of hypospadias, internal genital
structures, gonadal pathology, and etiology of the intersex condition are
all part of the equation in gender determination.
TABLE 39-3 STUDIES TO EVALUATE AMBIGUOUS GENITALIA
Immediate studies
Karyotype
Pelvic ultrasonography
Serum
Electrolytes
17-Hydroxyprogesterone
17-OH pregnenolone
Testosterone
11-Deoxycortisol
Dihydrotestosterone
Mullerian inhibiting substance
FSH/LH
Follow-up studies
hCG stimulation testing
Cortrosyn stimulation testing
Genito-urethrogram and other radilogic studies

Exploratory laparotomy and gonadal biopsy
Skin biopsy to evaluate androgen action
Genetic/molecular studies for specific mutations
Figure 39-2 An algorithm for evaluating sexual ambiguity in infants

with mullerian structures.
P.921
To evaluate further the specific etiology of the sexual ambiguity,

secondary studies may be necessary. The algorithms in Figs. 39-2 and
39-3, which are based on the initial ultrasound findings, delineate the
steps that may be necessary to make a definitive diagnosis. These
algorithms do not include patients with normal male or female
phenotypes that are discordant with the genotype. Surgical exploration
frequently will be required in cases of true hermaphroditism or partial
gonadal dysgenesis but may be done at a later time. It should be
stressed that the final histopathologic diagnosis is not necessary for sex
assignment.
After the evaluation is complete, including consultations from an
endocrinologist and an urologist, the appropriate sex assignment is
determined by a consensus of opinion from the team that might also
include a geneticist, a psychiatrist or psychologist, the pediatrician, and
clergy or other support personnel. Parental input is also considered in
the decision, especially in cases in which the appropriate sex assignment
is uncertain. Gender identity, and future sexual function and fertility are
major determining factors. The attending physician should discuss the
condition fully with the parents, including expectations for future sexual
function and fertility and whether any hormonal medications or surgery
are recommended.
Gender assignment for most infants with ambiguous genitalia is
straightforward when chromosomal sex and gonadal sex correlate with
the internal structures. The external genitalia may require reconstructive
surgery to improve function and cosmetic appearance. The timing of
surgery has become controversial and highly politicized as a result of
heightened concerns about ethical issues of informed consent by the
child, the possibility of gender dysphoria and sex reassignment, and the
risk for postsurgical loss of genital sensation. With little long-term data
to support early vs. late reconstructive surgery, it may be prudent to
postpone surgery until gender identity is clear and
P.922
to have the full participation of the family (and child) in the decision.
Hormonal therapy may be required to induce secondary sexual
maturation, but is usually not needed during the neonatal period. Rarely,
as in cases of partial androgen resistance syndromes, true

hermaphroditism, or mixed gonadal dysgenesis, gender assignment
contrary to chromosomal or gonadal sex must be considered. In these
cases, careful consideration must be given to the likelihood of gender
role and sexual function as an adult (34).
Figure 39-3 An algorithm for evaluating sexual ambiguity in infants

without mullerian structures.
Severe micropenis or agenesis previously warranted a female sex

assignment despite the presence of testes or a normal 46, XY karyotype,
but this approach is now under reappraisal. Recent reports of
dissatisfaction with a female sex assignment in some 46, XY individuals
with cloacal exstrophy or other nonhormonally mediated causes of
aphallia reinforce the need to explore new paradigms for sex assignment.
These paradigms will need to include other factors that affect adult
gender identity such
P.923
as the role of prenatal hormones on central nervous system (CNS) sex
differentiation.
DISORDERS OF THE HYPOTHALAMUS AND

PITUITARY
Development of the Hypothalamic-Pituitary
Axis
The hypothalamus arises by proliferation of neuroblasts in the
intermediate zone of the diencephalic wall and formation of the
supraoptic and periventricular nuclei. The anterior pituitary, or
adenohypophysis, arises embryonically from an invagination of the oral
ectodermal cavity called the Rathke pouch. This diverticulum arises at 3
weeks of gestation, and by 5 weeks has migrated upward to its final
position and separated completely from the oral cavity. Concordantly,
the posterior pituitary, or neurohypophysis, is formed from a downward
invagination of the floor of the diencephalon. Neural fibers migrate from
the hypothalamus down to the posterior pituitary to form the
neurohypophyseal tract. The hypothalamus regulates the pituitary by
secreting both stimulatory and inhibitory hormones. The stimulatory
hormones include growth hormone-releasing hormone (GHRH),
thyrotropin-releasing hormone (TRH), corticotropin-releasing hormone
(CRH), and gonadotropin-releasing hormone (GnRH). The main inhibitory
hormones are somatostatin, which inhibits growth hormone release and
prolactin inhibitory factor, which inhibits prolactin release. In response
to these hypothalamic hormones, the anterior pituitary secretes

prolactin, growth hormone (GH), thyroid-stimulating hormone (TSH),
adrenocorticotropic hormone (ACTH), LH, and FSH. The posterior
pituitary secretes vasopressin and oxytocin. The hypothalamic and
pituitary glands are functional after week 12 of gestation.
TABLE 39-4 ETIOLOGY OF DISORDERS OF THE HYPOTHALAMICPITUITARY AXIS
Malformations
Cleft lip and palate
Optic nerve atrophy
Septooptic dysplasia
Transphenoidal encephalocele
Holoprosencephaly
Anencephaly
Trauma associated with breech delivery
Congenital infection
Rubella
Toxoplasmosis
Tumor
Hypothalamic hamartoblastoma (i.e., Pallister Hall syndrome)
Rathke pouch cyst
Craniopharyngioma
Glioblastoma
Isolated or combined familial or idiopathic pituitary hormone
deficiency
Autosomal recessive or X-linked recessive familial
panhypopituitarism
Most disorders of the hypothalamic-pituitary axis in the newborn period,

except for the syndrome of inappropriate secretion of antidiuretic
hormone (SIADH), are those of insufficiency related to malformations,
trauma, infection or genetically inherited disorders, as outlined in Table

39-4. This differs from older children and adults who may have either
functionally active tumors that secrete pituitary hormones or infiltrative
disease or tumors that interfere with normal pituitary function.
DISORDERS OF THE ANTERIOR PITUITARY

Anterior pituitary dysfunction is difficult to detect in the newborn. The
predominant features of anterior pituitary insufficiency are
hypoglycemia, micropenis and, occasionally, cholestatic jaundice. The
hypoglycemia may be quite severe and comparable to that seen in
infants with congenital hyperinsulinism. The infants may even have a
brisk glycemic response to glucagon causing further confusion (35). The
cholestatic jaundice is initially unconjugated, then becomes
predominantly conjugated, and will often only resolve after hormone
replacement. There may be combined deficiency of multiple pituitary
hormones or isolated deficiency of a single hormone. The molecular basis
of multiple hormone deficiency is established for a number of genetic
defects and was reviewed recently by Parks and associates (36).
Evidence for the importance of the POU family of pituitary transcription
factors in establishing pituitary lineages came from identifying the roles
of Pit-1 in the Snell dwarf mouse (37), Prop-1 in the Ames dwarf mouse
(38) and P-Lim knockout mice (39). In humans P-Lim is associated with
deficiencies of all anterior pituitary hormones except ACTH in conjunction
with retinal colobomas. Pit-1 mutations cause pituitary gland hypoplasia
and deficiencies of GH, TSH, and prolactin. Prop-1 (Prophet of Pit-1) is
essential for Pit-1 expression, and therefore has an identical clinical
picture. Hypopituitarism in conjunction with optic nerve hypoplasia and
absence of the septum pellucidum comprise the syndrome of Septo-optic
dysplasia (SOD). Pituitary function can vary from intact to
panhypopituitarism including diabetes insipidus. Mutations in a
homeodomain protein, HESX-1, have been found in some patients with
SOD associated with mild panhypopituitarism or isolated GH deficiency
(40). SOD is often suggested by wandering nystagmus in the newborn,
reflective of optic nerve hypoplasia and blindness.
Growth Hormone Deficiency

Growth hormone deficiency in the neonate may present as hypoglycemia,
micropenis, or both. Micropenis is defined as penile size less than 2.5 cm
stretched length in the term
P.924
infant. Congenital deficiency of growth hormone does not cause
intrauterine growth retardation and often does not affect linear growth
until 6 to 9 months of age. Intrauterine growth is primarily determined
by maternal factors, including nutritional status, placental function, and
gestational infection or drugs. During early postnatal life, thyroid
hormone, insulin, and nutrition are more important growth determinants
than growth hormone. A family history of short stature is pertinent
because familial autosomal dominant inheritance of growth hormone
deficiency is well recognized.
Gonadotropin Deficiency
Gonadotropin deficiency can occur as either isolated hypogonadotropic
hypogonadism or combined multiple pituitary hormone deficiency.
Although infants with combined deficiencies present with micropenis,
those with isolated gonadotropin deficiency may not be recognized at
birth. The genitalia can be normal male in Kallmans syndrome
(hypogonadotropic hypogonadism and anosmia), a syndrome caused by
mutations in the KAL gene encoding anosmin-1. Female infants are
asymptomatic at birth and may not be identified until puberty fails to
occur. Other causes of micropenis associated with gonadotropin
deficiency include syndromic conditions, such as Noonan syndrome and
Prader-Willi syndrome.
Adrenocorticotropic Hormone Deficiency

Adrenocorticotropic hormone deficiency rarely presents as an acute
adrenal crisis; the cortisol insufficiency is typically mild and may cause
hypoglycemia or hyponatremia without hyperkalemia, and occasionally,
prolonged direct hyperbilirubinemia. Isolated ACTH deficiency is
extremely rare but has been linked to the CRH gene locus (41). The
combination of both growth hormone and ACTH deficiency may cause
hypoketotic hypoglycemia of such severity that it is difficult to

differentiate from congenital hyperinsulinism (35).
Thyroid-Stimulating Hormone Deficiency

Thyroid-stimulating hormone deficiency is essentially asymptomatic in
the newborn. On newborn screening tests, the serum thyroxine (T4)
concentration is low or low normal with the TSH in the normal range.
This finding may be misinterpreted as the euthyroid sick syndrome (see
Disorders of the Thyroid) in a stressed neonate. Furthermore, secondary
hypothyroidism may be missed if primary TSH screening is used. TSH
deficiency is usually associated with other pituitary deficiencies and
rarely occurs in isolation. In an infant with any of the CNS abnormalities
outlined in Table 39-4, secondary hypothyroidism should be considered
and may be missed with routine newborn screening procedures.
Diagnosis
The diagnosis of hypothalamic and pituitary deficiency may require
stimulation testing if random values are nondiagnostic. Growth hormone
is tonically elevated in the first few days of life, thus a random growth
hormone greater than 10 ng/mL suggests adequate growth hormone
secretion. If a random value is low, growth hormone provocative testing
is needed to confirm deficiency. In normal newborn infants, growth
hormone values increase to greater than 25 ng/mL with stimulation
testing. ACTH deficiency causing adrenal insufficiency is unlikely if a
random cortisol is greater than 20 mcg/dL, because serum cortisol is low
in newborns, without diurnal variation. In general, ACTH or CRH
stimulation testing is necessary to test the hypothalamic-pituitaryadrenal axis. Random sex steroids, FSH, and LH may be diagnostic at 1
to 3 months of age when the hypothalamic-pituitary-gonadal axis is
active transiently, otherwise gonadotropin-releasing hormone stimulation
testing is needed to assess LH and FSH secretion.
In those infants suspected of anterior pituitary deficiency,
ultrasonography through the open fontanelle may discern mid-line
malformations of the brain, although magnetic resonance imaging or
computed tomography scanning is more sensitive. If septo-optic
dysplasia is a consideration, ophthalmologic examination should be
performed.
Treatment
Anterior pituitary deficiency may not be detected clinically during the
neonatal period if the hypoglycemia is mild, the micropenis, obviously
not a clinical feature in hypopituitary females, is marginal, and jaundice
is not severe. Treat-ment considerations, therefore, are based on the
severity of symptoms. The child who is severely hypoglycemic will
require immediate growth hormone and glucocorticoid replacement albeit
at relatively modest doses. Recombinant growth hormone is injected
subcutaneously, at a dose of 0.04 mg/kg daily. Glucocorticoid
replacement with 8 to 10 mg/m 2 of oral hydrocortisone per day is often
sufficient. This dose should be at least tripled for acute illness. If the
testes are nonpalpable, MIS determination will ascertain their presence
(32). An MIS value in the normal male range for age verifies the
presence of testes. In male infants with micropenis, testosterone
enanthate at a dose of 25 mg every month can be administered to
stimulate penile growth. If the penile response is inadequate after a
three month course, this can be repeated.
DISORDERS OF THE POSTERIOR PITUITARY

Vasopressin or antidiuretic hormone (ADH) and oxytocin are the two
major posterior pituitary endocrine hormones. Oxytocin has no known
function in the neonate although
P.925
ADH helps to regulate intravascular volume and osmolality. ADH is
synthesized in the supraoptic and periventricular nuclei of the
hypothalamus by 12 weeks of gestation. It is bound to neurophysin and
is transported along the neurons of the neurohypophyseal tract to the
posterior pituitary, where it is stored and released as necessary. ADH
increases the permeability of the collecting tubules of the kidney to
water and urea. Its secretion is stimulated by hyperosmolar states and
volume depletion and inhibited predominantly by volume overload. The
two main disorders of ADH secretion are diabetes insipidus and the
syndrome of inappropriate SIADH.
Diabetes Insipidus
Diabetes insipidus (DI) in the newborn may be as a result of central ADH
insufficiency or renal unresponsiveness to ADH (nephrogenic DI). This
section will discuss only central DI.
DI in the neonate may present with failure to thrive, irritability, fever,
vomiting, hypernatremia, and a history of polyhydramnios. Polyuria is
difficult to appreciate in newborn infants because healthy newborn
infants can void up to 20 times a day (42). However, sustained urine
outputs greater than 60% of fluid input, and single-void volumes of
greater than 6 mL/kg suggest DI. In a child with hyperosmolar serum,
the diagnosis is confirmed by finding inappropriately dilute urine that
becomes more concentrated after vasopressin administration. Failure to
respond to vasopressin is suggestive of renal DI. Water deprivation tests
should not be done in newborns as acute dehydration and hypernatremia
may cause permanent CNS injury.
A list of causes of central DI is given in Table 39-5. Secondary DI is
more common than primary in the neonatal period and should be
strongly suspected in infants with certain malformations.
TABLE 39-5 ETIOLOGY OF CENTRAL DIABETES INSIPIDUS
Primary
Familial
X-linked recessive
Autosomal dominant
Idiopathic
Secondary
Malformation sequences
Optic atrophy
Septooptic dysplasia
Holoprosencephaly
Birth trauma
Periventricular hemorrhage
Infection
Meningitis
Encephalitis
Infiltrative disease (in older infants)
Histiocytosis X
Granulomatous disease
Germ cell tumors (in older children)
Treatment
Treatment of DI requires strict management of fluid balance. Infants with
DI require enormous quantities of free water; it is not unusual to provide
several times the usual maintenance quantities of water as 5% glucose
intravenously, although providing nutrition and electrolytes by the oral
route. Desmopressin is a long-acting analogue of vasopressin. A number
of different formulations are available; oral or subcutaneous may be
easiest in infants. Sublingual administration can be helpful in patients
with cleft lip and palate. Sublingual or subcutaneous dosing starts at 1 to
2 mcg once or twice daily and oral doses are about 10- to 20-fold higher.
The dose and dose interval must be carefully titrated in each child by
monitoring fluid intake, urine output, serum electrolytes and osmolality,
and state of hydration. Management should include a breakthrough
period of diuresis daily to avoid fluid overload although providing
sufficient milk intake to meet caloric needs. Rapid shifts in the serum
sodium caused by excessive fluid input or urine output should be

avoided. An alternative approach, which minimizes the risk of water
overload, is to use a diluted formula without administering any
vasopressin. This treatment is based on the principle that hunger rather
than thirst is the driving force behind fluid intake in the neonate.
Providing the total daily caloric intake as one-third strength formula will
usually result in stable fluid balance. This approach requires two to three
hourly feeding even during the night, thus sleep may be disrupted.
Moreover the volume of fluid needed can compromise growth in some
infants. For emergency treatment of severe dehydration, intravenous
aqueous pitressin infusion, instead of desmopressin, is recommended as
the short half life of pitressin allows precise control of fluid balance.
Syndrome of Inappropriate Antidiuretic

Hormone Secretion
The secretion of ADH is normally higher in premature infants (43) but
can be further increased in sick premature infants for the reasons
outlined in Table 39-6. A common mechanism in many pathologic cases is
intravascular volume depletion, which is detected by stretch receptors in
the left atrium. Thus, the elevated ADH levels are appropriate for the
volume status, but inappropriate for the osmolar status. SIADH, by
definition, occurs in a volume replete or overloaded state when there is
dilutional hyponatremia associated with inappropriately concentrated
urine with continued sodium loss (urine sodium >20-30 mEq/L). This
occurs in the absence of volume depletion, renal failure, or adrenal
insufficiency. True SIADH is uncommon in neonates (44), and should be
differentiated from hyponatremia caused by ADH levels that are
appropriately elevated in response to volume depletion. It is vitally
important to limit water and sodium intake and prevent hyponatremia in
SIADH, but equally important to adequately treat the volume depletion
states associated with appropriately increased ADH secretion.
TABLE 39-6 CAUSES OF ELEVATED LEVELS OF ANTIDIURETIC

HORMONE IN THE NEWBORN
Birth asphyxia
Acute deterioration of hyaline membrane disease and
bronchopulmonary dysplasia
Respiratory syncytial virus infection
Pneumothorax
Pulmonary interstitial emphysema
Artificial ventilation
Acute blood loss
Periventricular hemorrhage
Surgery
Pain
Syndrome of inappropriate ADH secretion
ADH, antidiuretic hormone.
P.926
Hyponatremia occurs commonly in newborn premature infants who have
a higher fractional excretion of sodium than term infants. The most
common nonphysiological cause of hyponatremia is renal sodium wasting
as a result of diuretics. The differential diagnosis of hyponatremia in the
newborn includes prerenal failure, renal failure, adrenal insufficiency and
SIADH. The SIADH, if it occurs, is associated more commonly with sepsis
and central nervous system infection in older infants, but perhaps in
critically ill neonates as well. Unlike volume depletion states, SIADH is
treated by fluid restriction. If volume depletion is evident, combined with
polyuria, urinary sodium loss, and hyponatremia, salt wasting should be
suspected.
DISORDERS OF THE ADRENAL GLAND

Development and Function of the Adrenal
Gland
The adrenal gland consists of a cortex and medulla, which each function
independently and secrete two different classes of hormones. The fetal
adrenal cortex is of mesodermal origin, whereas the chromaffin cells of
the adrenal medulla are of neuroectodermal origin. Diseases of the
adrenal medulla during the neonatal period are extremely rare, thus this
section will focus on disorders of the adrenal cortex.
The anlage of the adrenal cortex arises as two large masses on either
side of the aorta, at the level of the first thoracic nerve, immediately
adjacent to the medullary cells that have migrated from the neural crest.
Fetal adrenal cortical cells can be identified by 4 weeks gestation. By 7
weeks of gestation, the medullary cells begin to migrate to the interior of
the adrenal cortex. The coelomic epithelium envelops the original cortical
cells and remains as an outer shell. The fetal adrenal gland is
steroidogenically active and large during gestation, but involutes during
the latter half of pregnancy and especially after birth, suggesting a role
in the maintenance of pregnancy. The adult adrenal cortex slowly
develops from the outer shell although the fetal zone undergoes
involution. DAX-1, a gene on the X chromosome, is essential for
development of the definitive zone of the adrenal cortex (3). Mutations
in this gene are responsible for congenital adrenal hypoplasia.
The trophic hormonal control of the fetal adrenal is not clear. In
anencephalic fetuses, the fetal adrenal appears to develop normally
during the first 12 weeks of gestation, then involutes. In patients with
enzymatic defects of cortisol biosynthesis however, the hyperplasia and
increased activity of the adrenal glands during the first 12 weeks of
gestation suggests that ACTH must play some role during that time.
The adrenal cortex secretes three main types of steroid hormones,
glucocorticoids, mineralocorticoids, and androgens. The glucocorticoids,
of which cortisol (hydrocortisone) is the most important, exert their
major physiologic effects on carbohydrate, protein, and fat metabolism.
The mineralocorticoids, desoxycorticosterone and aldosterone, maintain

salt and water balance by promoting sodium retention in exchange for
hydrogen and potassium in the distal convoluted tubules of the kidney.
The adrenal androgens, dehydroepiandrosterone (DHEA), 4androstenedione, and 11-hydroxyandrostenedione, are anabolic and
responsible for the development of sexual hair in girls at puberty. The
slightly higher levels of adrenal androgens during the neonatal period
may be secondary to the relative deficiency of 3-hydroxysteroid
dehydrogenase in the fetal zone of the fetal adrenal cortex, which is
reflected in the higher concentrations of 5 steroids (e.g., DHEA, 17-OH
pregnenolone) especially in premature infants.
The production of adrenocortical steroids is controlled by a
hypothalamic-pituitary-adrenal homeostatic mechanism. Hypothalamic
CRH stimulates release of pituitary ACTH, which in turn, stimulates
cortisol biosynthesis. Increased levels of cortisol exert down-regulation
of the axis, probably at the level of the hypothalamus.
Aldosterone secretion is controlled by the renin-angiotensin system
rather than ACTH. Acute changes in pressure receptors control the
release of renin from the juxtaglomerular cells of the kidney. Circulating
renin, in turn, increases angiotensin II, which acts on the zona
glomerulosa of the adrenal cortex to stimulate aldosterone secretion and
vascular contractility. The increased blood volume and higher pressures
within the arterial receptors exert negative-feedback inhibition of the
renin-angiotensin system. Secondary mechanisms such as a low sodium
or high potassium intake also increase aldosterone excretion. ACTH will
cause a transient, albeit unsustained, increase in aldosterone excretion,
and aldosterone secretion will be diminished in the absence of ACTH.
Finally, cortisol itself may have a permissive role in aldosterone action at
the tissue level.
P.927
ADRENAL INSUFFICIENCY
The disorders of the adrenal cortex during the neonatal period consist
almost entirely of those conditions that cause adrenal insufficiency. The
inborn errors of steroid biosynthesis (CAH) can cause excessive

production of various steroids, but Cushing syndrome or cortisol excess
is rare during the neonatal period. Cushing syndrome may occur
secondary to exogenous steroids such as dexamethasone, but Cushing
disease has not been described in the very young infant. Adrenal cortical
tumors occur in older infants but have not been reported in the neonate.
Adrenal insufficiency can result from hypopituitarism, adrenal
hemorrhage and other adrenal injury, ACTH receptor abnormalities,
inherited degenerative disorders, or inborn errors of steroid biosynthesis.
Adrenocorticotropic Hormone Insufficiency

In a number of neonatal deaths associated with shock and peripheral
vascular collapse in conjunction with severe hyponatremia and
hyperkalemia, the adrenal glands were noted to be hypoplastic at
autopsy. Because some of these cases were reported in infants with
anencephaly or with partial or total pituitary aplasia, the lack of ACTH
was thought to be responsible for the failure of development of the
definitive zone after birth. The possibility, however, of another critical
trophic factor is suggested by the findings in patients with congenital
hypopituitarism. These patients have decreased cortisol production, but
normal mineralocorticoid function and rarely develop hyperkalemia. They
develop hypoglycemia, poor feeding, and failure to thrive, but can, in
general, maintain water and electrolyte balance and respond to sodium
deprivation with an increase in aldosterone excretion.
Familial isolated glucocorticoid insufficiency. This is an autosomal
recessive disorder that can present during the neonatal period or later in
childhood with shock, hyperpigmentation, hypoglycemia, and failure to
thrive. These patients have cortisol insufficiency and cannot increase
serum cortisol or urinary 17-hydroxysteroid excretion in response to
ACTH stimulation. They can, however, respond to sodium deprivation
with increased aldosterone excretion and decreased sodium excretion.
Some pedigrees have a defect in the melanocortin-2 receptor (ACTH
receptor) although others may have postreceptor defects (45). In one
family with five affected siblings, two had intact glucocorticoid function
during early infancy and developed glucocorticoid deficiency at a later
age, suggesting an inherited degenerative process of the adrenal glands

as the pathogenetic basis in this family (46).
Damage
Adrenal insufficiency can occur during the newborn period as a result of
damage to the relatively large and hyperemic adrenal glands. Trauma in
association with a difficult delivery, particularly breech delivery;
hemorrhagic diseases; or infectious processes can damage the adrenal
glands. Minor hemorrhage or unilateral damage may not cause adrenal
insufficiency, and may present subsequently as calcification of the
adrenal glands detected on an abdominal radiograph obtained for other
purposes. All patients with shock in association with hyponatremia
should be suspect for adrenal insufficiency. The highly sensitive ACTH
determinations, using monoclonal antibodies and immunoradiometric
assays, can detect elevated levels of plasma ACTH that are diagnostic of
primary adrenal insufficiency.

CAH is a genetic disorder involving a deficiency of one of several
enzymes required for normal steroid biosynthesis. Deficient cortisol
synthesis secondary to an enzymatic deficiency causes increased ACTH
production, which in turn, stimulates a compensatory hypertrophy of the
adrenal cortex and increase in steroidogenesis. This partially
compensates for the block in the biosynthetic pathway, but also leads to
the increased production and accumulation of precursor steroids
preceding the enzymatic defect. Although ACTH chiefly regulates the
glucocorticoid pathway, the synthesis of mineralocorticoids and
androgens are variably affected depending on the specific enzyme
involved.
The principal biochemical reactions in the conversion of cholesterol into
active adrenocortical steroids require a series of hydroxylations (Fig. 394) that are mediated by cytochrome P450 oxidases. The initial step
requires that the StAR protein form contact sites between the outer and
inner membranes of mitochondria to transport cholesterol into the
mitochondria and start the steroidogenic process. Five enzymes are
necessary for the synthesis of cortisol: P450scc, P450c17, P450c21,
P450c11, and 3-hydroxysteroid dehydrogenase enzymes (47). The

genes for all of these enzymes have been cloned and defects in any of
them can cause insufficient glucocorticoid production and CAH. The
clinical manifestations of this syndrome correspond to the particular
enzyme affected and are summarized in Table 39-7.
Virilization
Virilization of the female is secondary to the elevated adrenal androgens
caused by those enzymatic defects subsequent to 17-hydroxylation. In
most cases, the labioscrotal folds are partially fused with clitoral
enlargement, which may be bound down by chordee. Occasionally,
virilization may be so severe that a phallic urethra develops. In male
infants, virilization is generally is not apparent during the neonatal
period, and, the diagnosis in the milder nonsalt-losing forms of this
disorder, can be undetected for several years. Boys can present later
with secondary sexual changes, increased somatic growth, and welldeveloped musculature. The classic, and most prevalent virilizing form of
CAH is a defect in cytochrome P450c21 (21-hydroxylase deficiency),
accounting for almost 90% of recognized cases (48).
P.928
Mutations in P450c11 and 3-hydroxysteroid dehydrogenase also cause
female virilization.
Incomplete Masculinization
Failure of complete masculine development occurs in those forms of
adrenal hyperplasia in which the androgen pathway is affected.
Incomplete masculinization in the male, which requires fetal testosterone
production, suggests that the enzymatic defects occur in both the
adrenal gland and the testis (28). In the 3-hydroxysteroid
dehydrogenase defect, secreted steroids consist almost entirely of
compounds with 5-3-hydroxy configuration (49). Fetal testosterone
production by the testis is also impaired, causing incomplete
masculinization in the male (50). The marked elevation of 5-3hydroxyadrenal androgens, especially DHEA, however, is converted
peripherally to active androgens that virilize the female infant.
Elevation
Figure 39-4 The biosynthet ic pathway of adrenal steroid. The classic enzyme
terminology is represented by the alphabetical letters with the appropriate
cytochrome P450 oxidases in parentheses. A, 20,22-desmolase (P450scc); B, 3hydroxysteroid dehydrogenase: C, 17-hydroxylase (P450c17); D, 21-hydroxylase
(P450c21); E, 11-hydroxylase (P450c11); F, 17,20-lyase (P450c17); G, 17-keto
reductase; H+, 18-hydroxylase + 18-oxidase (P450c11).
P.929
of serum 17-hydroxypregnenolone is diagnostic of 3-hydroxysteroid
dehydrogenase deficiency, although 17-hydroxyprogesterone
concentrations also are markedly elevated (50).
TABLE 39-7 CLINICAL AND BIOCHEMICAL FINDINGS OF THE COMMON VARIANTS OF CONGENITAL
ADRENAL HYPERPLASIA
Phenotype
Enzyme Deficiency
(Classic)
46XX
46XY
Other Clinical
Manifestations
Predominant Steroids
Congenital lipoid
hyperplasia
Female
Female
Salt-wasting crisis
Low level-all steroids No response

to ACTH
3-Hydroxysteroid
dehydrogenase deficiency
Virilized
Hypospadias
Salt-wasting crisis
Dehydroepiandrosterone 17-OH
pregnenolone Increased 5 - 4
ratio of steroids
21-Hydroxylase deficiency
Late virilization in female
Salt-wasting crisis
Virilized
Male
Pseudoprecocious
puberty in male
17-OH progesterone
Androstenedione Testosterone
11-Hydroxylase
deficiency Hypertension
Virilized
Male
Pseudoprecocious
puberty in male
11-Deoxycortisol 11Deoxycorticosterone
Androstenedione Low renin
17-Hydroxylase
deficiency
Female
Female
Sexual infantilism
Hypertension
Corticosterone 11Deoxycorticosterone (Low renin)
ACTH, adrenocorticotrophic hormone.
Hypertension
Hypertension has been associated with enzymatic blocks resulting in
excessive secretion of mineralocorticoids. A defect of cytochrome
P450c11 (11-hydroxylase deficiency) causes an accumulation of
desoxycorticosterone, a potent mineralocorticoid, and 11-deoxycortisol
(47). The P450c17 defect (17-hydroxylase deficiency) blocks 17hydroxylation of progesterone, interfering with cortisol and androgen
biosynthesis, and shunting steroid production to the mineralocorticoid
pathway (51). Genital development in females is unaffected but males
are undervirilized. The hypertension resulting from excess
mineralocorticoid production, however, is an inconstant feature and it is
not known if hypertension is present during the newborn period. Whether
the hypertension is related to the duration of excessive secretion of
mineralocorticoid, the severity of the defect, or variations in sodium
intake is also unclear.
Salt Loss
Mineralocorticoid insufficiency and severe sodium loss are seen in the
salt-losing form of 21-hydroxylase deficiency, and 3-hydroxysteroid
dehydrogenase deficiency. The electrolytes initially are normal, but,
within the first week of life, serum sodium will slowly decrease with a
concomitant rise in serum potassium. These infants may manifest acute
adrenal crisis with shock, peripheral collapse, and dehydration, by a
week of age.
The underlying metabolic defects for two clinical variants of the 21hydroxylase enzyme defect are now understood. Bongiovanni and
Eberlein postulated that both are the result of the same enzymatic defect
(52). In the salt-loser, there is almost complete 21-hydroxylase
deficiency, whereas in the simple form, there is sufficient 21-hydroxylase
to permit aldosterone synthesis. A single gene mediates the
hydroxylation of both progesterone and 17-hydroxyprogesterone.
Different mutations of the P450c21 gene account for the heterogeneity of
21-hydroxylase deficiency disorders, including the nonclassic late-onset
variant, although there is phenotypic variability with the same genotype
(47,48).
A few instances of aldosterone deficiency caused by a specific defect of

18-dehydrogenase, the last enzymatic process in aldosterone synthesis,
have been described (53,54). There is salt and water loss, without the
other clinical consequences of congenital adrenal hyperplasia. These
disorders are secondary to mutations of the P450c11 enzyme.
Congenital Lipoid Hyperplasia

Previously, this disorder was thought to be as a result of a deficiency of
the enzyme, 20,22 desmolase (P450scc) that mediates the conversion of
cholesterol to pregnenolone. The etiology is now recognized as a genetic
defect of the StAR protein, a mitochondrial protein needed for transport
of cholesterol across the mitochondrial membrane. StAR is needed for
acute steroidogenesis, thus in some individuals, glucocorticoid
production is initially preserved, but the accumulation of cholesterol
esters causes continuing destruction of the adrenal and onset of adrenal
insufficiency at older ages (55). Male and females both have female
appearing external genitalia, hyponatremia,
P.930
hyperkalemia and dehydration. However, the age at presentation varies
from the newborn period to several months of age and later. No steroids
are produced and the adrenal glands are markedly enlarged, filled with
cholesterol esters.
Prenatal Diagnosis and Treatment of

CAH, 21-hydroxylase deficiency, can be diagnosed prenatally using
molecular techniques. Once the diagnosis of 21-hydroxylase deficiency
has been established in a propositus, molecular analysis of the P450c21
gene should be done in the propositus and the parents. The parents
should have ACTH stimulation testing to confirm biochemically that they
are genetic heterozygotes for 21-hydroxylase deficiency. The
management of the pregnancy, the techniques used for diagnosis, the
treatment of the female fetus with 21-hydroxylase deficiency, and the
problems with these techniques have been reviewed (56,57). In brief,
the mother is started on dexamethasone early in the first trimester and,

subsequently, chorionic villus sampling or amniocentesis is performed to
determine the sex and genetic studies. If molecular and genetic
techniques confirm the diagnosis and the fetus is female, dexamethasone
treatment is continued to term. Long-term consequences of prenatal
steroids remain unknown and therapy can be associated with cushingoid
symptoms in the mother, thus prenatal management and therapy should
be monitored closely and performed at centers with expertise.
Diagnosis of Adrenal Insufficiency

The diagnosis of acute adrenal insufficiency is difficult to make in the
newborn. There must be a high index of suspicion in any acutely ill infant
with shock, peripheral collapse, and a rapid and weak pulse, and in any
infant with poor feeding, failure to thrive, intermittent pyrexia, or even
hypoglycemia and convulsions. A subtle sign of congenital adrenal
hypoplasia is hyperpigmentation, especially in the extensor creases and
genitalia; however, this sign is seldom recognized until the diagnosis has
been made. Decreased serum sodium and chloride and increased serum
potassium levels are suggestive of mineralocorticoid deficiency. Isolated
hyponatremia does not exclude glucocorticoid insufficiency, and should
be viewed as a possible sign of adrenal insufficiency. Certainly,
ambiguous external genitalia at birth always should suggest the
possibility of CAH.
Serum cortisol levels are low in all newborns, especially in premature
infants and lack diurnal variation, therefore random cortisol
determinations usually are not diagnostic. In clinical situations highly
suggestive of adrenal insufficiency, it is recommended that a rapid, 1hour low dose ACTH stimulation test be performed and that
pharmacologic doses of glucocorticoids be administered, along with fluid
and electrolyte resuscitation, after testing. Plasma concentrations of
ACTH are elevated in those infants with primary adrenal insufficiency,
including CAH. A plasma sample for ACTH determination should be
obtained before ACTH testing.
Delineation of the specific enzyme defect in CAH can be determined by
measuring serum concentrations of the various steroidal precursors (see
Table 39-7). Cord blood concentrations of 17-hydroxyprogesterone are

normally elevated to the range of 900 to 5,000 ng/dL. Serum values
rapidly decrease by the second or third day of life to less than 100 ng/dL
in newborns (58), but can increase to greater than 200 ng/dL at 1 to 2
months of age in male infants. Newborn values above 1,000 ng/dl are of
concern although 17-hydroxyprogesterone values are higher in stressed
newborns, especially preterm sick newborns, in whom they can be
greater than 600 ng/dL (59). These values, however remain significantly
lower than those in patients with 21-hydroxylase deficiency, which are
often markedly above 2,000 ng/dL. Serum 17-hydroxyprogesterone is
used for newborn screening for CAH using a filter paper technique (60).
It is important to recognize that elevated serum concentrations of 17hydroxyprogesterone are not diagnostic of the 21-hydroxylase defect.
Serum 17-hydroxyprogesterone can be mildly elevated in the 11hydroxylase defect and can be markedly elevated in the 3hydroxysteroid dehydrogenase defect secondary to peripheral conversion
of 17-hydroxypregnenolone to 17-hydroxyprogesterone (50). Serum
values of 17-hydroxypregnenolone are especially elevated in the
premature infant, and values up to 2,000 ng/dL are normal (59).
Treatment of Adrenal Insufficiency

The immediate need of a critically ill infant in adrenal crisis is for
cortisol. If possible, cortisol should be withheld until the diagnosis can
be established, either by ACTH testing or by obtaining serum for
determination of the appropriate steroids and ACTH. However, if a
newborn infant is in shock and in extremis, glucocorticoids should be
immediately given as a lifesaving measure. In the usual situation, salt
and water alone will relieve the clinical crisis. Intravenous isotonic saline
in 5% glucose water should be infused at a rate of 100 to 120 mL/kg
during the first 24 hours. If the infant is in severe shock, the use of
plasma or 5% albumin, 10 to 20 mL/kg to restore intravascular volume,
and cortisol, is often necessary. Hydrocorti-sone hemisuccinate or
phosphate, 1.5 to 2 mg/kg, should be given intravenously immediately.
Constant infusion of hydrocortisone hemisuccinate or phosphate (50 mg/
m 2 /day) should be continued. Hydrocortisone hemisuccinate, 2 mg/kg,
can be given intramuscularly if intravenous access is a problem. The

infant with severe shock may require a vasopressor, although
vasopressor drugs may not be efficacious until hydrocortisone is
administered.
Hydrocortisone or cortisone acetate, 10 to 12 mg/m 2 /day, is the
mainstay for long-term treatment of patients with adrenal insufficiency.
For acute illness, stress doses of hydrocortisone should be given
parenterally at triple the maintenance dose.
P.931
Often a mineralocorticoid is a necessary adjunct for the chronic
treatment of adrenal insufficiency. The dose of the oral
mineralocorticoid, 9-fludrocortisone (Florine) is 0.05 to 0.1 mg/day,
which is sufficient for most forms of adrenal insufficiency. In the saltlosing forms of congenital adrenal hyperplasia, occasionally higher doses
of fludrocortisone are necessary. Even the compensated (nonsalt-losing)
variant may benefit from low dose mineralocorticoid.
Iatrogenic Adrenal Insufficiency

During the neonatal period, pharmacologic doses of glucocorticoids are
often needed for adjunctive treatment of a number of diseases, such as
bronchopulmonary dysplasia. The dose and duration of glucocorticoid
therapy that will cause iatrogenic adrenal insufficiency is not known,
particularly in infants. High-dose glucocorticoid therapy for a brief
duration (<1 week) probably will not cause adrenal insufficiency,
therefore a slow taper is not needed unless the clinical course of the
primary condition deteriorates. Treatment longer than 14 days may
result in at least transient adrenal insufficiency. After a prolonged
course, the dose of glucocorticoids can be decreased by one-half every
several days until a physiologic replacement dose (10 mg of
hydrocortisone/m 2 /day orally) is achieved. The dose can then be lowered
more gradually, by 20% every 4 or 5 days.
Adrenal function may be suppressed for some time after prolonged
pharmacologic glucocorticoid therapy. Again, there are no studies
correlating dose, duration of pharmacologic treatment, and time needed
for recovery of adrenal function after high-dose glucocorticoid therapy in

infants. There are anecdotal reports of adrenal crisis occurring during
stress more than 6 months after discontinuation of pharmacologic
glucocorticoid therapy. Although it is possible to evaluate periodically the
adrenal response to exogenous ACTH to determine when iatrogenic
adrenal insufficiency has resolved, alternatively, pharmacologic doses of
glucocorticoids can be used empirically during situations of stress for at
least a year after discontinuation of prolonged high-dose glucocorticoid
therapy. A minimal dose of glucocorticoid to be used during stress
situations is 30 mg of hydrocortisone/m 2 / day, orally.
DISORDERS OF THE THYROID

Development and Function of the Thyroid
The fetal thyroid begins as a thickening of epithelium at the base of the
tongue that migrates down the trachea, leaving the thyroglossal duct as
an embryonic remnant. During its caudal migration, the thyroid assumes
a more bilobate shape. The developing thyroid is able to concentrate
iodide by 12 weeks of gestation, and to organify iodide and synthesize
thyroxine (T4) and triiodothyronine (T3) by 14 weeks of gestation. Free
T4 and, more so, T3 can cross the placenta in either direction. It is
probable that the gradient is from mother to fetus during the first 12
weeks and then the gradient in transfer of thyroid hormones is from
fetus to mother except in the hypothyroid fetus (61,62). There is no
placental transfer of maternal or fetal TSH, although thyroid-stimulating
immunoglobulins (TSI) will cross the placenta. The hypothalamicpituitary feedback mechanisms are operative by the latter part of
gestation. The hypothalamus secretes a tripeptide, TRH, which stimulates
pituitary secretion of TSH. TSH, in turn, stimulates thyroid hormone
production by regulating every step of thyroid hormone biosynthesis and
release, from iodide accumulation to proteolysis of thyroglobulin. The
thyroid hormones exercise negative feedback control of TSH response to
TRH at the pituitary level but also inhibits at the hypothalamic level.
The biosynthesis of thyroid hormones is illustrated in Fig. 39-5.
Circulating plasma iodide is concentrated by the thyroid gland, then
oxidized by a thyroid peroxidase and bound to tyrosine to form
monoiodotyrosine (MIT) and diiodotyrosine (DIT). The iodotyrosines are

held by peptide linkage to thyroglobulin and coupled to form T3 and T4.
T3 and T4 are then cleaved from thyroglobulin by thyroid proteases and
secreted into the circulatory system. A large percentage of circulating T3
arises by deiodination of T4 by diodinases 1, 2, and 363. Diodinase 1
(D1) and Diodinase 2 (D2) are differentially localized in the pituitary and
peripheral tissues. At the pituitary level, D2 partially regulates TSH
production by modulating T3 concentrations. Diodinase 3 (D3) converts
T4 to reverse T3 and T2, biologically inactive iodothyronines. Within the
thyroid, iodotyrosines and iodothyronines are deiodinated by
dehalogenase enzymes and remain within the intrathyroidal iodide pool
to be reused. The iodide released from peripheral deiodination enters the
circulatory system to be reconcentrated by the thyroid gland or excreted
by the kidneys. The iodothyronines are transported in the plasma by
proteins. Thyroxine-binding globulin (TBG), an -globulin, is the major
carrier of T4, but binds T3 to a lesser extent. Thyroxine also is bound by
T4-binding prealbumin and by albumin. At the cellular level, the
unbound, free T3 and T4 are biologically active. Genetic disorders,
acquired conditions, or drugs that quantitatively change the
concentration of TBG will alter circulating total T4 without affecting the
physiologic thyroid status.
THYROID FUNCTION TESTS

Thyroid function tests in infants are elevated compared to values in older
children. This is secondary to the cold-stimulated surge of TSH and TRH
in the immediate postnatal period and the higher TBG concentrations
secondary to maternal estrogen effect. Total T4 ranges from 7.3 to 22.9
mcg/dL during the first month of life, with mean values greater than 10
mcg/dL (Tables 39-8 and 39-9). Thyroid function tests normally are
lower in premature and sick newborns
P.932
than in healthy term newborn infants (Table 39-10), as a result of lower
concentrations of TBG (64,65).
Figure 39-5 Thyroid hormone synthesis. The roman numerals represent

described enzymatic defects of thyroxine synthesis.
Congenital Hypothyroidism
The causes of congenital hypothyroidism are many and include genetic
and sporadic disorders of embryogenesis, inborn errors of T4
biosynthesis, and environmental factors. Congenital hypothyroidism can
be classified into the following subgroups:
agenesis (athyrotic cretinism) or dysgenesis of the thyroid gland

(thyroid hypoplasia, thyroid ectopia)
endemic goitrous hypothyroidism
inborn errors of T4 synthesis (familial goitrous cretinism)
drug-induced hypothyroidism
thyroid hormone resistance
thyroid unresponsiveness to TSH
secondary hypothyroidism (pituitary hypothyroidism)
tertiary hypothyroidism (hypothalamic hypothyroidism).
TABLE 39-8 RANGE OF MEAN VALUES FOR THYROID AND THYROID-STIMULATING HORMONES DURING
THE NEONATAL PERIOD IN FULL-TERM INFANTS
T 4 (g/dL)
T 3 (ng/dL)
TBG (mg/dL)
Cord blood
10.9 (713)
48 (1290)
5.4 (1.29.6)
2 h of age
22.1
217
2472 h of age
17.2 (12.421.9)
125 (89256)
5.4
2 wk of age
12.9 (8.216.6)
250
5 (19)
6 wk of age
10.3 (7.914.4)
163 (114189)
4.8 (27.6)
TSH (U/mL)
9.5 (2.420)
86
7.3(<2.516.3)
2.5 (<2.56.3)
T 3 , triiodothyronine; T 4 , thyroxine; TBG, thryoxine-binding globulin; TSH, thyroid-stimulating hormone.
Agenesis or Dysgenesis of the Thyroid Gland

Disordered embryogenesis of the thyroid gland is the most frequent
cause of congenital hypothyroidism in the United States. Although
congenital endemic goitrous hypothyroidism may have been more
prevalent throughout the world at one time, the frequency of this
disorder has declined with the introduction of iodine into endemic areas.
Genetic mutations in thyroid transcription factors have been identified in
familial forms of thyroid agenesis, although athyrosis generally is a
sporadic disorder (66). Thyroid antibodies have been detected with
increased incidence among mothers of children with hypothyroidism and
may cause either transient or persistent congenital hypothyroidism (67).
Most mothers with thyroid antibodies have unaffected children, however,
and, conversely, most mothers of children with congenital
hypothyroidism do not have thyroid antibodies. Thyroid hypoplasia has
been reported in children with congenital toxoplasmosis,
P.933
but in utero infectious disorders have not been implicated commonly as a
cause of thyroid dysgenesis.
TABLE 39-9 THE UPPER LIMIT OF NORMAL FOR MEASUREMENTS OF
THYROID-STIMULATING HORMONE IN THE FIRST 5 DAYS OF LIFE a
Standard TSH Cutoff
Age at
Collection
(d)
Age-adjusted TSH
Cutoff
No. of
Infants
TSH Value
(mU/L)
No. of
Infants
TSH Value
(MU/L)
01
19
>20
>30
54
>20
26
>25
136
>20
45
>25
30
>20
30
>20
34
>20
34
>20
Total
273
143
Retrospective study of 37,927 infants with definite abnormal

results during a 6-month period (1988-1989).
TSH, thyroid-stimulating hormone.
From Allen DB, Sieger JE, Litsheim T, Duck SC. Age-adjusted

thyrotropin criteria for neonatal screening of hypothyroidism. J
Pediatr 1990;117:310, with permission.
a
Endemic Goitrous Hypothyroidism

Although endemic goiter is one of the widespread nutritional diseases in
the world, the introduction of iodine into various foods, including infant
formulas, has markedly decreased the incidence of endemic goitrous
hypothyroidism. The dietary requirements for iodine vary, but 40 to 100
mcg/day is sufficient for most children. In areas with endemic goiters,
factors other than iodine (e.g., enzymatic defects, other genetic factors,
and other dietary factors such as goitrogens), probably contribute to
goiter formation. Such factors are suggested by the evidence that
females are more commonly afflicted than males, that not everyone
within the endemic area is afflicted despite similar iodine intake, and the
variable incidence of endemic cretinism in different areas. In the Alps,
deaf-mutism is a common finding in association with endemic cretinism,
suggesting possibly an associated enzymatic defect of organification of
iodide. When cretinism occurs in conjunction with an endemic goiter, the
signs and symptoms are similar to the dysgenetic form of cretinism
except for the presence of a goiter and an elevated radioactive iodine
uptake.
TABLE 39-10 MEAN VALUES FOR THYROID AND THYROIDSTIMULATING HORMONES IN CORD BLOOD OF FULL-TERM AND
PREMATURE INFANTS
T 4 (g/dL)
T 3 (ng/dL)
TSH (U/mL)
10.9
48
9.5
35 wk of gestation
9.5
29
12.7
32 wk of gestation
7.6
15
Term
T 3 , triiodothyronine; T 4 , thyroxine; TSH, thyroid-stimulating

hormone.
Inborn Errors of Thyroxine Synthesis

The inherited disorders of T4 synthesis involve deficiencies in one or
more of the enzymes necessary for hormonogenesis or release of thyroid
hormones, resulting in hypothyroidism (69). A compensatory increase in
TSH production produces hyperplasia and enlargement of the thyroid
gland, creating the clinical picture of familial goitrous cretinism.
Iodide Trap Defect

The thyroid gland has the ability to concentrate iodide, so that the
intrathyroidal iodide concentration may be 40-fold greater than in serum.
In this rare inherited defect of T4 synthesis, this ability is lost. As would
be anticipated, the 24-hour radioactive iodine uptake is negligible.
Thyroid scans do not demonstrate the presence of thyroid tissue,
although there may be some uptake, but the gland is detected by
ultrasonography. Several other organs, including the salivary glands,
share the ability to concentrate iodide, and this defect can be
distinguished from athyrosis because the salivary iodide concentration

also is low and there usually is a goiter that can be detected by
ultrasonography. This defect essentially reflects iodine deficiency, and
can be compensated using high doses of iodide, although treatment with
thyroxine is easier and probably more effective.
Organification or Peroxidase Defect

A defect in the organification of iodide is one of the more frequent
disorders of hormonogenesis. In this defect, the thyroid has an increased
uptake of iodide but is unable to oxidize it or combine it with tyrosine.
This leads to the accumulation of free iodide in the gland that can be
discharged or released with administration of anions such as perchlorate
or thiocyanate, although iodine bound to tyrosine or thyronines cannot
be discharged. These findings have led to a simple test for the
organification defect. The patient is given a tracer amount of radioactive
iodine, which is rapidly concentrated into the gland in a patient with the
organification
P.934
defect. When the radioactivity over the gland has stabilized, a potassium
perchlorate or thiocyanate given orally (0.5-1 g) will displace the
unorganified iodine, causing a rapid discharge of the radioactive iodine
from the thyroid gland.
A variant of this form of familial goiter secondary to a defect in
organification is associated with neurosensory hearing loss (or Pendred
syndrome). The clinical picture differs from the full organification defect,
in that patients with the Pendred syndrome often have only small goiters
and the perchlorate discharge is incomplete.
Coupling Defect
Coupling of MIT and DIT into T4 and T3 is a complex intermediate step
involving several processes, and should not be considered a defined
single enzymatic deficiency. The inability of the thyroid gland to couple
MIT and DIT into T4 and T3 leads to the accumulation of large amounts
of MIT and DIT in the gland, with the small amounts of T4 and T3
synthesized being immediately released into the circulation. Radioactive
iodine uptake by the thyroid gland is rapid and high. Definitive diagnosis
requires thyroid biopsy and chromatographic analysis of the
iodotyrosines and iodothyronines. Chromatographic analysis of thyroid
gland tissue detects large amounts of MIT and DIT with only trace
quantities of T4 and T3.
Dehalogenase Defect
The deiodination of the iodotyrosines and iodothyronines occurs within
the thyroid and in the liver, kidneys, and other organs. The inherited
inability of the thyroid to deiodinate MIT and DIT causes leakage of these
precursors from the gland and depletion of iodide stores. This loss of
iodide causes decreased hormone synthesis, resulting in a compensatory
rise in TSH, thyroid hyperplasia, and increased synthesis of MIT, DIT,
and iodothyronines. The goitrous hypothyroidism in this defect is not
caused by a biosynthetic block but by iodine deficiency, which can be
treated with large amounts of iodine. However, equally efficacious and
easier is to use thyroid hormone replacement therapy. Radioactive iodine
is rapidly accumulated and turned over. Because this defect is
extrathyroidal and intrathyroidal, radioactive MIT and DIT appears
unchanged in the urine.
Abnormal Thyroglobulin
Thyroglobulin is synthesized exclusively within the thyroid. The defects
of thyroglobulin formation incorporate a group of disorders, including
errors of thyroglobulin synthesis, and decreased synthesis. Deficient
protease activity for thyroglobulin degradation also has been postulated
to result in deficiency of thyroid hormone release. These disorders are
characterized by abnormal circulating and intrathyroidal iodoproteins.
These peptides sometimes have been described as albumin-like, and
have been identified as the iodoalbumin thyroalbumin, in which the
major iodinated compounds appear to be monoiodohistadines and
diiodohistadines (70,71). The thyroglobulin abnormality is thought to
cause iodination of inappropriate proteins, mainly albumin, with a
subsequent low yield of T4. A compensatory increase in TSH secretion
causes thyroid hyperplasia and a rapid turnover of T4 or albumin.
Proteolysis of the iodohistadinethyroalbumin results in a high secretion

of iodohistadine, which can be detected in the urine.
Drug-induced Neonatal Goiter

Many drugs have been demonstrated to be goitrogenic. In the newborn
infant, the most commonly implicated drugs are iodides and thiourea
derivatives used for treatment of maternal thyrotoxicosis. The use of
these drugs not only has caused goiter in the newborn but also has been
associated with scattered reports of hypothyroidism (72). Although the
correlation between the dose of the drug and the occurrence of goiter is
poor, prolonged administration of thiourea drugs to the mother increases
the risk of fetal goiter. One recommendation for minimizing this risk is to
lower the dose of thiourea drugs during the last trimester and add
thyroid hormone concurrently (73). In infants of hyperthyroid mothers, it
is necessary to distinguish the drug-induced goiter from the TSI-induced
goiter. A low T4 suggests that the goiter is secondary to the drug,
whereas a high T4 is more compatible with a TSI-induced goiter and
maternal antibody mediated neonatal hyperthyroidism. A mixed picture
can also occur with an initially low T4 that increases within a few days as
serum concentrations of the drug decline sufficiently. The stimulating
effects of TSI antibodies, which have a much longer half-life, can
manifest as thyrotoxicosis at a few days to a week of age (discussed
under congenital thyrotoxicosis). Treatment usually is not necessary for
the infant with a drug-induced goiter unless the goiter is asphyxiating or,
more rarely, the infant is hypothyroid. Thyroid hormone will cause the
goiter to subside.
Concerns have been raised regarding the appropriate use of antithyroid
agents in the lactating mother. Propyl-thiouracil (PTU) is the preferred
drug (74), but if alternatives are needed because of adverse drug
reactions, carbimazole and methimazole are only transmitted at low
levels in breast milk and also appear to be safe (75,76). Nonetheless,
thyroid function should be monitored in the infant.
RESISTANCE TO THYROID HORMONES

Refetoff and associates reported a family with deaf-mutism, stippled
epiphyses, delayed bone age and goiter that appeared clinically
euthyroid (77). Serum T4 was elevated but thyroid hormone-binding

proteins and hormone biosynthesis were normal. This family had the
variant of thyroid hormone resistance characterized by generalized tissue
unresponsiveness to thyroid hormone. Other patients with primarily
central resistance of the hypothalamus and pituitary gland to thyroid
hormone are mildly hyperthyroid during infancy and childhood (78).
Mutations in the
P.935
isoform of the thyroid hormone receptor have been identified in a
number of families with both peripheral and central resistance to thyroid
hormones (79).
UNRESPONSIVENESS TO THYROIDSTIMULATING HORMONE

Stanbury and associates reported a severely retarded 8-year-old boy
with a normal thyroid gland, a low PBI, normal radioactive iodine uptake,
and high endogenous levels of biologically active TSH (80). Exogenous
TSH neither stimulated the thyroid gland in vivo nor increased glucose
metabolism by thyroid slices in vitro. Thyroid-stimulating hormone
unresponsiveness of the thyroid gland was postulated by these
investigators as an explanation for this clinical syndrome. Mutations in
the TSH receptor have now been identified as the cause of TSH
resistance (79). Patients can be either euthyroid or present with
congenital hypothyroidism.
SECONDARY AND TERTIARY

HYPOTHYROIDISM
Central hypothyroidism is caused by the failure of secretion of TSH and
TRH from the pituitary and hypothalamus, respectively. Newborn infants
with these disorders may be missed in newborn screening programs,
which rely on primary TSH screening or those directed to primary
hypothyroidism, by screening for both low T4 and elevated TSH levels.
All patients with the midline abnormalities outlined in Table 41-4 should
be suspected of possible secondary or tertiary hypothyroidism.
Therefore, they should have more complete thyroid studies (including a

free T4 determined by dialysis and ultra sensitive TSH level) because of
the risk that they might be missed by the newborn screening test.
Isolated TSH deficiency is rare and identification of central
hypothyroidism should lead to evaluation of other pituitary functions.
SYMPTOMS OF HYPOTHYROIDISM
Symptoms of agenesis of the thyroid gland are readily detectable by 6
weeks of age; however, some infants will have clinical manifestations at
birth or during the immediate neonatal period (81). Infants with ectopic
or residual thyroid tissue or inborn errors of T4 synthesis often will
produce enough thyroid hormone to delay the onset of clinical symptoms
and are typically asymptomatic when identified by newborn screening.
The signs during the early neonatal period are subtle and include
prolonged neonatal jaundice, mottling of the skin, poor suck, poor
feeding, lethargy, umbilical hernia, bradycardia, constipation, and
intermittent cyanosis. Occasionally infants with congenital
hypothyroidism can demonstrate severe respiratory distress. Later, the
more classic symptoms of cretinism appear. The progressive myxedema
causes coarsening of the facies, with puffy eyelids, flattened nasal
bridge, and enlarged tongue. The cry is hoarse secondary to myxedema
of the larynx and epiglottis. Lethargy, hypotonia, constipation, poor
feeding, poor weight gain, dry hair, and pallor become more notable with
time.
There is considerable evidence for the essential role of the thyroid
hormones in the growth and development of the central nervous system
(82). The final outcome of mental development in children with
congenital hypothyroidism depends on the severity and duration of
thyroid insufficiency and the time of initiation of therapy and dose of
hormone administered. The prognosis appears to be worse if signs of
hypothyroidism are clinically evident at diagnosis. Thus, a delay in
treatment until three months of age is associated with a poorer cognitive
outcome than those treated earlier (83). Klein and associates found no
differences in IQ testing or other psychometric parameters in children
with congenital hypothyroidism treated before 1 month of age as
compared to matched normal controls (84). Others have found

developmental delay even in those infants treated within 1 month of age
who had severe hypothyroidism as defined by a thyroxine concentration
less than 2 mcg/dL and retarded bone maturation (85). In general
although, the prognosis for cognitive outcome is good in those children
with congenital hypothyroidism are started on adequate doses of thyroid
hormone within the first month of life.
DIAGNOSIS
The incidence of congenital hypothyroidism has been estimated to be 1 in
4,000 births. In view of the desirability of early diagnosis and treatment,
newborn screening using filter paper spots is standard in the United
States (86). Previously, it was a common approach is to use an initial T4
screen. If the T4 is in the lowest 10% of the samples being tested or
below a specific value, a repeat T4 and a TSH are determined on the
same sample. If the repeat T4 is still low (<10 mcg/dL) or the TSH is
elevated, confirmatory tests are requested. More recently, many of the
states are using an approach to detect primary hypothyroidism,
screening for elevation of TSH. This latter approach provides better
detection of primary hypothyroidism, eliminating the infants who are
premature, euthyroid sick or have thyroxine binding protein deficiency.
The potential problems with newborn screening for congenital hypothyroidism have been reviewed (87). Following a positive screen, a serum
sample should be obtained for more specific thyroid functions. The
diagnosis should be confirmed by serum free T4, TSH, and thyroid
hormone binding index.
In those states still using T4 as a primary screen, almost 34% of low T4
values detected by newborn screening are not the result of true
hypothyroidism, but represent diminished levels of TBG in patients with
TBG deficiency or prematurity. In those cases suggestive of deficient
thyroid hormone
P.936
binding, a direct measurement of TBG should be determined and a test
such as the T3 resin uptake to evaluate for problems of thyroxine binding
other than TBG deficiency. A low T4 and a normal TSH may represent
secondary or tertiary hypothyroidism or TBG deficiency. Thyroxinebinding globulin deficiency is an X-linked disorder and occurs in 1 in
2,000 screening studies of boys. Low T4 levels also are found normally in
premature infants and severely ill newborn infants, and are not
necessarily indicative of hypothyroidism (see Euthyroid Sick Syndrome).
A T4 greater than 7 mcg/dL is regarded as normal in premature or sick
infants (65,86). Note the need to avoid using specimens obtained in the
first 24 to 48 hours after birth because of the normal surge of TSH (see
Tables 39-8 and 39-9). Radiographic skeletal age is often useful, because
50% of full-term infants with congenital hypothyroidism will not have the
osseous centers normally present at birth. It is important to perform a
thyroid scan or ultrasound on all patients with congenital hypothyroidism
to identify those patients with inborn errors of T4 synthesis so that
appropriate genetic counseling may be given.
TABLE 39-11 INCIDENCE OF VARIOUS FORMS OF CONGENITAL
HYPOTHYROIDISM
Disorder
Incidence
Congenital thyroid agenesis or dysgenesis
1:4000
Inborn errors of thyroxine synthesis
1:30,000
Hypothalamic-hypopituitary hypothyroidism
1:66,000 a
Hypothalamic-hypopituitary hypothyroidism incidence is based on
newborn screening studies so may be underdetected.
The incidence figures for various forms of congenital hypothyroidism, as

determined by the newborn screening studies, are listed in Table 39-11.
It is probable that the hypothalamic-hypopituitary forms of

hypothyroidism have been underestimated by the newborn screening
studies, because many of the TSH-deficient patients have normal serum
T4 concentrations at birth. Also, the syndromes of T4 resistance and
thyroid resistance to TSH can be missed by the newborn screening
method. Probably the most common cause of undetected congenital
hypothyroidism is omitting the screening study. This is more likely to
occur in intensive care situations, because of the magnitude of other
ongoing problems.
TREATMENT AND PROGNOSIS

The dose of thyroid hormone prescribed should be sufficient to achieve
high euthyroid levels of serum T4 within 2 weeks of starting therapy.
During the newborn period, the starting dose of l-thyroxine is 10-12
mcg/kg/day, which usually equals 37.5 mcg/day for most full-term
infants. Serum T4 normalizes before the serum TSH, therefore serum T4
should be used to guide therapy during the first 4 weeks to avoid
overdosing with L-thyroxine. After 4 to 6 weeks of therapy, however,
TSH level is the best measure for monitoring treatment. If the TSH
remains elevated, the dose of l-thyroxine should be increased.
Recently controversy has arisen over the issue of transient
hypothyroxinemia in preterm infants and neurological outcome. A
retrospective analysis of T4 levels in premature infants suggested that
severe transient hypothyroxinemia during the immediate neonatal
period was associated with problems in neurologic and mental
development at two years of age (88). None of the infants were
subsequently diagnosed as having permanent hypothyroidism. Despite
adjusting for many variables it is still not clear whether those children
with the impaired neurologic and mental outcome had lower thyroxine
levels causing the worse outcome or because these children were sicker
which resulted in the worse outcome and lower thyroxine levels. Perhaps
the answer lies in a paper, which showed that thyroid replacement in a
randomized double blind, placebo controlled study of 200 infants, less
than 30 weeks gestation did not improve developmental outcome (89).
No firm recommendations may be made at this time and further studies

are awaited.
EUTHYROID SICK SYNDROME

The euthyroid sick syndrome is a reflection of adaptive physiologic
processes that occur during acute and chronic illness. Thyroid hormones
increase basal metabolism, cardiac output, and oxygen consumption.
Reduced production of thyroid hormones, especially T3, which reduces
oxygen consumption and basal metabolic rate, is beneficial for certain
illnesses (e.g., catabolic or hypoxic conditions). In animal studies,
hypophysectomized rats survive longer during oxygen deprivation than
intact animals. The euthyroid sick syndrome has been noted in both
premature and full-term sick newborn infants.
The euthyroid sick syndrome is characterized by a low normal T4
concentration, an extraordinarily low T3 concentration, and a normal TSH
level. The latter two findings distinguish the euthyroid sick syndrome
from primary and secondary hypothyroidism. T3 levels generally are in
the low normal range in both primary and secondary hypothyroidism,
with TSH levels markedly elevated in primary hypothyroidism. Unlike
older infants and children, reverse T3 is not useful in the diagnosis of
euthyroid sick syndrome, because it is elevated normally in the newborn
infant. Euthyroid sick syndrome does not require any treatment other
than correction of the primary disease.
CONSUMPTIVE HYPOTHYROIDISM SECONDARY

TO GIANT HEMANGIOMA
Infants with large hemangioma may develop hypothyroidism, which is not
present at birth, and therefore is not detected by newborn screening
studies (90). The location of the hemangioma may be in the liver and
ectodermally.
P.937
The infants develop hypothyroidism as the hemangioma grows during
early infancy (within a few months to a year), which is detrimental to
brain development if undetected. The cause of the hypothyroidism is as a
result of markedly increased activity of the type D3 iodothyronine

deiodinase that rapidly converts T4 to reverse T3 and T3 to T2. Both
reverse T3 and T2 are biologically inactive, resulting in insufficient T4
and T3 for metabolic needs and increased TSH. This consumption of T4
and T3 ameliorates with involution of the giant hemangioma, which
sometimes need to be surgically excised. The hypothyroidism must be
treated with significant doses of synthetic thyroxine to overcome the
rapid inactivation of T4.
CONGENITAL THYROTOXICOSIS
Thyrotoxicosis in the neonatal period is relatively uncommon. Affected
infants almost always are born of mothers who have either active or a
previous history of Graves disease. Neonatal thyrotoxicosis may also
present in infants born to mothers with Hashimoto thyroiditis. Fewer
than 5% of infants born to mothers with Graves disease will have
thyrotoxicosis in the newborn period. Neonatal thyrotoxicosis is caused
by the placental transfer of maternal thyroid stimulating
immunoglobulins, which can be demonstrated in over 90% of studied
cases (91).
Neonatal thyrotoxicosis is manifested by poor weight gain or excessive
weight loss, goiter, irritability, tachycardia, flushing, and exophthalmos.
A number of these infants tend to be small for gestational age. The
infant of a thyrotoxic mother with a high normal T4 should be followed
closely. A low or suppressed TSH is further evidence of neonatal
thyrotoxicosis. The onset of symptoms usually occurs within the first
week of life, but may be delayed until the second week, particularly if
the mother has been on antithyroid drugs that can also cross the
placenta. Arrhythmias, such as paroxysmal atrial tachycardia, cardiac
failure, and rarely death have been reported with severe thyrotoxicosis
(92). In several reported cases, there has been a rapid advance in
skeletal maturation, with advanced bone age and premature closure of
the cranial sutures (93,94). Neonatal thyrotoxicosis is a self-limiting
condition and abates as the levels of TSI antibodies decrease, thus the
prognosis is good. Most infants are asymptomatic by 2 months of age
and most cases will have resolved by 9 months.
Major therapeutic concerns in neonatal thyrotoxicosis are tracheal

obstruction secondary to goitrous encroachment, and cardiac failure.
Subtotal thyroidectomy is rarely required to relieve tracheal obstruction.
Iodide (one drop every 8 hours of saturated solution of potassium
iodide), along with a -adrenergic blocking agent such as propranolol
hydrochloride, can be used to control the thyrotoxicosis. Iodide has the
advantage of interfering not only with T4 synthesis but also with the
release of thyroid hormones. In the most severe cases, digitalis,
sedation, or glucocorticoids may be necessary to prevent cardiovascular
collapse.
REFERENCES
1. Tilmann C, Capel B. Cellular and molecular pathways regulating
mammalian sex determination. Recent Prog Horm Res 2002;57:1.
2. Little M, Wells C. A clinical overview of WT1 gene mutations. Hum
Mutat 1997;9:209.
3. Achermann JC, Meeks JJ, Jameson JL. Phenotypic spectrum of
mutations in DAX-1 and SF-1. Mol Cell Endocrinol 2001;185:17.
4. Ford CE, Jones KW, Polani PE, et al. A sex-chromosome anomaly in a
case of gonadal dysgenesis. Lancet 1959;1:711.
5. Jacobs PA, Ross A. Structural abnormalities of the Y chromosome in
man. Nature 1966;210:352.
6. Sinclair AH, Berta P, Palmer MS, et al. A gene from the human sex
determining region encodes a protein with homology to a conserved DNS
binding motif. Nature 1990;346:240.
7. Clarkson MJ, Harley VR. Sex with two SOX on: SRY and SOX9 in testis
development. Trends Endocrinol Metab 2002;13:106.
8. Kwoc C, Weller PA, Guioli S et al. Mutations in SOX9, the gene
responsible for campomelic dysplasia and autosomal sex reversal. Am J
Hum Genet 1995;57:1028.
9. Jordan BK, Mohammed M, Ching ST, et al. Up-regulation of WNT-4
signaling and dosage-sensitive sex reversal in humans. Am J Hum Genet
2001;68:1102.
10. Johnston AW, Ferguson-Smith MA, Handmaker SD Jr, et al. The

triple-X syndrome: clinical, pathological and chromosomal studies in
three mentally retarded cases. Br Med J 1961;2:1047.
11. Zenteno-Ruiz JC, Kofman-Alfaro S, Mendez JP. 46, XX sex reversal.
Arch Med Res 2001;32:559.
12. Lus HA, Rudd FH. Chromosomal abnormalities in the human
population: estimation of rates based on New Haven Newborn Study.
Science 1970;169:496.
13. Carr DH, Gedeon M. Population cytogenetics in human abortuses. In:
Hook EB, Porter IH, eds. Population cytogenetics. New York: Academic
Press, 1977:1.
14. Palmer CG, Reichmann A. Chromosomal and clinical findings in 110
females with Turner syndrome. Hum Genet 1976;35:35.
15. Goldberg MD, Scully AL, Solomon IL, et al. Gonadal dysgenesis in
phenotypic female subjects. Am J Med 1968;45:529.
16. Saenger P, Albertsson Wikland K, Conway GS, et al.
Recommendations for the diagnosis and management of Turner
syndrome. J Clin Endocrinol Metab 2001;86:3061.
17. Preger L, Steinbach HL, Moskowitz P. Roentgenographic abnormalities
in phenotypic females with gonadal dysgenesis. A J Radiology
1968;104:899.
18. Quigley CA, Crowe BJ, Anglin G, et al. Growth hormone and low dose
estrogen in Turner syndrome: results of a United States multi-center trial
to near-final height. J Clin Endocirnol Metab 2002; 87:393.
19. Serhal PF, Craft IL. Oocyte donation in 61 patients. Lancet 1989;
1:1185.
20. Moshang T, Vallet HL, Cintron C, et al. Gonadal function in XO/XY or
XX/XY Turner's syndrome. J Pediatr 1972;80:460.
21. Pallister PD, Opitz JM. The Perrault syndrome: autosomal recessive
ovarian dysgenesis with facultative, non-sex-limited sensorineural
deafness. Am J Med Genet 1979;4:239.
22. Sternberg WH, Barclay DL, Kloepfer HW. Familial XY gonadal
dysgenesis. N Engl J Med 1968;278:695.
23. Krob G, Braun A, Kuhnle U. True hermaphroditism: geographical

distribution, clinical findings, chromosomes and gonadal histology. Eur J
Pediatr 1994;153:2.
24. Grumbach MM, Ducharme J, Moloshok RE. On the fetal masculinizing
action of certain oral progestins. J Clin Endocrinol Metab 1959;19:1369.
25. Murset G, Zachman M, Prader A, et al. Male external genitalia of a
girl caused by virilizing adrenal tumor in the mother. Acta Endocrinol
1970;65:627.
26. Haymond MW, Weldon VV. Female pseudohermaphroditism secondary
to a maternal virilizing tumor. J Pediatr 1973;82:682.
27. Shozu M, Akasofu K, Harada T, et al. A new cause of female
pseudohermaphroditism: placental aromatase deficiency. J Clin
Endocrinol Metab 1991;72:560.
28. Miller WL. Disorders of androgen biosynthesis. Semin Reprod Med
2002;20:205.
29. McPhaul MJ. Androgen receptor mutations and androgen insensitivity.
Mol Cell Endocrinol 2002;198:61.
P.938
30. Imperato-McGinley J. 5alpha-reductase-2 deficiency and complete

androgen insensitivity: lessons from nature. Adv Exp Med Biol
2002;511:121, discussion 131-134.
31. Quigley CA, De Bellis A, Marschke KB, et al. Androgen receptor
defects: historical, clinical, and molecular perspectives. Endocr Rev
1995;16:271.
32. Lee MM, Donahoe PK, Silverman BL, et al. Measurements of serum
mullerian inhibiting substance in the evaluation of children with
nonpalpable gonads. N Engl J Med 1997;336:1480.
33. Eberenz W, Rosenberg HK, Moshang T, et al. True hermaphroditism:
sonographic determination of ovotestes. Radiology 1991; 179:429.
34. Diamond M, Sigmundson HK. Sex reassignment at birth. Long term
review and clinical implications. Arch Pediatr Adolesc Med 1997;151:298.
35. Stanley CA. Hyperinsulinism in infants and children. Pediatr Clin
North Am 1997;44:363.
36. Parks JS, Brown MR, Hurley DL, et al. Heritable Disorders of the
Pituitary. J Clin Endocrin Metab 1999;84:4362.
37. Pfaffle R, Kim C, Otten et al. Pit-1: clinical aspects. Horm Res
1996;45[Suppl]:25.
38. Sornson MW, Wu W, Dasen JS, et al. Pituitary lineage determination
by the Prophet of Pit-1 homeodomain factor defective in Ames mouse.
Nature 1996;384:327.
39. Bach I, Rhodes SJ, Pearse RV II, et al. P-Lim, a LIM homeodomain
factor is expressed during pituitary organ and cell committment
andsynergizes with Pit-1. Proc Natl Acad Sci U S A 1995;92:2720.
40. Thomas PQ, Dattani MT, Brickman JM, et al. Heterozygous HESX1
mutations associated with isolated congenital pituitary hypoplasia and
septo-optic dysplasia. Hum Mol Genet 2001;10:39.
41. Kyllo J, Collins MM, Vetter KL, et al. Linkage of congenital isolated
adrenocorticotropic hormone deficiency to corticotropin releasing
hormone locus using sequence repeat polymorphisms. Am J Med Genetics
1996;62:262.
42. Goellner MH, Ziegler EE, Fomon SI. Urination during the first three
years of life. Nephron 1981;28:174.
43. Rees L, Brook CGD, Shaw JCL, et al. Hyponatremia in the first week
of life in preterm infants: parts I and II. Arch Dis Child 1984;59:414.
44. Judd BA, Haycock GB, Dalton N, et al. Hyponatremia in premature
babies and following surgery in older children. Acta Paediatr Scand
1987;76:385.
45. Clark AJL, Weber A. Adrenocorticotropin insensitivity syndromes.
Endocrine Rev 1998;19:828.
46. Moshang T Jr, Rosenfield RL, Bongiovanni AM, et al. Familial
glucocorticoid insufficiency. J Pediatr 1973;82:821.
47. Miller WL, Levine LS. Molecular and clinical advances in congenital
adrenal hyperplasia. J Pediatr 1987;111:1.
48. White PC, Speiser, PW. Congenital adrenal hyperplasia due to 21hydroxylase deficiency. Endocr Rev 2000;21:245.
49. Simard J, Rheaume E, Sanchez R, et al. Molecular basis of congenital
adrenal hyperplasia due to 3-hydroxysteroid dehydrogenase deficiency.
Mol Endocrinol 1993;7:716.
50. Cara J, Moshang T, Bongiovanni AM. Elevated 17

hydroxyprogesterone and testosterone in a newborn male with 3hydroxysteroid dehydrogenase deficiency. N Engl J Med 1985;313:618.
51. Auchus RJ. The genetics, pathophysiology, and management of
human deficiencies of P450c17. Endocrinol Metab Clin NA 2001; 30:101.
52. Eberlein WR, Bongiovanni AM. Defective steroidal biogenesis in
congenital adrenal hyperplasia. Pediatrics 1958;21:661.
53. Ulick S, Gautier E, Vetterik K, et al. An aldosterone biosynthetic
defect in a salt-losing disorder. J Clin Endocrinol Metab 1964;24:669.
54. Visser HK, Cost WS. A new hereditary defect in the biosynthesis of
aldosterone: urinary C 21-corticosteroid pattern in three related patients
with a salt-losing syndrome, suggesting an 18-oxidation defect. Acta
Endocrinol 1964;47:589.
55. Bose HS, Sugawara, Strauss JF III, et al. The pathophysiology and
genetics of congenital lipoid hyperplasia. N Engl J Med 1996;335:1870.
56. Pang S, Pollack MS, Marshall RN, et al. Prenatal treatment of
congenital adrenal hyperplasia due to 21-hydroxylase deficiency. N Engl
J Med 1990;322:111.
57. Speiser PW, Laforgia N, Kato K, et al. First trimester prenatal
treatment and molecular genetic diagnosis of congenital adrenal
hyperplasia (21-hydroxylase deficiency). J Clin Endocrinol Metab
1990;70:838.
58. Weiner D, Smith J, Dahlem S, et al. Serum adrenal steroid levels in
full term infants. J Pediatr 1987;110:122.
59. Lee MM, Rajagopalen L, Berg GJ, et al. Serum adrenal steroid
concentrations in premature infants. J Clin Endocrinol Metab
1989;69:1133.
60. Pang S, Wallace MA, Hofman L, et al. Worldwide experience in
newborn screening for classical congenital adrenal hyperplasia due to 21hydroxylase deficiency. Pediatrics 1988;81:866.
61. Vulsma T, Gons MH, Vijlder JJM. Maternal-fetal transfer of thyroxine
in congenital hypothyroidism due to a total organification defect or
thyroid agenesis. N Engl J Med 1989;321:13.
62. Dussault J, Row VV, Lickrish G, et al. Studies of serum
triiodothyronine concentration in maternal and cord blood: transfer of
triiodothyronine across the human placenta. J Clin Endocrinol Metab

1969;29:595.
63. Pittman CS, Chambers JB Jr, Read VH. The extrathyroidal conversion
rate of thyroxine to triiodothyronine in normal man. J Clin Invest
1971;50:1187.
64. Fisher DA. Thyroid function in premature infants. Clin Perinat
1998;25:999.
65. Adams LM, Emery JR, Clark SJ, et al. Reference ranges for newer
thyroid function tests in premature infants. J Pediatr 1995;126:122.
66. Macchia PE, De Delice M, Di Lauro R. Molecular genetics of congenital
hypothyroidism. Curr Opin Genet Dev 1999;9:289.
67. Blizzard RM, Chandler RW, Landing BH, et al. Maternal
autoimmunization to thyroid as a probable cause of athyrotic cretinism.
N Engl J Med 1960;262:327.
68. Zakaria M, McKenzie JM, Eidson MS. Transient neonatal
hypothyroidism: characterization of maternal antibodies to the
thyrotropin receptor. J Clin Endocrinol Metab 1990;70:1239.
69. de Vijlder JJ, Ris-Stalpers C, Vulsma T. Inborn errors of thyroid
hormone biosynthesis. Exp Clin Endocrinol Diabetes 1997;105:32.
70. Savoie JC, Thompoulos P, Savoie F. Studies on mono and diiodohistidine: I. The identification of histadines from thyroidal
iodoproteins and their peripheral metabolism in the normal man and rat.
J Clin Invest 1973;52:106.
71. Savoie JC, Massin JP, Savoie F. Studies on mono and diiodohistidine: II. Congenital goitrous hypothyroidism with thyroglobulin
defect and iodohistidine-rich iodoalbumin production. J Clin Invest
1973;52:116.
72. Burrow GN. Neonatal goiter after maternal propylthiouracil therapy. J
Clin Endocrinol Metab 1965;5:403.
73. Herbst AL, Selenkow JA. Hyperthyroidism during pregnancy. N Engl J
Med 1965;273:627.
74. Kampmann JP, Johansen K, Hansen JM, et al. Propylthiouracil in
human milk: revision of a dogma. Lancet 1980;1:736
75. Lamberg BA, Ikonen E, Osterlund K, et al. Antithyroid treatment of
maternal hyperthyroidism during lactation. Clin Endocrinol 1984;21:81.
76. Azizi F. Effect of methimazole treatment of maternal thyrotoxicosis

on thyroid function in breastfeeding infants. J Pediatr 1996;128:855.
77. Refetoff S, DeWind LT, DeGroot LJ. Familial syndrome combining
deaf-mutism, stippled epiphyses, goiter, and abnormally high PBI:
possible target organ refractoriness to thyroid hormone. J Clin Endocrinol
Metab 1967;27:279.
78. Bode HH, Danon M, Weintraub BD, et al. Partial target organ
resistance to thyroid hormone. J Clin Invest 1973;52:776.
79. Refetoff S, Weiss RE, Usala SJ. The syndromes of resistantance to
thyroid hormone. Endocrin Rev 1993;14:348.
80. Stanbury JB, Rocmans P, Butler UK, et al. Congenital hypothyroidism
with impaired thyroid response to thyrotropin. N Engl J Med
1968;279:1132.
81. Lowrey GH, Aster RH, Carr EA, et al. Early diagnostic criteria of
congenital hypothyroidism. Am J Dis Child 1958;96:131.
82. Oppenheimer JH, Schwartz JH. Molecular basis of thyroidhormonedependent brain development. Endocrinol Rev 1997;18: 462.
83. Klein AH, Meltzer S, Kenny FM. Improved prognosis in congenital
hypothyroidism treated before age 3 months. J Pediatr 1972;81:912.
84. New England Congenital Hypothyroidism Collaborative.
Characteristics of infantile hypothyroidism discovered on neonatal
screening. J Pediatr 1984;102:539.
P.939
85. Glorieux J, Desjardins M, Letarte J, et al. Useful parameters to

predict the eventual mental outcome of hypothyroid children. Pediatr Res
1988;24:6.
86. Committee on Genetics, American Academy of Pediatrics. Screening
for congenital deficiency of thyroid hormone. Pediatrics 1977;60:389.
87. Willi SM, Moshang T Jr. Diagnostic dilemmas: results of screening
tests for congenital hypothyroidism. Pediatr Clin North Am 1991:38;555
88. Reuss ML, Paneth N, Pinto-Martin JA, et al. The relation of transient
hypothyroxinemia in preterm infants to neurologic development at two
years of age. N Engl J Med 1996;334;821.
89. van Wassenaer AG, Kok JH, de Vijlder JJ, et al. Effects of thyroxine
supplementation on neurologic development in infants born at less than
30 weeks' gestation. N Engl J Med 1997;336:21.
90. Huang SA, Tu HM, Harney JW, et al. Severe hypothyroidism caused
by type 3 iodothyronine deiodinase in infantile hemangiomas. N Engl J
Med 2000;343:185-189.
91. Foley TP Jr, White C, New A. Juvenile Graves' disease: usefulness
and limitations of thyrotropin receptor antibody determinations. J Pediatr
1989;110:378.
92. Riopel DA, Mullins CE. Congenital thyrotoxicosis with paroxysmal
atrial tachycardia. Pediatrics 1972;50:140.
93. Farrehi C. Accelerated maturity in fetal thyrotoxicosis. Clin Pediatr
1968;7:134.
94. Hollingsworth DR, Mabry CC, Eckard JM. Hereditary aspects of
Grave's disease in infancy and childhood. J Pediatr 1972;81:446.
Chapter 40
Gastrointestinal Disease
Jon A. Vanderhoof
Terence L. Zach
Thomas E. Adrian
Although most pediatric gastroenterologists are uncomfortable with primary care
of the sick premature infant, they often are valuable consultants to the
neonatologist. In evaluating a complex gastrointestinal (GI) or hepatobiliary
problem, a gastroenterologist often uses an organ system-specific developmental
pathophysiologic approach. In looking at a problem from a somewhat different
perspective than the neonatologist, the opinion of the consultant may augment
the analysis of the primary physician. It remains the responsibility of the
neonatologist to put the consultant's view into perspective as it relates to the
other complex problems of the sick infant.
The gastroenterologist also may offer his or her skills in invasive procedures to aid
in the diagnosis of GI and liver disease. Upper and lower GI endoscopy, liver
biopsy, rectal suction biopsy, esophageal, antroduodenal, and anorectal motility
studies, and even endoscopic retrograde cholangiopancreatography can be
performed in term infants and, depending on the skill and training of the
gastroenterologist, in premature infants as well. Small diameter neonatal
endoscopes are now available which facilitate obtaining small bowel biopsies from
the distal duodenum and jejunum.
In some institutions, gastroenterologists with special expertise in nutrition provide
assistance in nutritional support of parenteral nutrition-dependent or
malnourished infants. Their role becomes especially important in infants with GI or
liver disease who may require long-term follow-up, such as the infant with
progressive liver disease, or home on parenteral nutrition, such as the infant with
short bowel syndrome.
DEVELOPMENT OF THE GASTROINTESTINAL

TRACT
Subsequent to the development of the individual organs of the GI tract,
specialized features of the system begin to become apparent, mostly in the
second and third trimesters (1). At approximately 14 weeks of gestation,
differentiation of the pancreatic endocrine and exocrine tissues begins, and crypts
and villi begin to form in the small intestine. A few weeks later, the colon, initially
populated with villi similar to those in the small intestine, begins to develop its
more characteristic surface, with gradual loss of villi. As these morphologic
changes occur, numerous functional processes begin, some of which mature early
in utero, some only at birth, and some during the first year of life.
Carbohydrate Absorption
The functional maturation of the digestive process is complex (2). There are
marked differences in maturation of the digestive and absorptive processes of
different nutrients (Table 40-1). In the neonate, most dietary carbohydrate is
presented in the form of lactose, the predominant carbohydrate in virtually all
mammalian milk. Lactose and other disaccharides are digested by enzymes
located on the brush border membrane in mature enterocytes; those located on
the distal and midportions of the small intestinal villi. Component
monosaccharides are released after hydrolysis by disaccharidases. Lactase
hydrolyzes lactose to glucose and galactose, and both subsequently are
transported by active carrier-mediated transport. Other disaccharidases include
maltase, which hydrolyzes maltose to two glucose units, glucoamylase, which
hydrolyzes glucose oligosaccharides to glucose monomers, and sucrase, which
hydrolyzes sucrose to fructose and glucose. Sucrase is actually a double enzyme,
the other part of the molecule being isomaltase, which hydrolyzes a-1-6 bonds of
a-limit dextrins. Disaccharidase activities are highest in the proximal and
midjejunum and decrease distally.
Lactase activity develops later in gestation than the other disaccharidases. Lactase
activity is low until the final weeks of gestation. Although other disaccharidase
levels can be detected somewhat earlier in gestation and reach nearly adult levels
between 26 and 34 weeks of gestational age, lactase levels are only 30% of fullterm levels by that point in gestation.Because of the delayed maturation of
lactase, specialized infant formulas for preterm infants have been designed, with a
significant percentage of carbohydrate presented as sucrose or glucose polymers
rather than lactose. Alternatively, the addition of lactase to formulas for
P.941
preterm infants may enhance weight gain (3). The predominant enzyme for
digestion of starches and glucose polymers is pancreatic amylase, which is nearly
absent during the first 4 to 6 months of life and gradually matures during the
latter half of the first year. An alternative pathway therefore must exist for the
digestion of these glucose polymers (4,5).
TABLE 40-1 DIGESTIVE AND ABSORPTIVE FUNCTION IN INFANTS
RELATIVE TO ADULTS
Process
Premature Infant
Full-term Infant
Adult
Salivary enzymes
Normal
Normal
Normal
Gastric acid production

Bile acid secretion
to normal
Normal
Pancreatic enzyme
Normal
production
Lactase production
Normal
Normal
Sucrase and isomaltase
Normal
Normal
Normal
production
Salivary glands produce an amylase that may be important in the digestion of

complex carbohydrates in the newborn. This enzyme is detectable at 20 weeks of
gestation and is present in significant quantities in premature infants. As with
pancreatic amylase, however, the ability of the newborn to secrete salivary
amylase is substantially reduced and matures throughout the first year of life.
Salivary amylase is inactivated by gastric acid, but probably retains some activity
in the stomach of premature infants. Glucoamylase is a brush border enzyme
capable of digesting glucose units from the nonreducing ends of starch and
dextrin. Glucoamylase is present in neonates and infants at 50% to 100% of adult
levels.
Finally, it is probable that some malabsorbed carbohydrate is digested in the colon
through the colon salvage pathway. Colonic anaerobic bacteria are capable of
metabolizing carbohydrates to produce short-chain fatty acids that then are
absorbed through the colonic mucosa. Considering the relative pancreatic
insufficiency and lactase deficiency in the newborn infant, the colon salvage
pathway may be an important mechanism by which infants absorb carbohydrates.
Fat Absorption
Fat absorption is a complex process, primarily because fat is insoluble in the
aqueous environment of the small intestinal lumen (6). Solubilization, therefore, is
an important part of the fat assimilation process. The first phase of fat absorption
is that of enzymatic digestion or lipolysis. Because most dietary fat is present in
the form of triglycerides, otherwise known as triacylglycerols, these first must be
hydrolyzed by pancreatic lipase.
Phospholipids are hydrolyzed concurrently by pancreatic phospholipase. Colipase,
a cofactor secreted by the pancreas, also is required, facilitating the action of
lipase by binding to bile salt-lipid surfaces and improving the interaction of lipase
with triglyceride. The efficiency of this process is augmented by the release of
cholecystokinin (CCK) from the duodenal epithelium, which occurs in response to

the presence of lipid and protein in the duodenum. Cholecystokinin stimulates
pancreatic secretion, gallbladder contraction, and simultaneous relaxation of the
sphincter of Oddi, to mix large quantities of bile acids and digestive juices with
lipids. Pancreatic lipase levels are reduced in preterm infants and intrauterine
growth-retarded infants, significantly impairing lipolysis (7). Lingual lipase,
secreted from the salivary glands, may facilitate lipolysis in the premature infant
and partially compensate for the infant's relative pancreatic insufficiency (8).
Nonetheless, fat absorption is significantly impaired in newborn infants and, to a
greater extent, in premature infants, due at least in part to pancreatic
insufficiency.
Closely linked with the process of enzymatic digestion of fats is micellar
solubilization by bile acids (9). Bile acid molecules are complex structures with
both hydrophobic and hydrophilic ends. Bile acids interface with lipids to render
them water soluble by positioning the hydrophobic portion in close proximity to
the lipid globules although allowing the hydrophilic portion to remain free to
interact with the aqueous environment. Lipids then become enclosed in discshaped water-soluble micelles that contain fatty acids, monoglycerides,
phospholipids, cholesterol, and fat-soluble vitamins.
Solubilization is particularly important because of the presence of the intestine's
unstirred water layer. This stagnant layer of water overlies the microvillus
membrane of the intestinal epithelial cells and is the primary barrier to lipid
transport. The actual thickness of the unstirred water layer is complex and difficult
to measure, but the layer is significantly reduced by the constant agitation of the
fluid in the GI tract as a result of gut motility and villus contraction. Because of
the convolutions in the small intestine caused by the presence of villi and
microvilli, the total surface area available to interface between the intestinal
surface and the unstirred layer is much greater than the interface between the
unstirred water layer and the aqueous intraluminal environment. Penetration
through the unstirred layer by the disc-shaped micelles is the rate-limiting step for
lipid absorption. Disease processes that increase the unstirred layer thickness will
markedly inhibit fat absorption in much the same manner as disease states that
render the supply of bile acids inadequate for micellar solubilization. Bile acids
commonly are deficient in cholestatic liver diseases, such as neonatal hepatitis or
biliary atresia, and in rare cases of congenital bile acid deficiency. Bile acids are
deconjugated rapidly and reabsorbed in the presence of small intestinal bacterial
overgrowth. Bile acid deficiency therefore may occur in patients with disorders
that cause intestinal stasis and bacterial overgrowth, such as short bowel
syndrome. In disorders of mucosal injury, the unstirred layer thickness may be
increased, making penetration of the fat-containing micelles difficult and further
exacerbating fat malabsorption.
P.942
Bile acids are extremely important in the fat absorption process. In the absence of
bile acids, only about one-third of dietary triglycerides, a very small percentage of
fatty acids, and virtually no cholesterol or fat-soluble vitamins are absorbed.
Medium-chain triglycerides may be better absorbed because of their enhanced
water solubility, which allows penetration of the unstirred water layer without
micellar solubilization. In both preterm and term infants, bile acid synthesis is
limited and the bile salt pool size is low (10). Moreover, preterm infants may have
an ineffective bile salt transport process in the distal ileum, resulting in impaired
enterohepatic circulation of bile salts (11). Consequently, the bile acid
concentration may be less than adequate for the formation of micelles and
solubilization of fat. Thus, penetration of the unstirred layer is less efficient in the
term infant and further impaired in the preterm infant compared to adults.
After lipids are enclosed in the bile acid micelle and reach the lipid bilayer
membrane of the small intestinal mucosal cell, absorption into the cell occurs by
passive diffusion. Because of the convolutions of the GI tract, a large surface area
exists for lipid assimilation. In the absence of disease, this process progresses in
the term and preterm infant relatively uninhibited. In disorders in which the
absorptive surface area is reduced or damaged, however, such as short bowel
syndrome or any form of diffuse enterocolitis, fat, carbohydrate, and, to a limited
degree, protein are malabsorbed.
Within the enterocyte, monoglycerides and esterified fatty acids are immediately
resynthesized to triglycerides. These triglycerides, along with apoproteins,
phospholipids, free cholesterol, some diglycerides, and esterified cholesterol, are
stabilized within chylomicrons. The outer structure of the chylomicron then fuses
with the basolateral membrane and is extruded into the lamina propria, in which it
is carried by the lacteals and lymphatic channels and deposited into the blood
stream.
Protein Absorption
The assimilation of protein begins in the stomach through the action of
hydrochloric acid and pepsin. The maturational aspects of this process have been
the subject of substantial study and some controversy. Conflicting data exist
regarding the status of acid secretion in the newborn infant. Newborn infants
appear to be capable of secreting acid, although the process is somewhat
immature (12). In premature infants, it is probable that the process is impaired to
a greater extent (13). Pepsinogen, the proenzyme for pepsin, which facilitates
protein digestion in the stomach, is secreted in preterm infants, but in much lower
concentrations than in term infants (14).
The gastric aspects of protein digestion are relatively inconsequential, compared
to the much more complete process in the small intestine. Enterokinase, produced
in the duodenal mucosa, activates the pancreatic proteolytic enzyme trypsinogen,
converting it to trypsin, which then activates essentially all of the other enzymes
involved in protein digestion. Enterokinase levels have been demonstrated in
human fetuses as early as 21 weeks of gestation (15). Enterokinase expression is
diminished during fetal development, however, and is only 10% of adult levels in
the term newborn. Additionally, pancreatic and duodenal proteolytic enzymes are
present in preterm and term infants in lower concentrations than in older children
and in adults. These enzymes initiate hydrolysis of proteins, and the hydrolysis
process is completed by brush border and cytosolic peptidases. Protein is
absorbed in the form of amino acids and dipeptides through active transport
processes that appear to be well developed by 28 weeks of gestational age.
Despite the relative immaturity of multiple phases of the protein assimilation
process, both preterm and term infants are quite capable of absorbing adequate
quantities of dietary protein. In small infants, the protein malabsorption resulting
from mucosal injury is probably far less consequential than the malabsorption of
the other major macronutrients.
Micronutrient Absorption
Absorption of micronutrients matures at varying rates in infancy. Water is
absorbed passively in response to sodium and other electrolytes, as it is in older
children and adults (16). Experimental evidence suggests that the intestinal
epithelium may be more secretory during early infancy, and the increased
susceptibility of infants to diarrheal disorders probably is at least partially related
to this process.
Mineral absorption depends on the form in which the mineral is presented to the
infant. Iron, for example, is absorbed extremely well from breast milk. Even the
preterm infant is capable of absorbing nearly 50% of the iron in breast milk,
whereas only a small percentage of iron is absorbed from cow-milk formulas,
necessitating iron supplementation. Calcium and phosphorous also are well
absorbed from breast milk (17,18). Magnesium, copper, and, to a lesser extent,
zinc are well absorbed by both term and preterm infants (19). In general,
minerals are absorbed somewhat better from breast milk than from cow milk.
Most vitamins appear to be absorbed adequately in both term and preterm
infants, although fat-soluble vitamin deficiency is common in disorders affecting
fat absorption, especially disorders causing bile acid deficiency.
Gut Motility
Although nutrient assimilation is heavily dependent on the development of
digestive and absorptive function, actual feeding depends greatly on the
maturation of gut motility (20,21,22). Neuroblasts migrate in a cranial-to-caudal
direction between weeks 5 and 12 of gestation. There is gradual maturation of gut
motility throughout the fetal period and the first several years of postnatal life. In
the fetus, normal propulsive motility in the gut probably does not appear until
approximately 30 weeks of age. Interdigestive phenomena, known as migrating
motor
P.943
complexes, can be demonstrated by approximately 33 weeks of gestation. Motor
activity in the neonatal gut differs significantly from that in adults, in that the
propagation rate of the migrating motor complex is substantially slower in
neonates, and the complex is not abolished by feeding as in older children.
Sucking and swallowing reflexes begin early during fetal development, but the
maturation of the process is not completed until after birth. The fetus is able to
swallow amniotic fluid as early as 11 to 12 weeks of gestation. Actual sucking
probably does not occur until approximately 18 to 24 weeks. This type of sucking
is termed nonnutritive sucking, differentiating it from the more effective nutritive
sucking mechanism that develops by 34 to 35 weeks of gestation. The onset of
nutritive sucking closely parallels a rapid increase in growth of the fetal stomach
(23) and the acquisition of mature patterns of gastric antral and small intestinal
motility.
By the time a term infant is born, sucking movements are followed in an orderly
progression by swallowing, esophageal peristalsis, relaxation of the lower
esophageal sphincter, and relaxation of the gastric fundus. The first stage of
swallowing is an involuntary reflex in both the term and preterm infant. Early
introduction of oral feeds may accelerate the time from gavage to full nipple
feeding (24).
Some data suggest that nonnutritive sucking may play an important role in weight
gain in preterm infants. The mechanism of this effect may be related to
maturational changes in the infant GI tract, and sucking may facilitate gastric
emptying and other GI functions, primarily through stimulation of secretion of GI
regulatory peptides.
Maturation of GI motility may have important implications for a number of
conditions. Gastroesophageal reflux is common in both term and preterm infants,
and probably relates to diminished lower esophageal sphincter function or
inappropriate relaxation of the lower esophageal sphincter, often in association
with delayed gastric emptying. The maturation of both lower esophageal sphincter
function and gastric emptying has been studied extensively, with somewhat
equivocal results. Depending on the technique used to measure sphincter
function, the lower esophageal sphincter tone has been shown to be either low or
normal in both preterm and term infants (25). Hyper-tonic carbohydrate solutions
appear to delay gastric emptying in infants, much as they do in adults.
GASTROINTESTINAL HORMONES AND ENTERIC
NEUROPEPTIDES
GI peptide hormones appear to play an important role in the structural and
functional development of the gut, and in the control of alimentary functions. The
function of a vast endocrine system is integrated with that of the enteric nervous
system, which itself uses other regulatory peptides as local messengers.
Endocrine cells producing gastrin, somatostatin, motilin, and glucose-dependent
insulinotropic peptide (GIP) are detectable in the fetus at 8 weeks of gestation,
with gastrin- and somatostatin-producing cells being most numerous (26). By 14
weeks, all of the endocrine cell types are present in the intestinal mucosa,
although the anatomic distribution is more widespread than that seen in the adult
(26). By the end of the second trimester, the distribution of gut endocrine cells
resembles that of the adult (26). Peptidergic nerves are first demonstrable in the
myenteric plexus at about 12 weeks of gestation, correlating with the known
developmental pattern of enteric nerve plexuses (26). These enteric nerves then
migrate through to the submucous plexus. By the third trimester, all of the
regulatory peptide systems are well developed (27). At birth, the molecular forms
of the GI regulatory peptides and their distribution in the gut are similar to those
of the adult (27).
Surges of gut hormones appear to be responsible for the marked growth and
functional change that occur in the alimentary tract in early neonatal life.
Substantial changes in GI hormone secretion are seen during this period,
triggered by the switch from intravenous to enteral feeding (28).
Gastrin is an important regulator of gastric secretion and is trophic to the gastric
mucosa. At birth, cord blood levels of gastrin are already four to five times higher
than those in the adult, and prefeed basal levels remain elevated for several
weeks (29,30). Furthermore, gastrin levels increase in response to the first milk
feed (31). After 3 to 4 weeks of life, basal gastrin levels decline, a change
accompanied by development of marked elevation of levels following feeding
(29,32). Gastric acid is detectable in the stomach at birth and reaches a peak in
the first day or two of life (30,33). Thereafter, acid output decreases for a period
of about 1 month despite the hypergastrinemia and rapid growth of the stomach.
It has been suggested that the lack of responsiveness to gastrin could be as a
result of a lack of receptors in the oxyntic gland mucosa. Perhaps a more likely
explanation, however, is that secretion is suppressed by an inhibitor, such as
peptide YY (PYY) or neurotensin, thus enabling gastrin to stimulate growth of the
gastric mucosa without hyperstimulation of acid secretion (34,35).
Basal levels of the duodenal hormone, secretin, are higher at birth than in adults,
and, during the first 3 weeks of life, a more marked postprandial response
develops than is seen in the adult (36). Because secretin is considered to be a
major factor in triggering the neutralization of acid chyme entering the duodenum,
the increase in circulating secretin levels may be of considerable importance in
mucosal protection during this period. It is notable that the postnatal surge of
secretin, unlike that of the other alimentary hormones, occurs even in the absence
of feeding, indicating the importance of this mucosal cytoprotective function (32).
Cholecystokinin, released from the upper small intestine, stimulates pancreatic
enzyme secretion and contracts the gallbladder. Additionally, CCK has marked
trophic effects on the pancreas and appears to be responsible for regeneration
after resection or acute pancreatitis (37). The
P.944
observed postnatal surge of plasma CCK concentrations, therefore, may be of
importance in stimulating growth of this organ (38). Interestingly, CCK levels
decrease during kangaroo care in the NICU setting, in contrast to the increase
seen with nasogastric tube feeding (39).
Also released from the small intestine, motilin is a hormonal peptide with powerful
motor functions. These motor functions include acceleration of gastric emptying
and stimulation of the interdigestive myoelectric complexes (MMC) during the
interprandial period. The motilin receptor-mediated induction of phase III of the
MMC occurs by 32 weeks of gestational age. Motilin concentrations are low in cord
blood, but preprandial basal concentrations show a massive postnatal surge that
peaks at around 2 weeks of postnatal life (32). This peak is enhanced, but
delayed, in preterm neonates. It is likely that this increase in circulating motilin
concentrations is responsible for the known increase in motor activity of the gut
that occurs during the neonatal period. Interdigestive motor complexes appear
normal at birth in the term infant, but interdigestive cycles are incomplete in
preterm neonates (40). Premature babies exhibit abnormal motor activity, with
periods of motor quiescence and nonpropagating contractions. Thus, motor
activity is more immature in preterm infants than in term infants (40). The
relationship between maturation of the migrating motor complexes and the late
postnatal surge of motilin in preterm neonates is not clear.
The jejunal hormone, GIP, is thought to be largely responsible for the postprandial
increase in circulating insulin levels (41). Basal GIP concentrations are low at birth
and increase gradually throughout the first month of life, together with the
development of a marked postfeeding GIP response similar to that seen in the
adult after ingestion of a mixed meal (32,42). The development of the GIP
response to feeding in neonates is mirrored by the postprandial insulin response,
which increases through the first month of life to maintain glucose homeostasis
(42).
Neurotensin is an ileal peptide that has inhibitory effects on gastric secretion and
motility. Plasma neurotensin concentrations are higher in the neonate than in the
adult, and an enhanced postprandial response develops in the first month of life
(43). Both reduction of gastric secretion and slowing of the rate of gastric
emptying will decrease the rate at which acid chyme enters the duodenum and,
therefore, will result in a more steady absorption of nutrients from the gut. Thus,
neurotensin may be important in the adaptation of the neonate to enteral
nutrition. PYY is an important hormone from the distal intestine that inhibits
gastric emptying and slows small bowel transit (44). PYY also inhibits gastric and
small bowel secretion, leading to an increase in net absorption (34).
Concentrations of PYY are elevated in cord blood and rise postnatally to a peak
within the first 2 weeks postpartum (35). At their peak, plasma PYY
concentrations are about 50 times higher than fasting levels in normal adults (35).
There is evidence to suggest that gastric emptying and intestinal transit are rapid
during the first week of life, both in term and preterm infants. The triggering
mechanism for the changes that then take place is unknown, but it is likely that
factors such as PYY play a role (44). Additionally, the very potent inhibitory effect
of PYY on gastric secretion may account for the prevention of hypersecretion of
acid during the early neonatal period, despite the marked hypergastrinemia (34).
A truncated form of PYY (PYY7-36 amide) is produced by proteolytic cleavage of
the 36 amino acid peptide. A recent study demonstrated that this truncated form
of PYY potently inhibits food intake.
Enteroglucagon is one of three biologically active peptides produced by
posttranslational processing of the glucagon gene product in the small and large
intestine. Glucagon-like peptides I and II (GLP-I and GLP-II) are the two other
peptides secreted in parallel with enteroglucagon, which can serve as a marker for
production of all three. GLP-I has incretin effects and physiologically enhances
insulin secretion in response to ingested nutrients in the same manner as GIP and
is also a satiety hormone. GLP-II, on the other hand, is a trophic peptide that
increases growth of the small intestinal mucosa (45,46). Plasma enteroglucagon
concentrations show a very marked postnatal surge, which peaks within the first
week and is associated with the development of a marked postprandial response
(29,32). Because an increased rate of small intestinal growth occurs in the early
neonatal period, it is likely that GLP-II is important in neonatal alimentary
maturation. The resulting mucosal growth increases the absorptive area for the
uptake of nutrients from the gut lumen.
Temporally, the postnatal surges of gut hormones parallel the changes in GI
function that accompany the introduction of enteral feeding in the infant. It is
therefore of considerable interest that these surges are not seen in infants who
have never received enteral feeding nor with nonnutritive suckling (32,47).
Concentrations of all gut hormones, with the exception of secretin, remain low in
infants receiving only parenteral nutrition (32,47).
Precise mechanisms control the secretion of each gut hormone, and the amount of
a particular peptide liberated by a meal is adequate to stimulate the appropriate
digestive response (48). For example, a meal rich in long-chain triglycerides will
evoke a large CCK response, not seen with medium-chain fats (49). The high
circulating levels of CCK in turn stimulate pancreatic enzyme secretion and, by
gallbladder contraction, release of the bile salts necessary for digestion of the longchain fat. Medium-chain triglycerides, on the other hand, are rapidly hydrolyzed
by lingual and gastric lipases; they are water soluble, do not require micelle
formation, and are absorbed rapidly. Thus, bile salts and pancreatic enzymes are
not required for digestion of medium-chain triglycerides, and a large CCK
response is not seen when they are ingested (49). The gut endocrine system, with
its sparse distribution of overlapping cell types, is designed to produce an
integrated digestive response to the discontinuous stimulation of ingested food
(48). Because the type of food presented can influence the integrated hormonal
response, it is apparent that differences in nutrition in early neonatal life may
result in
P.945
changes in the growth and functional development of the neonatal alimentary
tract.
Although the fetus makes little demand on its GI tract, the situation changes
dramatically at birth, when demand for nutrients necessitates the rapid
maturation of the alimentary tract. This development of the GI tract is
characterized by the integrated maturation of its many functions. The observation,
however, that premature infants make a satisfactory transition from intravenous
nutrition through the placenta to extrauterine enteral feeding suggests that
external influences can exert a substantial influence. The massive postnatal
surges in circulating levels of hormones, which have trophic and secretory and
motor functions, is compelling circumstantial evidence of a profound gut endocrine
influence on alimentary development (32). This is supported further by the
observation that these hormonal surges are not seen in sick infants who are on
parenteral nutrition and have not been fed orally (47). It is likely that failure of
secretion of trophic gut hormones is responsible for the hypoplastic gut and
pancreas that accompany parenteral nutrition. Appropri-ate enteral stimuli or
hormone replacement eventually may alleviate this problem. Indeed, early
minimal enteral feeding, sufficient to stimulate hormonal surges appears to have
beneficial effects without any abdominal complications.
ABNORMALITIES OF THE GASTROINTESTINAL

TRACT
To avoid repetition, an attempt has been made to confine the abnormalities
described in this section to those that might require consultation by a pediatric
gastroenterologist. Some overlap with general surgery has been allowed,
however, to avoid extensive cross-referencing (see Chapter 44).
Abdominal Wall Defects

Major defects of the abdominal wall, omphalocele and gastroschisis, occur in
approximately 1 of 6,000 live births (50,51,52). In either case, a portion of the

infant's GI tract remains outside the abdominal cavity at birth. Omphalocele is a
failure of the extraembryonic intestine to reenter the abdominal cavity through
the umbilicus, a developmental anomaly that occurs between weeks 10 and 12 of
gestational age. The defect includes the umbilicus, and the viscera typically are
covered with a peritoneal sac. Occasionally, the sac may rupture, making the
disorder difficult to distinguish clinically from gastroschisis.
Omphalocele can be an isolated defect; however, omphalocele is often associated
with other congenital anomalies or syndromes. Genetic evaluation of patients
with omphalocele may provide information regarding causation or recurrence risk.
Periconceptional multivitamin use may reduce the risk of nonsyndromic
omphalocele (53).
Gastroschisis is an actual defect in the abdominal wall that occurs lateral to the
umbilicus. The infant with gastroschisis has a normal umbilical cord not involved
in the defect. The defect usually occurs to the right of the umbilical cord. Both
omphalocele and gastroschisis may occur in the presence of other intestinal
anomalies. Malrotation is present in association with omphalocele. Although
intestinal atresias more commonly are found in gastroschisis, atresias may be
associated with either anomaly. Abdominal wall defects frequently are diagnosed
prenatally by fetal ultrasound (54). The preferred route of delivery remains
controversial and may depend on the size of the defect (55). Vaginal delivery may
be acceptable for small defects, and cesarean section may be preferred in cases
with large defects (56).
Gastroschisis or omphalocele requires immediate pediatric surgical consultation.
Fluid losses and hypothermia are of primary immediate concern, especially in the
case of gastroschisis because no membrane covers the bowel. Large fluid and
heat losses are common, and intravenous fluid replacement should be initiated
immediately. The defect should be wrapped, using warm, sterile, moist saline
gauze or a sterile transparent plastic bag. Care must be taken to prevent twisting
and infarction of the bowel. A nasogastric tube should be inserted to minimize
intestinal distention.
Postoperative management includes sedation and mechanical ventilation for at
least 48 to 72 hours because of increased intraabdominal pressure. Careful
monitoring of fluids and electrolytes is essential. High fluid intake is required (57).
During the postoperative period, gut motility is slow to return, especially in the
case of gastroschisis. Patients with gastroschisis may demonstrate sluggish
motility for up to 8 months, and a protracted course of parenteral nutrition
commonly is required. Delayed onset of necrotizing enterocolitis (NEC) is not
uncommon and should be suspected if bloody stools are observed.
Disorders of the Esophagus
Gastroesophageal Reflux
Gastroesophageal reflux is the most common esophageal disorder in the neonatal
period (58). Gastric contents normally are retained within the stomach through
the action of the lower esophageal sphincter, a zone of high pressure in the distal
esophagus that remains tonically contracted except during deglutition (59). The
anatomy of the stomach and esophagus, and their relationship to the diaphragm
and related structures, may play a secondary role in retaining gastric contents
within the stomach. Although considerable controversy exists, there is evidence to
suggest that the lower esophageal sphincter may be fully functional in the normal
full-term infant. Some evidence suggests that sphincter pressure may be
decreased, either continuously or intermittently, in infants with gastroesophageal
reflux, facilitating reflux of gastric contents into the esophagus. There is
considerable controversy over the incidence of reflux in the premature infant.
Reflux appears relatively more common, but some data suggest that the lower
esophageal sphincter
P.946
may be competent. Delayed gastric emptying and other motility problems also
may play a role in reflux in premature infants.
In adults and older children, chronic esophagitis as a result of reflux of acid into
the distal esophagus is the major concern with gastroesophageal reflux. During
the neonatal period, however, esophagitis rarely occurs. Reflux typically presents
with continual regurgitation and spitting up or vomiting of small quantities of
formula after eating. The association between reflux and apnea of prematurity
remains controversial (60,61). Recurrent aspiration during reflux episodes may
occasionally result in pneumonitis or exacerbation of preexisting neonatal
pulmonary disease. If enough formula is regurgitated, the infant may fail to
thrive. In neonates, reflux also may be associated with delayed gastric emptying.
Delayed gastric antral distention occurs in some very premature infants in the
early postnatal period (62). Such delays in antral distention could contribute to
gastroesophageal reflux and feeding intolerance commonly seen in premature
infants less than 32 weeks' gestation. As in older children, reflux is encountered
more frequently in infants with neurologic abnormalities.
Gastroesophageal reflux may exist as a primary disorder as a result of lower
esophageal sphincter incompetence or intermittent relaxation, or it may be a
manifestation of another disorder. First of all, it must be realized that
gastroesophageal reflux may occur physiologically in all infants, although not with
the frequency and severity of pathologic reflux. Any disorder that limits gastric
emptying or causes a partial proximal small intestinal obstruction, such as annular
pancreas or pyloric stenosis, will result in some gastroesophageal reflux. Small
bowel disorders, including milk protein enterocolitis or infectious enteritis, will
cause vomiting and regurgitationin essence, gastroesophageal reflux. Finally, a
variety of systemic disorders, including certain inborn errors of metabolism,

chronic infection, chronic renal disease, and increased intracranial pressure, may
result in chronic emesis similar to gastroesophageal reflux. Drugs, such as
xanthines, which may be given because of apnea or lung disease, decrease lower
esophageal sphincter pressure and may exacerbate or even cause reflux.
Several diagnostic studies are available to diagnose gastroesophageal reflux in
infants; however, these studies as a rule do not separate primary from secondary
causes. For example, an infant with milk protein enterocolitis or pyloric stenosis
will have a positive test result for gastroesophageal reflux by any of the available
studies. The most widely available test for reflux is an upper GI series, which is
preferable to a barium swallow, because the latter only examines esophageal
motility. The stomach must be filled with barium to assess the patient accurately
for reflux. Unfortunately, assessing a child for gastroesophageal reflux
radiographically lacks sensitivity because of the short time interval during which
the child is observed, and it lacks specificity because of the likelihood of
physiologic reflux occurring during performance of an upper GI series. Therefore,
the primary role of an upper GI series is to exclude gastric outlet lesions such as
pyloric stenosis, or proximal small bowel partial obstructions such as duodenal
webs or annular pancreas.
Twenty-four-hour potential of hydrogen (pH) monitoring is the most widely
accepted means of assessing gastroesophageal reflux (63). The pH probe is
placed approximately 2 cm proximal to the lower esophageal sphincter, and distal
esophageal pH is recorded over a 24-hour period. The infant must be bolus-fed
during the study, to ensure adequate gastric distention to simulate the physiologic
state. Considerable controversy exists over appropriate feeding for children during
24-hour pH monitoring. The inconsistency of acid secretion in small infants makes
the procedure much less reliable during the neonatal period, and, consequently,
simultaneous measurement of intragastric pH often is helpful in determining the
validity of the study.
A 99Tc scintiscan may be used to screen for gastroesophageal reflux, although
this technique is not considered as reliable as 24-hour pH monitoring. The
technique is useful, however, for measuring gastric emptying delay, which may
coexist with gastroesophageal reflux in a number of infants. Endoscopy with
biopsy is a useful technique for detecting reflux in older infants; however,
endoscopic biopsies are less useful during the neonatal period because pathologic
reflux has not had sufficient time to cause esophageal mucosal injury. In older
children, the presence of intraepithelial eosinophils suggests reflux, but this sign
cannot be relied on in neonates, and biopsy specimens frequently are normal.
Contrary to earlier belief, eosinophils are now recognized as being strongly
associated with allergic disease of the gut, especially if present in the esophagus
in large numbers. Most such cases however appear later in childhood.
Treatment of gastroesophageal reflux is based on the severity of symptoms. If the
child is thriving well and the major complaint is frequent regurgitation and
spitting, the infant may be placed prone on an incline at approximately 30 degrees
with the head higher than the feet. It has been demonstrated that children
positioned in this manner will reflux less frequently. Although it may take several
weeks for symptoms to resolve, the risk of esophagitis is lessened and reflux
tends to resolve more quickly. If the volume of reflux is severe and the infant is
chronically irritable, has evidence of esophagitis, or is failing to thrive, then
inhibition of gastric acid secretion with agents such as antacids or histamine-2
(H2)-receptor antagonists may be necessary. Cimetidine and ranitidine are
available in liquid preparations, and they both work well. Data suggest that
aluminum antacids may elevate serum aluminum levels in small infants (64).
Proton pump antagonists such as omeprazole are even more potent suppressors
of acid secretory activity than H2-receptor antagonists. Suppression of gastric
acidity may actually be harmful as it increases gastric colonization which could
lead to increased risk of sepsis and pneumonia (65). Bethanechol, a
parasympathomimetic agent, has been demonstrated to increase lower
esophageal sphincter resting tone and improve weight gain in infants with failure
to thrive secondary to gastroesophageal
P.947
reflux (66). Unfortunately, bethanechol may have associated central nervous
system side effects, such as irritability and sleeplessness. Metoclopramide also has
been used to treat gastroesophageal reflux in infants. The effectiveness of
metoclopramide is controversial, and it probably is most helpful when delayed
gastric emptying coexists with reflux. Some clinicians thicken infants' formula with
cereal. Although this may reduce spitting, it usually does not reduce reflux or its
complications and results in nutrient imbalance in the infant's carefully formulated
diet. Infant formulas containing rice starch as part of the carbohydrate component
have been shown to have a modest beneficial effect in gastroesophageal reflux in
infants. They have a distinct advantage over addition of rice cereal in that they
maintain an appropriate macronutrient balance.
Gastroesophageal reflux may be treated successfully at any age with surgical
fundoplication in approximately 95% of cases. The most common surgical
procedures include the Nissen fundoplication, in which the stomach is wrapped
and sutured 360 degrees around the distal esophagus, and the Thal
fundoplication, which consists of a 270-degree wrap. Complications, including
gaseous distention of the stomach and dumping syndrome, may be less common
with the Thal procedure. Many surgeons now perform fundoplications
laparoscopically. Indications for an operation for gastroesophageal reflux include
recurrent aspiration pneumonia, failure to thrive secondary to severe vomiting
unresponsive to in-hospital medical management, or apparent life-threatening
apnea events associated with gastroesophageal reflux (67). Such procedures may
be performed laparoscopically by competently trained pediatric surgeons.
Differential diagnosis of the typical neonate with chronic recurrent vomiting

includes, in addition to gastroesophageal reflux, two major categories of disease.
The first is upper GI anomalies, including pyloric stenosis. Virtually all of these can
be eliminated by upper GI contrast studies; pyloric stenosis can be excluded
adequately by ultrasonography in the hands of an experienced pediatric
ultrasonographer. The second major diagnostic category is formula protein
intolerance. Infants with formula protein intolerance commonly vomit, especially
those with significant small bowel mucosal disease. Such infants often are irritable
and usually have loose, Hematest-positive (Ames, Elkhart, IN) stools. Proctoscopic
examinations of the rectum usually demonstrate colitis. This disorder is discussed
in detail later in this chapter.
Esophageal Anomalies
The other major category of esophageal disease that presents in the neonatal
period is tracheoesophageal fistula or esophageal atresia (68). These anomalies
occur in approximately 1 in 4,000 live births. In addition to a prenatal history of
polyhydramnios, increased salivation with coughing, choking, and cyanosis shortly
after birth should raise the suspicion of tracheoesophageal fistula-esophageal
atresia. The most common variety is that of atresia with the distal esophageal
pouch connected to the trachea through a fistula. Such infants frequently have a
stomach distended with air and respiratory symptoms as a result of tracheal
aspiration of refluxed gastric acid. Immediate pediatric surgical consultation is
required (see Chapter 44).
After surgery, gastroesophageal reflux is a virtual certainty. Patients with
tracheoesophageal fistula or esophageal atresia have incompetent lower
esophageal sphincter function and aperistaltic contractions in the midesophagus.
Although swallowing usually proceeds without much difficulty, gastroesophageal
reflux with chronic esophagitis and occasionally stricture formation are frequent
long-term complications. Subsequent esophageal dilatations and fundoplication
may be necessary.
Disorders of the Stomach and Duodenum

Congenital anomalies of the upper GI tract frequently present with vomiting. The
most common is pyloric stenosis, which occurs in approximately 1 in every 500
live births (69,70). The disease is most common in Caucasian males. A positive
family history often is present. Pyloric stenosis usually presents with nonbilious
projectile vomiting during the third to fourth week of life. The disorder often is
insidious in onset. After emesis, infants are hungry and will attempt to eat to
compensate for malnutrition. It is now clear that pyloric stenosis is caused by the
selective inadequate development of inhibitory neurons in the myenteric plexus of

the pyloric region that utilize vasoactive intestinal polypeptide and nitric oxide as
neurotransmitters to relax the sphincter. Eventually, nutrition deteriorates and
infants become dehydrated and alkalotic secondary to chronic vomiting of the
acidic gastric contents. Unconju-gated hyperbilirubinemia is present in a small
percentage. Patients with pyloric stenosis have normal or firm stools, in contrast
to infants with formula protein intolerance, who usually have loose stools with
evidence of malabsorption, inflammation, or both. Serum electrolytes reveal
potassium and chloride deficiency and metabolic alkalosis. Prior to the availability
of modern imaging techniques, the failure of a silver salt added to the infant's
urine to produce a white precipitate of silver chloride was a means of
differentiating pyloric stenosis from salt-losing adrenal hyperplasia. Physical
examination demonstrates visible peristalsis in the epigastric region. Careful
palpation of the abdomen while feeding may reveal a pyloric olive. The olive can
be felt best when the stomach is empty, particularly just after vomiting. Diagnosis
usually is confirmed by an upper GI series or ultrasonography, or both in the case
of ambiguity, before proceeding with surgical intervention (71).
Before operative correction of the pyloric stenosis, infants should be rehydrated
intravenously and the electrolyte imbalance and alkalosis corrected. Surgical
correction consists of longitudinal incision of the hypertrophied muscle (i.e.,
pyloromyotomy). After the operation is completed, the patient usually can be fed
within 6 to
P.948
12 hours. Rarely, if there has been prolonged gastric distention as a result of
delayed diagnosis, time to full refeeding may be prolonged. Pyloric stenosis has
been noted to recur in rare cases. Nonoperative medical management of pyloric
stenosis, consisting of anticholinergic drugs and small frequent feedings,
occasionally may be helpful. This therapy, sometimes used in Europe, is rarely
used in North America because of the excellent results of surgical intervention.
Other rare gastric anomalies also may present in the neonatal period. Various
forms of gastric atresia or hypoplasia have been described, most of which present
with vomiting at or shortly after birth. Congenital microgastria may occur in
association with a variety of other anomalies, including limb abnormalities,
asplenia, megaesophagus, situs inversus, midgut malrotation, and cardiac
anomalies. After major reconstructive surgery of the stomach, prognosis may be
quite good.
Acid Peptic Disease

Acid peptic disease may be seen in newborn infants (72,73,74). Ulcers in children
occur most commonly in the neonatal period or during the second decade of life.
In newborn infants, ulcers, whether gastric or duodenal, usually present with
hematemesis. Occasionally, blood loss may be substantial, manifested by

symptoms of hypovolemia and shock. Differential diagnosis of hematemesis in the
newborn includes swallowed maternal blood, or blood ingested from a cracked
nipple through breast-feeding. Detection of swallowed maternal blood can be
determined by assaying for adult hemoglobin in the gastric contents with the Apt
test.
Diagnosis of peptic ulcer disease in the neonate requires endoscopy. Radiographic
studies rarely are useful because the lesions are quite superficial and difficult to
image radiographically. Endoscopy can be performed easily in a newborn infant by
a skilled pediatric endoscopist using appropriate equipment. The smallest pediatric
upper GI endoscopes can be used safely in term infants under conscious
sedation. The minimum size of the infant in whom endoscopy can be performed
safely varies with the skill of the endoscopist, but endoscopy often also can be
performed safely in larger premature babies. A bronchoscope can be used in
smaller infants, although the examination usually is unsatisfactory. Neonatal
upper GI endoscopes are now available which facilitate diagnosis significantly.
Although ulcerations may be identified anywhere in the stomach and the
duodenum, multiple superficial gastric lesions are most common in newborn
infants. Ulcers may be primary, or they may be secondary, as in the case of drugs
known to irritate the upper GI tract, such as steroids or theophylline. Treatment
with antacids, or preferably H1-receptor antagonists such as cimetidine or
ranitidine, for a period of 2 to 6 weeks results in complete healing of the lesion.
Secondary ulcers may be treated in a similar fashion. In this instance,
continuation of the offending agent requires careful assessment of the risk-tobenefit ratio, because lesions will heal more rapidly if the agents are discontinued.
Neonatal infection with Helicobacter pylori is a rare phenomenon (75).
Spontaneous gastric perforation is a rare occurrence in the newborn. It occurs
most commonly during the first 5 days of life, especially in infants subjected to
severe stress or hypoxia (76). The constellation of symptoms typically includes a
sudden deterioration in clinical status between the second and fifth day of life,
characterized by refusal to eat, vomiting, abdominal distention, and respiratory
distress. Free intraperitoneal air (usually a large volume) and fluid are
demonstrable on plain radiographs of the abdomen. Immediate surgical
consultation should be sought.
Disorders of the Small Intestine

Of the various small intestinal disorders that present in the neonatal period,
congenital anomalies that produce obstruction are likely to present earliest.
Patients present with bilious vomiting, abdominal distention, and occasionally
obstipation. Bilious vomiting indicates obstruction distal to the ampulla of Vater.

Bilious vomiting associated with the passage of blood through the rectum
suggests vascular compromise of the small intestine, necessitating immediate
surgical intervention.
Malrotation or nonrotation of the gut is an anatomic defect produced by
incomplete rotation and fixation of the embryonic intestine after return from its
extraabdominal location at about week 10 of gestation (77). During development,
the intestine rotates 270 degrees around the axis of the superior mesenteric
artery to place the cecum in the right lower quadrant. When the cecum fails to
rotate completely, the mesenteric attachment of the small intestine is limited to
that supporting the superior mesenteric artery and vein. This permits the bowel to
twist on itself and produces a midgut volvulus. In the malrotated colon, adhesive
bands, otherwise known as Ladd bands, stretch anteriorly from the right
peritoneal gutter over the duodenum, in which they can produce obstruction.
Rotational anomalies may be associated with other intestinal anomalies, usually
duodenal stenosis or atresia or other small intestinal atresias. Cardiac,
esophageal, urinary, and anal anomalies also may be present. Rotational
abnormalities should be considered in the differential diagnosis of a high intestinal
obstruction identified radiographically. Unfortunately, the diagnosis often is
missed by plain abdominal radiographs because air may be present in several
loops of bowel distal to the obstruction. Rotational abnormalities are identified
more easily with an upper GI series, or barium enema. The radiographic hallmark
of malrotation is the identification of the cecum in the upper abdomen or to the
left of the midline. Symptomatic rotational abnormalities require urgent surgical
exploration, because a volvulus may result in loss of the entire midgut within
hours of presentation as a result of vascular occlusion. Intestinal dysmotility is
common following operative repair of a malrotation with or without an associated
volvulus (78).
P.949
Jejunal or ileal atresias range from membranous obstructions to complete atresia.
Atresias can be single or multiple (79,80). An apple peel or Christmas tree
deformity of the superior mesenteric artery results in an extensive jejunal atresia
followed by multiple ileal atresias that are vascularly supplied by a branch of the
ileocolic artery. Unlike duodenal atresia, relatively few anomalies are associated
with ileal atresia. Cystic fibrosis, however, is present in approximately 20% of
infants with jejunoileal atresia.
Small intestinal atresias in the neonatal period present with bilious vomiting. The
degree of abdominal distention varies with the site of the atresia. If the atresia is
distal, vomiting may be delayed for up to 24 hours after birth. Depending on the
location of the atresia, varying numbers of dilated loops of bowel with air-fluid
levels may be present on abdominal radiographs. Because it is difficult to
differentiate small intestine from colon on plain abdominal radiographs in
newborns, a contrast enema should be performed to exclude colonic lesions and

obstructions. Contrast enemas also are helpful in excluding disorders such as
meconium plug syndrome or associated rotational abnormalities.
Meconium Ileus
Meconium ileus occurs almost exclusively in patients with cystic fibrosis. It is
caused by abnormally viscid mucus glycoprotein in meconium (81). Approximately
10% to 20% of patients with cystic fibrosis have meconium ileus as the first sign
of their disease. Pathologically, the lumen of the distal small intestine is
obstructed by an accumulation of abnormal meconium. Infants present with
bilious vomiting and abdominal distention during the first 2 days of life. A palpable
sausage-like mass may be present, and rectal examination may identify hard, dry,
gray-tan meconium. Abdominal radiographs demonstrate some evidence of
complete obstruction, but the radiologic hallmark is the soap-bubble appearance
of trapped air within the tenacious meconium in the distal small bowel. A watersoluble contrast enema occasionally is therapeutic in disrupting the meconium
obstruction. Care should be taken to avoid dehydration, because contrast
substances are hypertonic and can result in massive pooling of fluid within the
bowel lumen. Surgical intervention is required if the contrast enema is
unsuccessful. When the diagnosis of meconium ileus is made, patients should be
thoroughly evaluated for cystic fibrosis.
Other disorders related to meconium may be seen in the neonatal period.
Meconium peritonitis may occur when intrauterine bowel perforation, secondary to
obstruction, has resulted in leakage of sterile meconium into the peritoneal
cavity. Common causes include atresia, volvulus, stenosis, cystic fibrosis,
meconium ileus, and Hirschsprung's disease. Small flecks of intraabdominal
calcification may be identified radiographically. Ascites occasionally occurs, but
may resolve spontaneously unless secondary infection develops. In severe cases,
meconium peritonitis can result in adhesions that require surgical intervention.
The most serious GI disorder occurring in neonates is NEC (79,80). Because NEC
appears predominantly in sick, low-birth-weight infants, the incidence has
increased in recent years as the mortality rate for the very-low-birth-weight infant
has decreased. It has been estimated that 90% of cases occur in premature
infants and that NEC may develop in 1% to 10% of infants hospitalized in
neonatal intensive care units (84). Significant intercenter differences in the
prevalence of NEC have been reported (79). The mortality rates vary from 10% to
50%. The age of onset of NEC is related to birth weight and gestational age.
Smaller, more immature infants (less than 28 weeks of gestation) tend to have
NEC at an older age than larger, more mature (greater than 31 weeks of age)
infants (85). Thus, the more premature the infant, the longer the duration of risk.
A diagnosis of NEC in a premature infant significantly increases the patient's
length of stay and imposes added financial burden (86).
The etiology of NEC is not fully known (87). Multiple factors appear to be involved,
including hypoxia, acidosis, and hypotension, which may lead to ischemic damage
of the mucosal barrier of the small intestine (88). Secondary bacterial invasion of
the mucosa may be involved in the pathogenesis of pneumatosis intestinalis.
Moreover, NEC has been observed to occur in epidemics in neonatal intensive care
units, further supporting the role of microbial agents in pathogenesis. A number of
conditions may predispose the larger infant to development of NEC, including
cyanotic congenital heart disease, obstructive lesions of the systemic cardiac
outflow (e.g., hypoplastic left heart, coarctation of the aorta), polycythemia,
umbilical catheters, exchange transfusions, perinatal asphyxia, maternal preeclampsia, and maternal use of cocaine. Infants with patent ductus arteriosus also
seem to be at greater risk. In this case, oxygenated blood is shunted from the
intestine. All of these factors suggest that mucosal injury and ischemia are
important in the development of NEC. The role of inflammatory mediators, such as
tumor necrosis factor- and platelet-activating factor, oxygen free radicals, and
local nitric oxide synthesis also have received attention (89,90,91,92).
Rapid onset of enteral feeding may be a risk factor for NEC, because of changes in
enteric blood flow and oxygen requirements during feeding (93,94). Early
introduction of small volumes of enteral feeding appears to significantly reduce
the risk of necrotizing enterocolitis when compared to aggressive advancement of
enteral feeding in at risk preterm infants. NEC occasionally is reported in infants
who have never been enterally fed. Several factors related to enteral feeding have
been studied, and a number of theories have been proposed on how enteral
feedings might precipitate NEC. Hyperosmolar formulas have been implicated in
the production of NEC, but these formulas differed
P.950
from standard formulas in other ways as well. Additionally, most hyperosmolar
formulas have been reformulated to minimize this risk. Formula feedings seem to
predispose to NEC more than breast-feeding, suggesting that breast milk factors,
including growth factors, antibodies, and cellular immune factors, might be
protective. It also is likely that formula within the GI tract may provide a
substrate for bacterial proliferation. The role of bacterial invasion in this disease
has been well recognized, but is likely to be a secondary event after compromise
of the intestinal mucosal barrier. Enterobacter sakazakii, a rare infection in
premature infants which is occasionally associated with the feeding of powdered
formula has been seen in some infants with necrotizing enterocolitis but a causal
relationship has not been established (95).
The shunting of blood away from the intestine in a fashion similar to the diving
reflex in aquatic mammals has been postulated as a potential mechanism for
producing the initial gut ischemia. This reflex might occur in response to a hypoxic
episode and has been studied extensively in animal models.
The association of NEC with prematurity implicates immaturity of the intestinal
mucosal barrier. A number of factors that affect the mucosal barrier are immature
in premature infants, including acid output, intestinal motility, and enzyme
production. Immaturity of the microvillus membrane itself, and differences in the
mucus secreted by the small intestine, may play a significant role. The mucosal
immune system is immature, and less secretory IgA is produced. Some interest
has arisen in the possible role of oral immunoglobulin administration for
prophylaxis against NEC (96).
The reported GI hormone abnormalities in NEC patients are difficult to interpret
because of the spectrum of ages at which the disease develops, the randomness
of blood sample timing, and the variation in quantity of enteral feedings.
Concentrations of GIP, neurotensin, and enteroglucagon in infants with NEC are
lower than those normally fed infants of comparable age, but gastrin, motilin, and
pancreatic polypeptide (PP) levels appear to be normal (97).
Clinical presentations vary widely. Abdominal distention usually is one of the
earliest and most consistent clinical signs. Other symptoms include bloody stools,
apnea, bradycardia, lethargy, shock, and retention of gastric contents as a result
of poor gastric emptying. Thrombocytopenia, neutropenia, and metabolic acidosis
may develop during bowel ischemia. Not every patient has every sign, however,
and clinical presentation may vary markedly. Diagnosis is confirmed by
radiographic demonstration of pneumatosis intestinalis or portal hepatic venous
air. Nonspecific radiographic findings include thickening of the bowel wall, dilated
loops of bowel, and ascites. The presence of reducing substances in the stool, as a
result of carbohydrate malabsorption, may be an early finding in NEC, as may
increased 1-antitrypsin levels, which indicate protein-losing enteropathy.
Suspicion of NEC dictates that all enteral feedings should be discontinued. An
orogastric tube is placed routinely to relieve distention of the alimentary tract.
Intravenous access must be secured to provide fluid, electrolytes, and nutrition,
because the patient will not be fed enterally for an extended period of time.
Intravenous antibiotics are administered to provide coverage for enteric
organisms. Inclusion of specific antianaerobic agents does not appear to be helpful
(98). The duration of oral intake restriction depends on the clinical status. Patients
who merely have poor feeding with increased residuals, and the presence of
minimal radiographic findings, may be fed within 48 to 72 hours. In the presence
of pneumatosis intestinalis and marked abdominal distention, 2 weeks of
parenteral nutrition may be required before judicious gradual reintroduction of
enteral feedings is considered.
Throughout the course of the disease, frequent radiographic evaluation of the
abdomen for evidence of intestinal perforation is required. Apnea, bradycardia,
abdominal wall discoloration or edema, or a sudden increase in intraabdominal

girth should give rise to the suspicion of bowel perforation. Frequent laboratory
evaluations include a complete blood count and platelet count to look for
thrombocytopenia and neutropenia, both of which suggest deterioration. In infants
with severe inflammation of the small intestine, large volumes of fluid and
electrolytes or blood products may be required to maintain perfusion and blood
pressure. This is especially true in infants in whom severe metabolic acidosis
develops secondary to poor perfusion. Ventilatory support usually is necessary.
Exploratory laparotomy with resection of dead bowel has been the traditional
surgical approach to patients with evidence of perforation or gangrenous bowel.
Peritoneal drainage prior to laparotomy may be of benefit in extremely low birth
weight infants or hemodynamically unstable patients (99).
Infants who require surgical intervention are at risk for postoperative
complications and complications associated with total parenteral nutrition (TPN).
The most common complications after surgery for NEC are sepsis, intestinal
strictures, short bowel syndrome, and wound infections (100). Intraabdominal
abscesses are relatively rare. In a number of infants, the mucosal inflammatory
process may progress to transmural necrosis that may, if it does not lead to
perforation, result in fibroblast proliferation, granulation tissue, and stricture
formation. Some clinicians routinely study the GI tract radiographically after NEC
has been treated medically. It is not uncommon to find asymptomatic ileal
stenosis or colonic stenosis in such patients. If symptoms of partial obstruction,
such as abdominal distention, failure to thrive, or poor feeding develop in infants
who have recovered from a bout of NEC, contrast studies are indicated; however,
it should be noted that ileal stenosis may not be detected by these studies.
Localized Intestinal Perforation

Localized intestinal perforation has recently been recognized as a clinical entity
distinct from NEC (101). Localized intestinal perforations are similar to NEC in that
they
P.951
occur almost exclusively in premature infants. In contrast, patients with localized
intestinal perforations are less likely to have symptoms of a severe illness, such as
metabolic acidosis or leukopenia, than patients with NEC (101). Patients with
localized intestinal perforations are more likely to survive to discharge than
patients with NEC.
The etiology of localized intestinal perforation is not known. Interruption of
regional blood flow to the bowel has been hypothesized. Patients with localized
perforation are more likely to have received higher dosages of indo-methacin or
had an umbilical artery catheter in place shortly before the perforation than
patients with NEC (101,102).
Short Bowel Syndrome

Short bowel syndrome is defined as a malabsorptive state that occurs after bowel
resection. Infants with short bowel syndrome fall into two categories: those with
congenital anomalies (e.g., gastroschisis, apple peel anomaly of the superior
mesenteric artery, intestinal atresia) and anatomically normal patients who
undergo bowel resection for NEC. The latter group tends to have fewer
complications and a better prognosis, when equal lengths of residual small
intestine remain.
After massive resection of the small intestine, the remaining small bowel
undergoes an adaptation process characterized by epithelial hyperplasia (103).
Within 1 to 2 days after resection, enterocytes begin replicating in the crypts.
Gradual morphologic changes occur in the small intestine, including a marked
lengthening of villi that results in increased mucosal surface area. This is followed
by an increase in absorptive capacity that eventually enables many to survive
without parenteral nutrition. The adaptation process is gradual, however, and may
require weeks to years.
The major gut hormone changes seen after ileal resection are marked increases in
plasma levels of PYY, entero-glucagon, and motilin (103,104). Increases of enteroglucagon (which reflect parallel changes in the enterotropic hormone, GLP-II and
of PYY, which inhibits gastric acid secretion and small intestinal secretion,
enhances absorption and delays gastric emptying and intestinal transit, are
appropriate responses in this condition. Preservation, or even enhancement, of
the PYY response would be valuable in diminishing the rapid transit and diarrhea
associated with this condition. Studies in experimental animals have revealed that
improvements in transit and fluid absorption are temporally related to the
increase in PYY response that occurs after bowel resection (105,106).
Hypergastrinemia also accompanies intestinal resection and is responsible for the
increased gastric secretion seen postoperatively. The increase of gastrin levels is
triggered by the reduction in small intestinal inhibitory factors and usually
subsides after a few weeks, when intestinal adaptation has taken place. Mucosal
hyperplasia does not occur in the absence of enteral nutrition. In fact, mucosal
atrophy may result if the patient is nourished only parenterally (107). Enteral
nutrition stimulates intestinal adaptation by several mechanisms (108). Highly
unsaturated long-chain fats stimulate intestinal adaptation to a greater extent
than protein or carbohydrate; the mechanism by which this occurs is poorly
understood. Nonetheless, careful attention to the provision of adequate enteral
nutrition is important.
Management of the short bowel syndrome is a multistage process (109). During
the early postoperative period, use of parenteral nutrition and careful attention to
fluid and electrolyte abnormalities are essential. High-volume ostomy losses must
be replaced with a solution of comparable electrolyte content to obviate the need

for frequent changes in electrolyte concentration in parenteral nutrition solutions.
The presence of an ostomy may create additional problems. These vary
somewhat, depending on whether the ostomy is in the ileum, from which the
volume output is likely to be much greater, or in the colon, in which case stool
consistency may vary markedly based on whether the ostomy was placed
proximally or distally. If available, the services of an enterostomal therapist, with
special training in rehabilitation of infants and children with ostomies, will assist
the parents in understanding the implications of the ostomy. In general, as would
be expected, caring for an infant with a small intestinal ostomy requires
meticulous monitoring of fluid and electrolyte balance which implies equally
meticulous planning if home care is considered. The ostomy should be placed
away from sites such as the iliac crest, costal margin, or umbilicus, so that ostomy
appliances will fit easily.
Most ostomy devices consist of an adhesive, nonallergenic wafer device with a
flange that allows the transparent drainage pouch to be secured to it over the
stoma. The pouch can be removed easily for draining by snapping it off the wafer,
or it can be emptied through a drainage port at the bottom of the bag. When
leakage occurs beneath the wafer, the device should be removed, and careful skin
care around the ostomy is necessary. The skin should be washed gently with a
soft cloth moistened with mild soap and tepid water. The skin should be dried
thoroughly after cleansing. Many protective ointments and powders are available
to apply around the stoma area to prevent skin breakdown. If skin irritation is
present, tincture of benzoin or steroid preparations available in spray form may be
used. The pouches often are reusable and may be washed with mild soap and
water and soaked in a deodorant solution to control odor problems. Appropriate
teaching will prevent many of the other problems encountered by ostomy
patients, such as skin excoriation, minor stoma bleeding, stoma prolapse, and
odor. Additional ostomy problems include prolapse and stenosis. The latter should
be suspected in the presence of abdominal distention and vomiting. In either
situation, the surgeon should be notified.
When enteral feeds are started, a slow continuous infusion of an extensively
hydrolyzed protein formula generally works best. Formulations such as Alimentum
(Ross Laboratories, Columbus, OH) or Pregestimil (Mead Johnson
P.952
Laboratories, Evansville, IN), which contain significant quantities of long-chain
fats, are ideal. These lactose-free preparations with hydrolyzed protein are
absorbed rapidly, yet provide adequate stimulation of intestinal adaptation. Stool
losses must be monitored carefully. If there is significant concern about allergic
disease, an amino acid based formula such Neocate, (Scientific Hospital Supplies,
Liverpool UK) may be advantageous. A marked increase in fluid losses or
significant evidence of carbohydrate malabsorption manifested by a low stool pH
or positive stool-reducing substances are contraindications for further increasing

the enteral infusion. Enteral infusions are increased as tolerated, and parenteral
nutrition is decreased in a gradual isocaloric fashion. Patients then can be weaned
to intermittent parenteral nutrition and prepared for home parenteral nutrition
therapy (110). Intermittent parenteral nutrition allows provision of parenteral
nutrients at night, primarily for the convenience of caregivers. Parenteral nutrition
initially is discontinued only for short periods of time, usually 4 to 6 hours, each
day. In small infants, this usually is delayed until the patient can tolerate
approximately 20% of calorie intake enterally to avoid hypoglycemia when he or
she is not receiving parenteral nutrition. The duration of parenteral nutrition can
be decreased gradually until all parenteral nutrition infuses over 10 to 12 hours at
night. It is wise to taper the parenteral nutrition rates up and down when placing
a patient on or off of parenteral nutrition, to prevent fluctuation in serum glucose
levels.
Home parenteral nutrition has markedly reduced the cost of long-term
management of short bowel syndrome patients and has decreased family stresses
and nosocomial infections (111). It has become a standard therapy in patients
with short bowel syndrome, and prolonged hospitalizations rarely are necessary.
To minimize complications, careful training of home nursing personnel is essential,
and coordination of activity between physician and nursing staff must be tightly
controlled.
Continuous enteral infusions are used in patients with short bowel syndrome for
several reasons. The percentage of calories absorbed from the continuous infusion
is greater than that possible with bolus feedings, because transport carrier
proteins are continually saturated. Continuous infusion provides constant
stimulation of mucosal adaptation and reduces the need for parenteral calories,
decreasing the risk of parenteral nutrition liver disease. Children should be fed
small quantities of formula orally to learn to suck and swallow. Eventually, solids
can be fed around the nasogastric tube. These manipulations often will speed the
transition from continuous enteral to oral feeding later in the course of therapy.
Numerous chronic complications arise in the treatment of short bowel syndrome,
including bacterial overgrowth, nutritional deficiency states, watery diarrhea,
parenteral nutrition liver disease, and catheter-related problems. Bacterial
overgrowth is defined as increased bacterial content in the small intestine (112).
Complications from bacterial overgrowth include increased malabsorption, -lactic
acidosis, and colitis-like or ileitis-like syndrome. Normal small bowel bacterial
counts vary from 103/mL proximally to much greater numbers in the ileum.
Normal antegrade peristalsis and gastric and mucosal immune factors prevent
excess bacterial proliferation. In short bowel syndrome, because of disruption of
normal anatomy and motility, overgrowth is likely and bacterial counts usually
exceed 105/mL. Bacterial overgrowth should be suspected whenever motility is
slowed, the bowel is dilated, or the ileocecal valve is absent. Organisms typically
include facultative bacteria and anaerobes. Bacteria deconjugate bile salts,

causing them to be reabsorbed, depleting the bile salt pool, impairing micelle
solubilization, and resulting in steatorrhea and malabsorption of fat-soluble
vitamins. More important, bacterial overgrowth causes mucosal inflammation,
exacerbating malabsorption of all nutrients. Protein-losing enteropathy and loss of
immunoglobulins may occur. Bacteria may compete with the host for nutrients,
such as vitamin B12.
Screening for bacterial overgrowth can be done with a fasting breath hydrogen or
glucose breath hydrogen test, or by detecting the presence of indican in the urine.
Measurement of breath hydrogen is a simple test in infants and children, although
collection of samples in the neonate requires special care. A fasting breath
hydrogen level greater than 42 ppm is seen only in small bowel bacterial
overgrowth (113). A breath hydrogen level greater than 20 ppm after the
administration of 2 g/kg of oral glucose, with measurement at 15-minute intervals
after ingestion, suggests bacterial overgrowth. Measurement of indican in the
urine is a simple screening test for bacterial overgrowth, because bacteria convert
dietary tryptophan to indican. Unfortunately, indicanuria may occur in other
conditions, and the technique lacks sensitivity (114). Bacterial overgrowth can be
diagnosed definitively by culture of aspirates from the small intestine, although
this technique is difficult to interpret as contamination and culture techniques may
frequently overestimate or underestimate the extent of overgrowth. Small
intestinal biopsies demonstrating inflammatory changes suggest bacterial
overgrowth.
Accumulation of d-lactate in the blood stream results in neurologic symptoms
varying from frank disorientation to coma (115). Bacterial overgrowth may cause
colitis-like or ileitis-like syndrome with large ulcerations characteristic of Crohn's
disease, but without granulomas (116). Broad-spectrum oral antibiotics (e.g.,
metronidazole, trimethoprim-sulfamethoxazole (TMP-SMZ), gentamicin) and
antiinflammatory agents often are beneficial. Broad-spectrum antimicrobial
coverage should be directed at the organisms present, usually anaerobes.
Antimotility agents may improve nutrient contact with the mucosa by lengthening
transit time, but tend to exacerbate bacterial overgrowth and should be used with
caution. Recently, use of probiotic bacteria has been proposed for the treatment of
certain patients with small bowel bacterial overgrowth (117). However, the overall
experience with probiotic therapy in bacterial overgrowth has been somewhat
disappointing.
P.953
Secretory diarrhea may be a problem in some children with short bowel
syndrome. This may be related to hypergastrinemia, which often occurs after
resection. Because the tight junctions in the ileum are less permeable than the
jejunum, the ileum plays a major role in fluid and electrolyte conservation, and
ileal resections are more likely to result in major fluid and electrolyte losses than
jejunal resections. Because most infants with NEC have ileal disease, this is a
major problem in neonates after bowel resection. Ileal resection also results in
malabsorption of bile acids, because the ileum is the primary site for bile acid
reabsorption. Malabsorption of bile acids into the colon may cause fluid secretion
and watery diarrhea, which may respond to a bile acid-binding resin such as
cholestyramine. Unfortunately, cholestyramine may further deplete the bile acid
pool, exacerbating steatorrhea.
Nutritional deficiency states may occur after parenteral nutrition is discontinued,
including deficiencies of fat-soluble vitamins A, D, and E, and the minerals iron,
zinc, calcium, and magnesium.
Parenteral nutrition hepatobiliary tract disease is the major complication that may
result in death in infants with short bowel syndrome (111). The mechanism by
which the liver injury occurs is unknown. In most instances, enteral administration
of a significant percentage of calories, usually between 20% and 30% of total
requirements, reduces the risk of parenteral nutrition liver disease.
Cholelithiasis develops in approximately 20% of infants receiving parenteral
nutrition for short bowel syndrome because of malabsorption of bile acids, altered
bilirubin metabolism, and gallbladder stasis. Cholangitis may occur in the presence
of partial obstruction. Early cholecystectomy should be considered if patients are
symptomatic with elevated direct bilirubin and liver enzymes.
Catheter-related infections and thrombosis are common in infants requiring longterm parenteral nutrition (118). In our experience, catheter-related infections
rarely are as a result of intestinal bacterial overgrowth and most commonly are
related to catheter care technique. Diligent parental instruction in catheter care
and in the signs and symptoms of sepsis is extremely important.
During later stages of therapy, additional surgery may be indicated (119). One of
the first questions usually concerns whether to close a stoma that was formed at
the time of initial surgery. If the colon remains, and especially if ileum exists as
well, reconnecting an ostomy may substantially conserve fluid and electrolytes,
but also may result in perianal disease. In infants with dilated segments of
proximal bowel, resecting a tight anastomosis or tapering the bowel to improve
flow of luminal contents often reduces bacterial overgrowth. A number of
procedures have been designed to slow transit time, including reverse segments
of bowel, one-way valves, or colon interposition, but none is considered reliably
effective, and all may increase bacterial overgrowth.
A procedure to increase the length of the bowel has been devised that involves
transecting the bowel longitudinally, preserving the blood supply to both sides of
the bowel, and creating a segment about twice the length and one-half of the
diameter. This allows reducing the diameter of the bowel without any loss of
mucosal surface area. Because it does not actually increase the mucosal surface
area, it is indicated primarily to reduce bacterial overgrowth without losing
absorptive surface in infants with dilated bowel. Our experience has been quite
rewarding with this procedure, with 12 of 14 recent patients demonstrating
significant improvement and several becoming independent of parenteral nutrition
after surgery (120). More simplified techniques of intestinal lengthening have
recently been described. In general, these procedures should not be, performed in
neonates, because they are likely to be successful only after significant bowel
dilation has occurred.
Intestinal transplantation has now become a reality. Well over 100 patients now
have been transplanted at a number of centers in the United States and Europe
(121). Patients with short bowel syndrome with evidence of parenteral nutritioninduced liver dysfunction should be referred early for intestinal transplantation.
Recent evidence suggests that isolated intestinal transplantation prior to the
development of irreversible parenteral nutrition-induced liver disease may be an
attractive alternative to the combined liver/bowel transplantation, which has
traditionally been utilized for such patients. However, long-term survival 5 years
following intestinal transplantation has only been in the range of 50%.
It is possible for infants to survive without transplantation or permanent
parenteral nutrition with surprisingly short segments of bowel (122,123). As a
general rule, patients with greater than 25 cm of normal bowel at the time of
neonatal resection who have an ileocecal valve, or with greater than normal 40
cm of normal bowel at the time of neonatal resection who have no ileocecal valve,
have a reasonable chance of eventually becoming independent of parenteral
nutrition. The ileocecal valve appears to play a major role in determining the longterm prognosis, primarily because of its ability to exclude colonic bacteria from
entering the small bowel and perhaps also because of its ability to delay transit
through the small intestine.
Mucosal Injury Disorders

Because of the limited small intestinal reserve in small infants, small intestinal
disease perhaps is most catastrophic in infancy. In small intestinal injury, all
nutrients are malabsorbed. Most symptoms, however, are related to carbohydrate
malabsorption because of the osmotic diarrhea produced when these malabsorbed
molecules are broken down further by intestinal bacteria into smaller and smaller
osmotically active particles. The osmotic gradient overrides the ability of the ileum
and colon to reabsorb fluid effectively, and watery diarrhea ensues.
Measurement of stool pH and reducing substances is an ideal means to screen for
small bowel mucosal disease in infancy. When carbohydrates are malabsorbed and
broken down into organic acids by colonic bacteria, the stool pH drops below 5.5.
Stool pH can be measured easily with
P.954
litmus paper, simply by inserting the paper into the stool. Measurement of
reducing substances in stool can be done by placing five drops of stool and ten
drops of water into a test tube and dropping in a Clinitest (Ames) tablet. Positive
reducing substances in stool confirms the presence of carbohydrate
malabsorption. Patients receiving formulas that are predominantly sucrose are
less likely to demonstrate positive reducing substances in their stools because
sucrose is a nonreducing carbohydrate.
Infectious Diarrhea
During the neonatal period, infectious diseases of the small intestine are relatively
uncommon. A number of viruses may cause diarrhea in small infants, including
rotavirus, enteric adenoviruses, and enteroviruses. Viral gastroenteritis usually
presents with watery stools, with evidence of carbohydrate malabsorption. The
predominant mucosa injury in viral gastroenteritis is in the proximal jejunum, in
which carbohydrates are absorbed. In contrast, bacterial pathogens generally
produce more distal injury that involves the colon and results in Hematest-positive
stools that contain leukocytes (Table 40-2). Bacterial causes of diarrhea include
Salmonella sp, Shigella sp, invasive Escherichia coli, and Campylobacter jejuni.
Clostridium difficile infection predominantly involves the large intestine and, in
severe cases, produces pseudomembranous colitis. Infection with C. difficile
usually follows a course of broad-spectrum antibiotics. Severe watery or bloody
diarrhea and colonic perforation may occur. Diagnosis is difficult in neonates,
because a very high percentage of small infants carry C. difficile without evidence
of disease.
Human Immunodeficiency Virus-Associated

Disease
Human immunodeficiency virus (HIV) infection in infants results in a variety of GI
problems. Failure to thrive is common. Chronic diarrhea and generalized
lymphadenopathy often are present. Other common presentations include
asymptomatic hepatosplenomegaly, which may occur in conjunction with severe
interstitial pneumonia and hypergammaglobulinemia.
TABLE 40-2 SCREENING STOOL STUDIES IN INFECTIOUS DIARRHEA
Bacterial
Viral
pH
5.5
5.5
Hematest
+++
Clinitest
Leukocytes
+++
, negative; , negative or positive, +++, strongly positive.
Chronic diarrhea in infants with acquired immunodeficiency syndrome (AIDS)

presents a very difficult management problem. Diarrhea in older children may
result from opportunistic infections, tumors, including Kaposi's sarcoma and
lymphoma, and direct HIV infection of the gut. Opportunistic organisms include
viral agents (e.g., cytomegalovirus, rotavirus, herpes simplex, coxsackievirus,
adenovirus), bacterial pathogens (e.g., Salmonella sp, Campylobacter-like
organisms, Listeria sp, Mycobacterium avium-intracellulare, Plesiomonas
shigelloides), fungal pathogens (e.g., Candida sp, Aspergillus sp), and parasitic
pathogens (e.g., cryptosporidium, Strongyloides sp, Giardia sp, amoebas,
Isospora belli). A broad spectrum of endoscopic and histologic findings is possible
because of the diverse nature of the disease. Extensive viral and bacterial cultures
and examinations for ova and parasites are warranted in infants and children with
AIDS.
Treatment consists of therapy directed at any specific infectious pathogen
identified, coupled with use of parenteral and enteral nutrition. As in almost all
chronic enteropathies, primary attention should be given to continuous enteral
infusion with an elemental diet or protein hydrolysate formula. If malabsorption
develops in the patient, as indicated by stools testing positive for reducing
substances or manifesting a pH below 5.5, supplemental parenteral nutrition is
needed to provide the remainder of caloric and other nutritional needs. The
presence of a significant secretory component of the diarrhea may make
treatment difficult, complicate the use of continuous enteral infusion, and require
careful attention to intake and output of fluid and electrolytes to maintain
biochemical homeostasis.
Hormonal Changes in Diarrhea

Infective diarrhea in infants is associated with a massive increase in circulating
concentrations of motilin, entero-glucagon, and PYY (124,125). These
abnormalities resolve once the patients are better. The gut endocrine system in
infants, however, responds in a manner different from that of adults. In infants

with diarrhea, plasma motilin concentrations exceed those that are known to
accelerate gastric emptying and increase small bowel motility (125). Motilin is
therefore likely to be involved in the motor abnormalities associated with this
condition. A hormonally triggered increase in transit rate may constitute a defense
mechanisms to rid the bowel of pathogens and secreted toxins. The extremely
high enteroglucagon levels in neonatal infective diarrhea appear to be related to
the extent of mucosal injury and its repair (126). Measurement of selected gut
hormones may give information on the extent of mucosal damage or the presence
of ongoing pathologic change.
Formula Protein Intolerance

One of the most common causes of chronic diarrhea in small infants is formula
protein-induced enterocolitis
P.955
(127,128). Small bowel involvement (i.e., enteritis), colon involvement (i.e.,
colitis), or concomitant small bowel and colon involvement are possible with this
disorder. Protein-induced mucosal injury has been reported with cow-milk protein,
soy protein, breast milk, and even beef protein. A number of infants who are
intolerant of protein hydro-lysate formulas, have responded well to an amino acidbased infant formula, Neocate (129).
Infants typically present before 3 months of age with bright red blood in the stool,
which may be normal or loose in consistency depending on the extent of
inflammation. Sigmoidoscopic examination may reveal gross friability or may
appear normal until the mucosa is wiped with a cotton swab, at which time the
mucosa readily bleeds. Rectal biopsy confirms inflammation, with the combination
of polymorphonuclear leukocytes and eosinophils in the lamina propria. Stools
may test positive for occult blood and may contain leukocytes. If the inflammation
extends into the small intestine, carbohydrate malabsorption may occur, infants
will have watery diarrhea, and stools will be acidic (pH less than 5.5) and test
positive for reducing substances. Small intestinal biopsy will demonstrate varying
degrees of mucosal injury, with shortening and blunting of the villi, inflammatory
infiltrates, and an increase in mitotic activity in the crypts. Disaccharidase levels in
the mucosal biopsies often are reduced.
Other symptoms that may occur as a result of formula protein-induced
enterocolitis are vomiting and irritability. Infants with formula protein intolerance
commonly vomit, especially those with small intestinal disease. It often is difficult
to differentiate this disorder from gastroesophageal reflux, because both groups
may present with irritability; however, babies with formula protein sensitivity
commonly have loose stools and abnormal sigmoidoscopic examinations. This is
an important distinction because the treatment is vastly different. Irritability, not
dissimilar from infantile colic, may occur in infants with formula protein
intolerance. In infantile colic, the irritability classically occurs at a specific time of
the day and responds symptomatically to repetitive stimuli. Infants with irritability
secondary to formula protein intolerance usually are inconsolably irritable, often
feed poorly, have loose stools, spit up, and have abnormal sigmoidoscopic
examinations or other evidence of small intestinal or colonic inflammation. A
careful history and physical examination and appropriate laboratory studies can
be quite specific in differentiating the two disorders.
Infants who manifest signs and symptoms of formula protein intolerance should
be placed on a protein hydro-lysate formula such as Nutramigen, Pregestimil, or
Ali-mentum, because a high percentage also will be intolerant of soy formula. A
small percentage of infants who do not respond to these formulas may improve on
an amino acid formulation such as Neocate.
Most infants with formula protein intolerance will outgrow their sensitivity by 1
year of age. Powell (130) has described a specific challenge procedure to confirm
the diagnosis of cow-milk protein intolerance. Patients should not have received
the suspected antigen for at least 2 weeks before testing and should be
asymptomatic. A standard dose (100 mL of cow milk or soy formula) is
administered and the child is monitored carefully for reaction. Stool specimens are
analyzed for occult blood, leukocytes, and reducing sugars, and a leukocyte count
is obtained 6 to 8 hours later. The test result is considered positive if diarrhea
develops within 24 hours, leukocytes or blood appear in stools, or the leukocyte
count rises by more than 4,000 cells/mL over baseline. Approximately 20% of the
infants have a delayed reaction. If the child originally had evidence of severe milk
protein sensitivity, it is wise to hospitalize the infant, start with small volumes (5
to 10 mL) of formula, and gradually increase the volume to avoid severe mucosal
injury, anaphylaxis, and shock. Skin testing rarely is useful in children with
formula protein intolerance who have predominantly GI symptoms.
Intractable Diarrhea of Infancy

A state of persistent diarrhea and malabsorption despite the institution of a
protein hydrolysate formula and in the absence of infectious pathogens is referred
to as intrac-table or protracted diarrhea of infancy (131). These patients
demonstrate a variety of histologic abnormalities on small intestinal biopsy,
including blunting or flattening of the villi, increased mononuclear cells with
occasional polymorphonuclear infiltrates, cuboidalization of the surface epithelium,
and a mild-to-moderate increase in mitotic activity in the crypts. Histologic lesions
vary substantially, however, and correlate poorly with ultimate prognosis (132).
Such infants have chronic weight loss and progressive malnutrition unless
appropriate therapy is instituted. Initial therapy involves slow institution of a
continuous enteral infusion of a diluted extensively hydrolyzed protein formula
such as Pregestimil or Alimentum (133). The rate is rapidly advanced to
approximately 150 mL/kg/d or more, provided the child is not on supplemental

parenteral nutrition. The concentration of the formula then is advanced
sequentially over 3 to 4 days, until the patient is tolerating full caloric
requirements and gaining weight. During this time, stool pH and reducing
substances are monitored, and evidence of carbohydrate malabsorption suggests
failure of enteral feedings. If this occurs, a period of parenteral nutrition is
indicated, usually through a central venous catheter, with gradual reintroduction
of enteral feedings by continuous infusion. In treating such infants, it is important
to avoid using formulas that contain intact milk or soy protein, because the
likelihood of protein sensitivity in such infants is high and further mucosal injury
may result. Lengthy periods of parenteral or continuous enteral infusion may be
necessary, but the ultimate prognosis is good.
In addition to intractable (i.e., protracted) diarrhea associated with formula
protein intolerance, other rare syndromes have been reported to result in mucosal
injury and chronic diarrhea. One is congenital microvillus atrophy or microvillus
inclusion disease, a disorder with
P.956
hypoplastic atrophy of the villi and shortening or depletion of microvilli (134). This
disorder requires electron microscopic diagnosis and has a very poor prognosis. A
similar disorder has been recognized electron microscopically and has been
associated with tufting of the microvilli, hence, the term tufting
enteropathy (135). A number of patients also have been described with an
intractable diarrhea-like syndrome in association with a mild immune defect, stiff
unmanageable hair, and peculiar facies (136). All of these conditions are quite
rare. Severe protracted diarrhea also may be associated with autoimmune
enteropathy; however, this disorder usually presents outside the immediate
neonatal period (137). Mucosal lesions are severe and the prognosis is poor.
Infants with this disorder often have multiorgan autoantibodies, including gut
epithelium, and frequently have pancreatic involvement with hyperglycemia or
hypoglycemia.
Disorders of the Colon

There is substantial overlap in small bowel and colonic disease in neonates. Many
of the congenital anomalies involve both the small and large intestines, and
certain mucosal injury disorders, including formula protein enterocolitis and NEC,
can involve both the small intestine and colon. There are some disorders that
primarily affect the colon. Most are congenital and involve anatomic obstructions,
such as atresias, or dysmotility, such as Hirschsprung's disease.
Anatomic Lesions
Colonic stenosis or atresia is a rare event, often associated with other skeletal
anomalies. Colonic duplication also is a rare entity, which may present with
delayed symptoms of obstruction. Duplications usually are cystic, gradually
enlarging masses, located posterior to the rectum, which may be confused with
tumors (138).
Motility Disorders
More frequent are the disorders that present with delayed passage of meconium
secondary to dysmotility. Meconium plug syndrome is one such entity, in which
inspissated meconium in the distal colon results in obstruction and dilatation
proximally. Delayed passage of meconium is the presenting symptom, and barium
enema examination reveals a large plug of meconium that often is evacuated after
the barium enema. Normal feeding and stooling usually follows removal of the
obstruction, but 20% to 30% of patients with meconium plug syndrome have
Hirsch-sprung's disease. If symptoms recur after removal of the meconium plug,
rectal suction biopsy is indicated.
Delayed passage of meconium also may occur with the neonatal small left colon
syndrome. Radiographic examination of these infants demonstrates normal-todilated proximal colon with constricted or smaller distal colon, with the constricted
area usually beginning around the splenic flexure. The line of demarcation is much
more abrupt than is seen in neonatal Hirschsprung's disease. The disorder is more
common in infants of diabetic mothers (139). It usually resolves spontaneously,
although placement of a colostomy may be necessary until normal motility
returns. Colonic motility eventually will return, usually within 2 to 12 weeks, and
the colostomy may be closed at that time.
Hirschsprung's disease, or congenital aganglionic megacolon, occurs in
approximately 1 in 5,000 live births, more commonly in males than in females
(140). The risk of recurrence in families is reported to be as high as 10%; higher
in infants with total aganglionosis. The frequency is ten times higher in infants
with trisomy 21. The disease is caused by a congenital absence of the ganglion
cells in both the submucous and myenteric plexuses. Ganglion cells regulate
normal colonic peristaltic activity. The absence of ganglion cells results in an
inability of the bowel to undergo coordinated relaxation. Impaired migration of
neural crest cells into the distal colon is thought to be the mechanism through
which Hirschsprung's disease develops, although there is some controversy about
this. The disorder almost always involves the distal rectum, but the extent varies
substantially. There also is controversy regarding whether or not skip areas can
occur. A few such cases have been reported, but they appear to be extremely
rare. In most instances, involvement does not extend proximal to the sigmoid
colon. In very rare instances, the involvement may extend beyond the colon into
the small intestine. The further the lesion extends, the more difficult the medical
management becomes.
Most cases of Hirschsprung's disease are not diagnosed in the neonatal period.
When they are, the most common clinical presentation is delayed passage of
meconium, with passage of the first stool beyond 24 hours of age. This
presentation probably is common but often overlooked. Infants also may appear
irritable, with poor feeding and failure to thrive, which, unfortunately, is the
typical presentation of a wide variety of small bowel and colonic disorders.
Some infants with Hirschsprung's disease may present with a life-threatening
complicationacute enterocolitis (141). Toxic megacolon is common. Although
enterocolitis may occur in the newborn period, it more commonly presents at 2 to
3 months of age. Mortality remains around 50%. The disorder presents with
sudden or gradual onset of diarrhea, followed by bloody stools and eventually the
clinical appearance of sepsis. The clinical overlap between infectious enterocolitis
or formula protein-induced enterocolitis is such that Hirschsprung's enterocolitis
also must be considered in the differential diagnosis of these more common
entities. Patients who present with bloody diarrhea in infancy and have negative
stool cultures, and who do not respond quickly to protein hydrolysate formula,
need to have a rectal biopsy performed. If Hirschsprung's disease is suspected,
surgical consultation should be obtained immediately, and attempts should be
made
P.957
to decompress the colon with a rectal tube or rectal irrigation. Controversy
continues in the surgical literature regarding what is the best operation for
Hirschsprung's disease (142).
Diagnosis of Hirschsprung's disease usually rests with rectal suction biopsy. A
small biopsy tube is inserted into the rectum and a small piece of tissue is
removed from a point 2 cm proximal to the mucocutaneous junction. If the biopsy
is obtained higher, patients with low-segment Hirschsprung's disease may be
missed. If the biopsy is taken more distally, it will be obtained in the
hypoganglionic zone, an area in which ganglion cells are normally sparse,
resulting in a false-positive biopsy for Hirschsprung's disease. The biopsy must be
deep enough to contain sufficient submucosa to identify ganglion cells. Superficial
biopsies are inadequate to diagnose Hirschsprung's disease. Because ganglion
cells are sparse, the biopsy must be serially sectioned, and 60 to 80 sections of
tissue examined. Ganglion cells in newborns are somewhat immature and difficult
to identify. Thus, a rectal suction biopsy is a reliable diagnostic tool, provided the
biopsy is obtained from an appropriate location and depth and an experienced
gastroenterologist or pathologist interprets the biopsy. If results are equivocal, a
full-thickness biopsy may be performed to establish the diagnosis.
Diagnosis also can be made by inflating a balloon in the distal rectum and
measuring relaxation of the internal anal sphincter, a process impaired in

Hirschsprung's disease. Although this technique is performed less commonly,
those who are skillful in its use believe it is as reliable as rectal biopsy. In
contrast, barium enema examination in the newborn is highly unreliable in the
diagnosis of Hirschsprung's disease because the transition zone has yet to
develop. Therefore, proximal colonic dilation usually is not apparent in the
neonate, and the clinician must look for irregular contractions in the rectosigmoid
as the primary hallmark of Hirschsprung's disease.
Treatment begins with placement of a decompressing colostomy proximal to the
transition zone between ganglionic and aganglionic bowel. Definitive surgery
usually is done at 8 to 12 months of age. A number of different operations have
been devised in which the aganglionic bowel is removed and the ganglionic bowel
attached into the distal rectum. Surgical treatment generally is successful in
restoring long-term fecal continence (143).
Patients with total colonic Hirschsprung's disease present major difficulties in
postoperative fluid and electrolyte balance. Infants frequently require prolonged
courses of parenteral nutrition. Parents should be warned about the protracted
nature of the disease, and infants should be observed closely for fluid, electrolyte,
and nutritional problems.
Several other intestinal motility disorders present in the neonatal period. Transient
hypomotility occurs in some premature infants and is characterized by markedly
delayed gastric emptying and absent or diminished small bowel motility. In most
instances, these abnormalities gradually resolve with time; support with
parenteral nutrition coupled with intermittent attempts at feeding is all that is
indicated. Alterations in calcium and magnesium metabolism such as parathyroid
dysfunction or hypothyroidism can cause diminished motility, and these disorders
should be excluded in infants with apparent motility disorders. Occasionally,
infants with chronic idiopathic intestinal pseudoobstruction syndrome can present
during infancy. This term applies to a number of neuropathic and myopathic
disorders that result in chronic progressive GI hypomotility. A variety of histologic
lesions have been described, and the prognosis for improvement is poor.
Occasionally, the disorder also may involve the urinary tract, with associated
megacystis and megaureter (144).
Pancreatic Disorders
Cystic Fibrosis
Disorders of the pancreas uncommonly present during the neonatal period. The
most common is cystic fibrosis, which occurs in approximately 1 in 1,600
Caucasian live births (145). This autosomal recessive disorder usually presents
later in childhood with failure to thrive or chronic pulmonary disease, but may
present with meconium ileus in the neonatal period. After the obstruction has
been relieved, therapy consists of compensating for the pancreatic insufficiency
through use of extensively hydrolyzed protein formulas such as Pregestimil or
Alimentum, which contain -medium-chain triglycerides as part of their lipid
component. Medium-chain triglycerides do not require digestion by pancreatic
enzymes for absorption and, consequently, may facilitate nutrient assimilation in
infants with pancreatic insufficiency. Despite the use of elemental formulas,
replacement pancreatic enzyme therapy from birth is necessary to aid the
digestion of endogenously secreted proteins.
In children with pancreatic insufficiency as a result of cystic fibrosis, the release of
the pancreatic hormone polypeptide (PP) is almost totally abolished. Fasting PP
levels are low and the normal response to milk feeding is absent (146). Plasma
insulin and GIP responses are reduced significantly in cystic fibrosis patients
compared to control subjects, even though the early glucose rise is greater in the
former group (141). Reduced GIP secretion in response to feeding may
exacerbate the glucose intolerance that accompanies pancreatic destruction.
Although plasma enteroglucagon concentrations are elevated in cystic fibrosis,
levels of other hormones such as gastrin, secretin, motilin, and glucagon are quite
normal (146).
The second most common cause of pancreatic insufficiency in infancy is
Shwachman syndrome, an autosomal recessive disorder characterized by
pancreatic insufficiency and bone marrow dysfunction with cyclic neutropenia
(145,148). This rare disorder should be considered in infants with steatorrhea and
neutropenia. Extremely rare isolated defects in pancreatic enzyme secretion,
including trypsinogen and lipase, also have been reported.
P.958
Liver Disease
Development
The liver is a complex organ serving multiple metabolic functions. From a
digestive standpoint, its primary function is that of an exocrine organ producing
bile for the emulsification of fats. Postnatally, the liver receives its blood from two
separate sources, approximately 25% from the hepatic artery and 75% from the
portal vein. The portal vein drains the splanchnic bed and allows the liver the
opportunity to regulate and metabolize substances absorbed by the intestine and
hormones produced in the GI tract.
Bile is composed primarily of water. The concentration of solids in the bile are
increased threefold by the gallbladder. The fetal liver is capable of synthesizing
bile acids from cholesterol slowly, and the rate of synthesis increases
progressively throughout gestation. The major bile salt in newborns is

taurocholate. Conjugation of bile salts with glycine in preference to taurine
gradually increases, and, by adulthood, most bile salts are conjugated with
glycine. The bile acid pool is very small in the preterm infant, but gradually
increases in the newborn and matures throughout infancy. This relatively small
bile acid pool results in reduced bile salt secretion and, in addition to the relative
pancreatic insufficiency, plays a role in the less efficient absorption of fat in the
newborn infant. Bile acids are reabsorbed in the ileum through an active transport
mechanism. There is some passive transport of bile acids in the jejunum and
colon as well. In the fetus, taurocholate is absorbed passively and active ileal
transport appears after birth.
Cholestatic Liver Disease

Most liver diseases in the neonatal period present with cholestasis, or conjugated
hyperbilirubinemia (149,150). Although elevation of the amino transferase (i.e.,
transaminase) enzymes is considered the hallmark of hepatocellular injury in older
children and adults, neonates may have significant hepatocellular injury even in
the presence of normal amino transferase levels. There are many causes of
prolonged conjugated hyperbilirubinemia in the neonatal period. These can be
subdivided into general categories, including the following: infectious disorders;
toxic insults such as parenteral nutrition and sepsis; metabolic disorders;
anatomic disorders, including congenital hepatic fibrosis and choledochal cyst; and
idiopathic infantile cholangiopathies, including primarily biliary atresia and
neonatal hepatitis (151).
Neonatal Hepatitis and Biliary Atresia

After extensive evaluation of the infant with cholestasis, a diagnosis of either
extrahepatic biliary atresia or idiopathic neonatal hepatitis is made in 70% to 80%
of cases (152, 153). Current thinking suggests that neonatal hepatitis and biliary
atresia form a continuum of a pathophysiologic process directed at various levels
of the hepatobiliary tract. Inflammation in the bile duct epithelium may result in
sclerosis and obliteration of the bile ducts and manifest itself as extrahepatic
biliary atresia. Primary hepatocellular inflammation is more likely to result in
neonatal hepatitis. Reovirus type III has been found in a number of infants with
both idiopathic neonatal hepatitis and biliary atresia. This virus has been
implicated as a causative agent in both disorders, although there is evidence that
raises questions about this hypothesis (154,155,156).
Idiopathic neonatal hepatitis is slightly more common than biliary atresia. Both
sporadic and familial varieties exist. Neonatal hepatitis, unlike biliary atresia, is
more common in low-birth-weight infants. Jaundice develops during the first week
of life in most cases. A wide variety of clinical presentations may occur, from
severe failure to thrive or fulminant hepatic failure to asymptomatic jaundice.

Acholic stools are uncommon, but may occur if cholestasis is severe. Physical
examination reveals a firm, enlarged liver and occasionally splenomegaly. The
presence of other signs of congenital infection may point toward a more specific
diagnosis. Liver biopsy may be helpful in making the diagnosis, but the histologic
findings are relatively nonspecific. In most cases, neonatal hepatitis can be
differentiated successfully from biliary atresia by percutaneous liver biopsy.
Clinical management is directed toward nutritional support and medical
management of clinical complications such as ascites or pruritus. The prognosis is
variable, with one-half of the cases resolving with little or no sequelae. Lifethreatening chronic liver disease may necessitate liver transplantation.
Extrahepatic biliary atresia accounts for about one-third of cases of neonatal
cholestasis. Familial cases are uncommon, as are cases in premature infants.
Patients typically present with jaundice during the first or second week of life.
Acholic stools are more common than in neonatal hepatitis. The liver is firm and
enlarged, and splenomegaly may be present, as in neonatal hepatitis. Infants
often appear clinically well, although progressive liver injury results in nutritional
deficiencies, failure to thrive, and ascites.
Differentiation between neonatal hepatitis and biliary atresia has been the subject
of controversy over the years. Serum amino transferase (i.e., transaminase)
levels are notoriously unreliable indicators of neonatal liver disease and may be
normal even in some patients with neonatal hepatitis. Extremely elevated glutamyltransferase levels are suggestive of the marked bile ductular proliferation
found in biliary atresia (157). Liver biopsy demonstrates inflammatory obliteration
of the extrahepatic biliary tree, with bile stasis and bile ductular proliferation in
the liver. Histologic features may overlap with neonatal hepatitis, particularly early
in the course of the disease, making it difficult to differentiate. Histologic
transition from neonatal hepatitis to biliary atresia has been reported and may be
a relatively common phenomenon. The duodenal intubation and aspiration test is
a simple, rapid, inexpensive
P.959
method to check for patency of the extrahepatic biliary tree. Collection of 12 2hour aliquots of duodenal drainage from a feeding tube placed in the duodenum
over a 24-hour period suggests biliary atresia if no yellow fluid is present.
Aspiration of bile-stained fluid from the duodenum suggests patency of the
extrahepatic biliary tract. This test is sensitive and specific for the evaluation of
infantile cholestasis (158). The test has reliability comparable to, or better than,
radionuclide imaging with scans, which may give abnormal results in neonatal
hepatitis during periods of severe cholestasis. Reliability of the radionuclide
studies may be improved by measurement of duodenal fluid counts collected by a
simple string test, in which infants swallow a string that absorbs fluid from the
small intestine that later can be analyzed (159). The combination of liver biopsy
and imaging study or duodenal drainage usually is used to determine whether the
patient is more likely to have biliary atresia or neonatal hepatitis. If bile drainage
cannot be confirmed or the typical histologic picture of neonatal hepatitis is not
apparent on liver biopsy, then surgical exploration and intraoperative
cholangiography usually is performed to establish a final diagnosis.
If the patient has biliary atresia, a hepatoportoenterostomy with Roux-en-Y
enteroanastomosis (i.e., Kasai procedure) is performed to attempt bile drainage.
The Kasai procedure rarely alters the long-term outcome of biliary atresia (160,
but may delay the necessity of liver transplantation. Some advocate not
performing a Kasai procedure. Most, however, believe that performing the
procedure does not worsen the prognosis at transplant and may allow the child to
live longer so that a more suitable liver donor may be located before
transplantation. For those infants with progressive liver disease after the Kasai
procedure, every attempt must be made to optimize their condition before
transplantation. Deficient absorption must be corrected by vitamin
supplementation (i.e., A, D, E, and K) require supplementation. Salt and protein
restriction may become necessary, as liver failure progresses.
Hepatic transplantation is the definitive treatment for biliary atresia and results in
long-term survival in 70% to 85% of infants.
Other Causes of Cholestasis

Rare causes of cholestasis must be excluded before confirming the diagnosis of
neonatal hepatitis or biliary atresia (Table 40-3) (151). Infectious diseases such
as cytomegalovirus, hepatitis B, HIV, rubella, herpes, toxoplasmosis, and syphilis
should be excluded by standard serologic or culture techniques. Metabolic
disorders to be considered include tyrosinemia, galactosemia, and hereditary
fructose intolerance. Obtaining urine for reducing substances to exclude
galactosemia and hereditary fructose intolerance, and succinyl acetone to exclude
tyrosinemia should be done immediately. Anatomic disorders such as choledochal
cysts can be diagnosed by ultrasonography. Other causes such as TPN cholestasis
or sepsis should be considered, based on the clinical presentation.
TABLE 40-3 STEPS IN EVALUATING NEONATAL CHOLESTASIS
1. Determine that hyperbilirubinemia is predominantly direct

2. Exclude metabolic and infectious causes of cholestasis
3. Perform ultrasound to exclude anatomic lesions
4. Obtain percutaneous liver biopsy and hepatobiliary scan or duodenal
drainage study
5. Explore whether studies suggest biliary atresia and perform Kasai
procedure if indicated
Occasionally, liver biopsy will demonstrate a marked reduction in the number of

intrahepatic bile ducts, revealing a disorder known as paucity of the intrahepatic
bile ducts or intrahepatic biliary hypoplasia. Some of these patients fall into the
category of Alagille syndrome, also known as syndromatic paucity or
arteriohepatic dysplasia (161). These patients exhibit unusual facial
characteristics, ocular abnormalities including posterior embryotoxon (i.e.,
prominent Schwalbe line), pulmonic stenosis, and vertebral arch defects, including
anterior vertebral arch fusion with butterfly vertebrae. Careful cardiac
examination, examination of the eyes by an ophthalmologist, and radiographic
examination of the lumbosacral spine should be obtained in patients with
suspected Alagille syndrome who have a paucity of intralobular bile ducts.
Prognosis for long-term survival in syndromatic patients is relatively good.
There are patients with nonsyndromatic paucity who have liver biopsy findings
similar to those with Alagille syndrome. In general, these patients have a much
poorer prognosis than those with syndromatic paucity. Life-threatening cirrhosis
develops in many, and these patients may require hepatic transplantation. As in
biliary atresia, liver transplantation has markedly changed the prognosis for these
patients.
A number of infectious disorders, both viral and bacterial, may present during the
neonatal period, and a specific diagnosis should be sought in such instances.
Hepatitis B typically is found in infants whose mothers were infected during the
third trimester. As many as 60% to 90% of infants born to hepatitis B surface
antigen-positive mothers may be infected. Transmission from mother to infant
most commonly occurs at the time of delivery. Mothers who are hepatitis B
antigen-positive are at extremely high risk to transmit hepatitis B virus to their
infants. Additionally, mothers who are chronic carriers of hepatitis B surface
antigen also may infect their infants. Most hepatitis B in infancy is asymptomatic.
Abnormal liver tests develop at approximately 6 to 8 weeks of age and may
persist for up to 1 year. Nearly 50% of these children remain hepatitis B surface
antigen-positive and are at risk for developing hepatocellular carcinoma. Such

children should be screened annually with -fetoprotein levels for evidence of liver
cancer, unless their hepatitis B surface antigen determinations revert to negative.
Maternal screening for hepatitis B surface antigen
P.960
is essential to prevent perinatal transmission of hepatitis B. Neonates born to
mothers who are hepatitis B antigen-positive should receive hepatitis B
hyperimmune globulin, 0.5 mL intramuscularly at the time of birth, and hepatitis B
vaccine within 12 hours of birth. Booster immunizations with hepatitis B vaccine
are recommended at 1 and 6 months of age (162).
Several other viruses may produce hepatitis during the neonatal period. Hepatitis
A may develop in infants of mothers with active icteric hepatitis A at the time of
delivery. Hepatitis C commonly is transmitted through blood contact and produces
a clinical spectrum similar to hepatitis B. Serologic tests and viral RNA assays are
available to detect hepatitis C. infants with hepatitis B or C run substantial risks
for chronic liver disease. Recent experience has been gained with the use of interferon in the treatment of chronic hepatitis B and C, and with lamivudine in
hepatitis B. Preliminary results look promising, but more experience is needed.
Other viruses, such as Epstein-Barr virus, cytomegalovirus, HIV, rubella, herpes
simplex, coxsackievirus, and adenoviruses may cause a wide spectrum of neonatal
liver disease (163). In most instances, infection with these viruses results in
spontaneous resolution without chronic injury.
Bacteria also may produce neonatal liver injury, because invasion of the liver may
occur. Specific hepatic infection may result from certain bacterial diseases such as
syphilis and listeriosis, and the parasitic disorder toxoplasmosis.
A number of metabolic diseases may present with neonatal cholestasis. The most
common is 1-antitrypsin deficiency. 1-Antitrypsin is the major protease inhibitor
in the hepatocyte. Deficiency of 1-antitrypsin occurs in a number of inheritable
phenotypes. These Pi or protease inhibitor phenotypes can be determined. Type
ZZ produces the most complete deficiency state and most cases of liver disease.
It is estimated that 10% to 20% of type ZZ patients develop liver disease (164).
Isolated cases of type MZ and MS also have been reported with liver injury (165).
The ZZ phenotype is inherited through an autosomal recessive mechanism and
occurs in 1 in 2,000 live births. Patients can be identified by the measurement of a
very low 1-antitrypsin level in the blood. Diagnosis is confirmed by the
determination of the ZZ phenotype, plus the classic histologic findings of periodicacid-Schiff-positive, diastase-resistant granules on liver biopsy.
Liver disease may develop in patients with cystic fibrosis, although very few of
these present during the neonatal period. Measurement of sweat chloride and
specific mutational analysis to exclude cystic fibrosis should be part of the
evaluation of neonatal liver disease.
Three metabolic diseases present with rather fulminant neonatal liver disease.
(See also Chapter 41.) These include galactosemia, hereditary fructose
intolerance, and tyrosinemia. These disorders should be expected when
coagulation abnormalities appear inappropriately severe relative to the apparent
degree of liver disease. Fructose intolerance does not present in the neonatal
period unless the infants is on a sucrose containing formula (sucrose is
metabolized to glucose and fructose). Patients with galactosemia have positive
urinary reducing substances if they are being fed lactose at the time of screening.
Patients with hereditary fructose intolerance also may test positive. Patients with
tyrosinemia can be screened by measuring succinyl acetone content in the urine.
Plasma and urine amino acids will demonstrate marked elevations of tyrosine,
although this may be a nonspecific finding in any infant with neonatal liver
disease. Patients with galactosemia respond to a galactose-free diet, and liver
injury usually resolves spontaneously. Neonatal sepsis, particularly gram negative
sepsis, is a frequent occurrence in these infants, and precautions should be taken.
Patients with tyrosinemia commonly undergo progressive liver and renal
dysfunction and are candidates for emergent liver transplantation once the
diagnosis is made.
Several lipid storage diseases produce neonatal liver disease. Niemann-Pick
disease, Wolman disease, cholesterol-ester storage disease, and Gaucher disease
are included in this group. Most present with an insidious onset later in life.
A number of disorders exist in which paroxysmal dysfunction occurs. The most
common is Zellweger syndrome, the cerebrohepatorenal syndrome. These
patients present with cholestasis, hepatomegaly, hypotonia, and dysmorphic
features, and may be diagnosed by demonstration of very-long-chain fatty acids
in the serum.
Defects in the urea cycle may present with hyperammonemia during the first 2
days of life. A sepsis-like picture with vomiting, lethargy, seizures, and coma
suggests this diagnosis. The most common form is ornithine transcarbamylase
deficiency. Serum ammonia levels are very high, provided the infant is being fed
protein. Diagnosis depends on plasma and urinary amino acid levels, and liver
biopsy must be assayed for specific enzymes (see Chapter 41). Protein intake
should be restricted and liver transplantation should be considered. Transient
hyperammonemia of the newborn also has been reported with spontaneous
resolution and no long-term neurologic sequelae. Permanent resolution of the
hyperammonemia usually occurs by 2 weeks of age.
Another cause of acute liver failure in the neonate is neonatal hemochromatosis or
neonatal iron storage disease (166,167). This rare, apparently inherited disorder
of iron storage and metabolism may present with acute and rather fulminant liver
failure characterized by severe cholestasis and coagulopathy. A variety of
histologic findings have been observed in the livers of these patients, but they
consistently have increased iron deposition in the liver and in other organs.
Salivary gland biopsy may be used to establish the diagnosis. The disorder is
rapidly progressive and often fatal unless transplantation can be performed early.
Recently, antioxidant cocktails have been advocated for such patients, and some
success has been reported (168).
Cholestasis may occur in any patient on chronic parenteral nutrition, but it is far
more common in sick premature infants who receive parenteral nutrition for long
periods
P.961
of time (169,170). The mechanism by which the liver injury occurs is unknown
and perhaps multifactorial (171). Several risk factors have been identified,
however, including recurrent infections, prematurity, and lack of enteral feeding.
Certain components of parenteral solutions have been implicated in causing liver
injury. Excessive caloric administration may play a role. Certain amino acids may
be more hepatotoxic, although many of these data are derived from animal
studies. Higher doses of protein may result in a more rapid rise in bilirubin, but
does not appear to alter ultimate risk of development of liver disease. Available
intravenous lipid preparations do not appear to cause cholestasis and may, in fact,
be beneficial in this regard.
The reason premature infants are more susceptible to liver disease while on
parenteral nutrition probably is related to developmental immaturity of several
hepatobiliary processes. These infants have reduced and altered bile acid
synthesis, decreased bile acid pool size, and, therefore, decreased intraluminal
bile acids. Gallbladder function also is impaired. Bile acid reabsorption from the
small bowel is underdeveloped. The premature liver also is less capable of
detoxifying potentially toxic secondary bile acids.
Lack of enteral feeding definitely predisposes to parenteral nutrition cholestasis.
GI hormones that stimulate bile flow depend on enteral feeding for their release.
Reduced gut motility in the unused bowel may contribute to bacterial proliferation
and the resultant production of toxic secondary bile acids. Infection, especially GI,
and GI surgery may potentiate the liver injury through related mechanisms.
Limited amounts of enteral feeding, as tolerated, may be very beneficial in
preventing liver injury in the parenteral nutrition-dependent infant.
Diagnosis of parenteral nutrition liver disease depends on exclusion of other
causes of cholestasis in the parenteral nutrition-dependent patient. Separation of
this disorder from other causes of cholestasis is difficult using standard laboratory
tests. Histologic study is nonspecific, but may be helpful in making the diagnosis
(172). The disease often is reversible once parenteral nutrition is discontinued. It
occasionally may progress to cirrhosis, and hepatocellular carcinoma has been
reported.
Treatment is accomplished best by discontinuing parenteral nutrition. If this
cannot be accomplished, the following steps should be taken:
Reevaluate solutions to ensure they are appropriately formulated and

balanced.
Use low-dose enteral feedings as tolerated to stimulate bile flow and gut
motility.
Cycle the parenteral nutrition so that it is given over only part of the day.
Use amino acid solutions specially formulated for infants.
Other potential therapies, yet unproven, include choleretics such as Phenobarbital

or ursodeoxycholic acid, hormone stimulation of bile flow, and bowel prokinetic
agents. Success with combined intestinal-liver transplants suggests that this
procedure may play an important role in infants with end-stage parenteral
nutrition liver disease (173).
REFERENCES
1. Lebenthal E, Keung YK. Alternative pathways of digestion and absorption in
the newborn. In: Lebenthal E, ed. Textbook of gastroenterology and nutrition in
infancy, 2nd ed. New York: Raven Press, 1989:3.
2. Lebenthal E, Tucker N. Carbohydrate digestion: development in early infancy.

Clin Perinatol 1986;13:37.
3. Erasmus HD, Ludwig-Auser HM, Paterson PG, et al. Enhanced weight gain in
preterm infants receiving lactase-treated feeds: a randomized, double-blind,
controlled trial. J Pediatr 2002; 141:532.
4. Cicco R, Holzman I, Brown D, et al. Glucose polymer intolerance in

premature infants. Pediatrics 1981;67:498.
5. Lebenthal E, Lee PC. Alternate pathways of digestion and absorption in early

infancy. J Pediatr Gastroenterol Nutr 1984;3:1.
6. Watkins JB. Lipid digestion and absorption. Pediatrics 1985;75: [Suppl]:151.
7. Boehm G, Bierbach U, Seuger H, et al. Activities of lipase and trypsin in

duodenal juice of infants small for gestational age. J Pediatr Gastroenterol Nutr
1991;12:324.
8. Jensen RG, Clark RM, de Jong FA, et al. The lipolytic triad: human lingual,
breast milk and pancreatic lipases: physiological implications of their
characteristics in digestion of dietary fats. J Pediatr Gastroenterol Nutr
1982;1:243.
9. Watkins JB, Ingall D, Szczepanik P, et al. Bile salt metabolism in the

newborn. N Engl J Med 1973;288:431.
10. Balistreri WF, Heubi JE, Suchy FJ. Immaturity of the enterohepatic
circulation in early life: factors predisposing to physiologic malabsorption and
cholestasis. J Pediatr Gastroenterol Nutr 1983; 2:346.
11. Acra SA, Ghishan FK. Active bile salt transport in the ileum: characteristics
and ontogeny. J Pediatr Gastroenterol Nutr 1990;10:421.
12. Euler AR, Byrne WJ, Meis PJ, et al. Basal and pentagastrin stimulated acid
secretion in human newborn infants. Pediatr Res 1979;13:36.
13. Hyman PE, Clarke DD, Everett SL, et al. Gastric acid secretory function in
preterm infants. J Pediatr 1985;106:467.
14. Agunod M, Yamaguchi N, Lopez R, et al. Correlative study of hydrochloric

acid, pepsin and intrinsic factor secretion in newborns and infants. Am J Digest
Dis 1969;14:400.
15. Antonowicz I, Lebenthal E. Developmental pattern of small intestinal

enterokinase and disaccharidase activities in the human fetus. Gastroenterology
1977;723:1299.
16. Younoszai MK, Sapario RS, Laughlin M, et al. Maturation of jejunum and
ileum in rats: water and electrolyte transport during in vivo perfusion of
hypertonic solutions. J Clin Invest 1978; 62:271.
17. Southgate DAT, Widdowson EM, Smits BJ, et al. Absorption and excretion of
calcium and fat by young infants. Lancet 1969; 1:487.
18. Senterre J, Putet G, Salle B, et al. Effects of vitamin D and phosphorus

supplementation on calcium retention in preterm infants fed banked human
milk. J Pediatr 1983;103:305.
19. Voyer M, Davakis M, Antener I, et al. Zinc balances in preterm infants. Biol
Neonate 1982;42:87.
20. Tomomasa R, Hyman PE, Itoh K, et al. Gastroduodenal motility in neonates:

response to human milk compared with cow's milk formula. Pediatrics
1987;80:434.
21. Berseth CL. Gestational evolution of small intestine motility in preterm

infants. J Pediatr 1989;115:646.
22. Worniak ER, Fenton TR, Milla PJ. The development of fasting small intestine
motility in human neonates. In: Roman C, ed. Gastrointestinal motility. London:
Lancaster Press, 1983:265.
23. Nagata S, Koyanagi T, Horimoto N, et al. Chronological development of the

fetal stomach assessed using real-time ultrasound. Early Hum Dev 1990;22:15.
24. Simpson C, Schanler RJ, Lau C. Early introduction of oral feeding in preterm
infants. Pediatrics 2002;110(3):517.
P.962
25. Vanderhoof JA, Rappoport PJ, Paxson CL Jr. Manometric diagnosis of lower
esophageal sphincter incompetence in infants: use of a small, single-lumen
perfused catheter. Pediatrics 1978;62:805.
26. Buchan AMJ, Bryant MG, Polak JM, et al. Development of regulatory
peptides in the human fetal intestine. In: Bloom SR, Polak JM, eds. Gut
hormones. New York: Churchill-Livingston, 1981:119.
27. Bryant MG, Buchan AMJ, Gregor M, et al. Development of intestinal
regulatory peptides in the human fetus. Gastroenterology 1982;83:47.
28. Lucas A, Bloom SR, Aynsley-Green A. Development of gut hormone

responses to feeding in neonates. Arch Dis Child 1980; 55:678.
29. Lucas A, Adrian TE, Christofides ND, et al. Plasma motilin, gastrin and
enteroglucagon and feeding in the human newborn. Arch Dis Child 1980;55:673.
30. Euler AP, Byrne WJ, Cousins LM, et al. Increased serum gastrin
concentrations and gastric hyposecretion in the immediate newborn period.
1977;72:1271.
31. Aynsley-Green A, Lucas A, Bloom SR. The effects of feeds of differing

composition on entero-insular hormone secretion in the first hours of life in
human neonates. Acta Paediatr Scand 1979;68: 265.
32. Lucas A, Bloom SR, Aynsley-Green A. Postnatal surges in plasma gut

hormones in term and preterm infants. Biol Neonate 1982; 41:63.
33. Miller BA. Observations on the gastric acidity during the first month of life.
34. Adrian TE, Savage AJ, Sagor GR, et al. Effect of peptide YY on gastric,
pancreatic and biliary function in humans. Gastroenterology 1985;89:494.
35. Adrian TE, Smith HA, Calvert SA, et al. Elevated plasma peptide YY in
human neonates and infants. Pediatr Res 1986;20:1225.
36. Lucas A, Adrian TE, Bloom SR, et al. Plasma secretin in neonates. Acta
Paediatr Scand 1980;69:205.
37. Johnson LR. Regulation of gastrointestinal growth. In: Johnson LR, ed.
Physiology of the gastrointestinal tract, 2nd ed. New York: Raven Press,
1987:301.
38. Calvert SA, Soltesz G, Jenkins PA, et al. Feeding premature infants with
human milk or preterm milk formula: effects on postnatal growth, intermediary
metabolism and regulatory peptides. Biol Neonate 1985;47:189.
39. Tornhage CJ, Serenius F, Uvnas-Moberg K, et al. Plasma somatostatin and

cholecystokinin levels in preterm infants during kangaroo care with and without
nasogastric tube feeding. J Pediatr Endocrinol Metab 1998;11(5):645.
40. Berseth CL. Gestational evolution of small intestine motility in preterm and
term infants. J Pediatr 1989;115:646.
41. Sarson DL, Wood SM, Holder D, et al. The effect of glucose-dependent
insulinotropic polypeptide infused at physiological concentrations on the release
of insulin in man. Diabetologia 1982;22:33.
42. Lucas A, Sarson DL, Bloom SR, et al. Developmental aspects of gastric
inhibitory polypeptide (GIP) and its possible role in the enteroinsular axis in
neonates. Acta Paediatr Scand 1980;69:321.
43. Lucas A, Aynsley-Green A, Blackburn AN, et al. Plasma neurotensin in term

and preterm neonates. Acta Paediatr Scand 1981;17:201.
44. Savage AP, Adrian TE, Carolan G, et al. Effects of peptide YY (PYY) on
mouth to cecum transit time and on the rate of gastric emptying in healthy
volunteers. Gut 1987;70:166.
45. Drucker DJ, Ehrlich P, Asa SL, et al. Induction of epithelial proliferation by
glucagon-like peptide 2. Proc Natl Acad Sci U S A 1996;92:7911.
46. Chance WT, Foley-Nelson T, Thomas I, et al. Prevention of parenteral

nutrition-induced hypoplasia by coinfusion of glucagon-like peptide-2. Am J
Physiol 1997;273:G559.
47. Lucas A, Bloom SR, Aynsley-Green A. Metabolic and endocrine

consequences of depriving preterm infants of enteral nutrition. Acta Paediatr
Scand 1983;72:245.
48. Adrian TE, Bloom SR. Effect of food on the hormones of the gastrointestinal
tract. In: Hunter JO, Jones V, eds. Food and the gut. Philadelphia: Baillire
Tindall, 1985:13.
49. Isaacs PET, Ladas S, Forgacs IC, et al. A comparison of the effects of
ingested medium- and long-chain triglyceride on gallbladder volume and the
release of cholecystokinin and other gut peptides. Dig Dis Sci 1987;32:481.
50. Martin LW, Torres AM. Omphalocele and gastroschisis. Surg Clin North Am
1985;65:1235.
51. Meller JL, Reyes HM, Loeff DS. Gastroschisis and omphalocele. Clin Perinatol
1989;16:113.
52. Yazbeck S, Ndoye M, Khan AH. Omphalocele: a 25 year experience. J

Pediatr Surg 1986;21:761.
53. Botto LD, Mulinare J, Erickson JD. Occurrence of omphalocele in relation to

maternal multivitamin use: a population-based study. Pediatrics 2002;109
(5):904.
54. Dykes EH. Prenatal diagnosis and management of abdominal wall defects.
Semin Pediatr Surg 1996;5:90.
55. Quirk JG, Forney J, Collins HB, et al. Outcomes of newborns with
gastroschisis: the effects of mode of delivery, site of delivery, and interval from
birth to surgery. Am J Obstet Gynecol 1996;174:1134.
56. Lenke RR, Hatch EI Jr. Fetal gastroschisis: a preliminary report advocating
the use of cesarean section. Obstet Gynecol 1986; 67:395.
57. Langer JC. Gastroschisis and omphalocele. Semin Pediatr Surg 1996;5:124.
58. Herbst JJ. Gastroesophageal reflux in infants. J Pediatr Gastro-enterol Nutr

1985;4:163.
59. Werlin SL, Dodds WJ, Hogan WJ, et al. Mechanisms of gastroesophageal
reflux in children. J Pediatr 1980;97:244
60. Peter CS, Sprodowski N, Bohnhorst B, Silny J, Poets CF. Gastroesophageal
reflux and apnea of prematurity: no temporal relationship. Pediatrics 2001;109

(1):8.
61. Barrington KJ, Tan K, Rich W. Apnea at discharge and gastro-esophageal

reflux in the preterm infant. J Perinatol 2002;22:8.
62. Carlos MA, Babyn PS, Marcon MA, et al. Changes in gastric emptying in
early postnatal life. J Pediatr 1997;130:931.
63. Sondheimer JM. Continuous monitoring of distal esophageal pH: a

diagnostic test for gastroesophageal reflux in infants. J Pediatr 1980;93:804.
64. Tsou VM, Young RM, Hart MH, et al. Elevated plasma aluminum levels in
normal infants using antacids containing aluminum. Pediatrics 1991;87:148.
65. Mehall JR, Nothrop R, Saltzman DA, et al. Acidification of formula reduces
bacterial translocation and gut colonization in a neonatal rabbit model. J Pediatr
Surg 2001;36:56.
66. Strickland AD, Chang JHT. Results of treatment of gastroesophageal reflux

with bethanechol. J Pediatr 1983;103:311.
67. Jolley SG, Halpern LM, Tunell WP, et al. The risk of sudden infant death
from gastroesophageal reflux. J Pediatr Surg 1991; 26:691.
68. Raffensperger JG. Esophageal atresia and tracheoesophageal stenosis. In:

Raffensperger JG, ed. Swenson's pediatric surgery, 5th ed. Norwalk, CT:
Appleton and Lange, 1990:697.
69. Benson CD, Lloyd JR. Infantile pyloric stenosis: a review of 1120 cases. Am
J Surg 1964;107:429.
70. Dodge JA. Genetics of hypertrophic pyloric stenosis. Clin Gastroenterol

1973;2:523.
71. Hernanz-Schulman M, Sells LL, Ambrosino MM, et al. Hypertrophic pyloric

stenosis in the infant without a palpable olive: accuracy of sonographic
diagnosis. Radiology 1994;193: 771.
72. Nord KS. Peptic ulcer disease in the pediatric population. Pediatr Clin North
Am 1988;35:117.
73. Drumm B, Rhoads JM, Stringer DA, et al. Peptic ulcer disease in children:
clinical findings, and clinical course. Pediatrics 1988;82: 410.
74. Murphy MS, Eastham EJ. Peptic ulcer disease in childhood: long-term
prognosis. J Pediatr Gastroenterol Nutr 1987;6:721.
75. Guelrud M, Mujica C, Jaen D, et al. Prevalence of Helicobacter pylori in

neonates and young infants undergoing ERCP for diagnosis of neonatal
cholestasis. J Pediatr Gastroenterol Nutr 1994; 18(4):461.
76. Bell JJ. Perforation of the gastrointestinal tract and peritonitis in the
neonate. Surg Gynecol Obstet 1985;160:20.
77. Smith EI. Malrotation of the intestine. In: Welch KJ, Randolph JG, Ravitch
MM, et al, eds. Pediatric surgery, 4th ed. Chicago: Year Book, 1986:882.
78. Feitz R, Vos A. Malrotation: the postoperative period. J Pediatr Surg

1997;32:1322.
79. Grosfeld JL. Jejunoileal atresia and stenosis. In: Welch KJ, Randolph JG,
Ravitch MM, et al, eds. Pediatric surgery, 4th ed. Chicago: Year Book, 1986:838.
P.963
80. Martin LW, Zerella JT. Jejunoileal atresia: a proposed classification. J Pediatr
Surg 1967;11:399.
81. Holgersen LO, Stanly-Brown EG. Idiopathic post-operative intussusception

in infants and childhood. Am Surg 1978;44:305.
82. Brown EG, Sweet AY. Neonatal necrotizing enterocolitis. Pediatr Clin North
Am 1982;29:1149.
83. Kliegman RM, Fanaroff AA. Necrotizing enterocolitis. N Engl J Med

1984;310:1093.
84. Hack M, Horbar JK, Malloy MH, et al. Very low birth weight outcomes of the
National Institute of Child Health and Human Development Neonatal Network.
85. Uauy RD, Fanaroff AA, Korones SB, et al. Necrotizing enterocolitis in very
low birth weight infants: biodemographic and clinical correlates. J Pediatr
1991;119:630.
86. Bisquera JA, Cooper TR, Berseth CL. Impact of necrotizing enterocolitis on
length of stay and hospital charges in very low birth weight infants. Pediatrics
2002;109(3):423.
87. Kliegman RM, Walsh M. Neonatal necrotizing enterocolitis: pathogenesis,

classification and spectrum of illness. Curr Probl Pediatr 1987;17:213.
88. Ballance WA, Dahms BB, Shenker N, et al. Pathology of neonatal necrotizing
enterocolitis: a ten-year experience. J Pediatr 1990; 117[Suppl 1, Pt 2]:S6.
89. Kliegman RM. Neonatal necrotizing enterocolitis: bridging the basic science
with clinical disease. J Pediatr 1990;117:833.
90. Caplan MS, Sun X-M, Hsueh W, et al. Role of platelet activating factor and
tumor necrosis factor-alpha in neonatal necrotizing enterocolitis. J Pediatr
1990;116:960.
91. Amin HJ, Zamora SA, McMillan DD, et al. Arginine supplementation prevents
necrotizing enterocolitis in the premature infant. J Pediatr 2002;140:425.
92. Neu J. Arginine supplementation and the prevention of necrotizing

enterocolitis in very low birth weight infants. J Pediatr 2002; 140:389.
93. Anderson DM, Kliegman RM. The relationship of neonatal alimentation

practices to the occurrence of endemic necrotizing enterocolitis. Am J Perinatol
1991;8:62.
94. Covert RF, Neu J, Elliott MJ, et al. Factors associated with age of onset of
necrotizing enterocolitis. Am J Perinatol 1989;6:455.
95. van Acker J, de Smet F, Muyldermans G, et al. Outbreak of necrotizing

enterocolitis associated with Enterobactr sakazakii in powdered milk formula. J
Clin Microbiol 2001;39(1):293.
96. Eibl MM, Wolf HM, Furnkranz H, et al. Prophylaxis of necrotizing enterocolitis
by oral IgA-IgG: review of a clinical study in low birth weight infants and
discussion of the pathogenic role of infection. J Clin Immunol 1990;10[Suppl
6]:72S.
97. Aynsley-Green A, Lucas A, Lawson GR, et al. Gut hormones and regulatory
peptides in relation to enteral feeding, gastroenteritis, and necrotizing
enterocolitis in infancy. Arch Dis Child 1990; 117[Suppl]:24.
98. Faix RG, Polley TZ, Grasela TH. A randomized controlled trial of parenteral
clindamycin in neonatal necrotizing enterocolitis. Pediatrics 1988;112:271.
99. Morga LJ, Shochat SJ, Hartman GE. Peritoneal drainage as primary
management of perforated NEC in the very low birth weight infant. J Pediatr
Surg 1994;29:310.
100. Horwitz JR, Lally KP, Chen HW, et al. Complications after surgical
intervention for necrotizing enterocolitis: a multicenter review. J Pediatr Surg
1995;30:994.
101. Buchheit JP, Stewart DL. Clinical comparison of localized intestinal

perforation and necrotizing enterocolitis in neonates. Pediatrics 1994;93:32.
102. Fujii AM, Brown E, Mirochnick M, et al. Neonatal necrotizing enterocolitis

with intestinal perforation in extremely premature infants receiving early
indomethacin treatment for patnet ductus arteriosus. J Perinatol 2002;22:535.
103. Dowling RH, Booth CC. Structural and functional changes following small
intestinal resection in the rat. Clin Sci 1967;32:139.
104. Adrian TE, Savage AP, Fuessl HS, et al. Release of peptide YY (PYY) after
resection of small bowel, colon or pancreas in man. Surgery 1987;101:715.
105. Besterman HS, Adrian TE, Mallinson CN, et al. Gut hormone release after
intestinal resection. Gut 1982;23:854.
106. Armstrong DN, Ballantyne GH, Adrian TE, et al. Adaptive increase in
peptide YY and enteroglucagon after proctocolectomy and pelvic ileal reservoir
reconstruction. Dis Colon Rectum 1991;34:119.
107. Wilmore DW, Dudrick SJ, Daly JM, et al. The role of nutrition in the
adaptation of the small intestine after massive resection. Surg Gynecol Obstet
1971;132:673.
108. Vanderhoof JA. Short bowel syndrome. In: Lebenthal EB, ed.
Gastroenterology and nutrition in early infancy, 2nd ed. New York: Raven Press,
1990:793.
109. Vanderhoof JA. Short bowel syndrome. In: Kassirer JP, ed. Current therapy
in internal medicine, 3rd ed. Philadelphia: BC Decker, 1991:550.
110. Vanderhoof JA. Clinical management of the short bowel syndrome. In:
Balistreri WF, Vanderhoof JA, eds. Pediatric gastroenterology and nutrition.
London: Chapman and Hall, 1990:24.
111. Goulet OJ, Revillon Y, Jan D, et al. Neonatal short bowel syndrome. J
Pediatr 1991;119[Suppl 1, Pt 1]:18.
112. Gracey M. The contaminated small bowel syndrome: pathogenesis,

diagnosis and treatment. Am J Clin Nutr 1979;32:234.
113. Perman JA, Modler S, Barr RG, et al. Fasting breath hydrogen
concentration: normal values and clinical adaptation. Gastroenterology
1984;87:1358.
114. Aarbakke J, Schjonsby H. Value of urinary simple phenol and indican

determinations of the stagnant loop syndrome. Scand J Gastroenterol
1976;2:409.
115. Hudson M, Packnee R, Mowat NA. D-lactic acidosis in short bowel

syndrome: an examination of possible mechanisms. Q J Med 1990;74:157.
116. Taylor SF, Sondheimer JM, Sokol RJ, et al. Noninfectious colitis associated
with short gut syndrome in infants. J Pediatr 1991; 119:24.
117. Young RJ, Vanderhoof JA. Probiotic therapy in children with short bowel
syndrome and bacterial overgrowth. Gastroenerology 1997;112:A916.
118. Caniano DA, Starr J, Ginn-Pease ME. Extensive short-bowel syndrome in

neonates: outcome in the 1980s. Surgery 1989;105: 119.
119. Thompson JS. Recent advances in the surgical treatment of the shortbowel syndrome. Surg Ann 1990;22:107.
120. Thompson J, Pinch L, Murray N, et al. Experience with intestinal

lengthening procedures. J Pediatr Surg 1991;26:721.
121. Vanderhoof JA. Short bowel syndrome in children and small intestinal
transplantation. Pediatr Clin North Am 1996;43:533.
122. Cooper A, Floyd TS, Ross AJ, et al. Morbidity and mortality of short bowel
syndrome acquired in infancy: an update. J Pediatr Surg 1984;19:711.
123. Dorney SFA, Ament ME, Berquist WE, et al. Improved survival in very
short small bowel of infancy with use of long-term parenteral nutrition. J Pediatr
1985;106:521.
124. Adrian TE, Savage AP, Bacarese-Hamilton AJ, et al. Peptide YY

abnormalities in gastrointestinal disease. Gastroenterology 1986; 90:379.
125. Besterman HS, Christofides ND, Welsby PD, et al. Gut hormones in acute
diarrhea. Gut 1983;24:665.
126. Lawson GR, Nelson R, Laker MF, et al. Gut regulatory peptides and
intestinal permeability in acute infantile gastroenteritis. Arch Dis Child
1992;67:272.
127. Walker-Smith J, Harrison M, Kilby A, et al. Cow's milk-sensitive

enteropathy. Arch Dis Child 1978;53:375.
128. Walker-Smith J. Cow's milk protein intolerance: transient food intolerance

of infancy. Arch Dis Child 1975;50:347.
129. Vanderhoof JA, Murray ND, Kaufman SS, et al. Intolerance to protein
hydrolysate infant formulas, an under-recognized cause of gastrointestinal
symptoms in infants. J Pediatr 1997;131:741.
130. Powell GK. Milk- and soy-induced enterocolitis of infancy. J Pediatr

1978;93:553.
131. Avery GB, Villavicencio O, Lilly JR, et al. Intractable diarrhea in early
infancy. Pediatrics 1968;41:712.
132. Goldgar CM, Vanderhoof JA. Lack of correlation of small bowel biopsy and
clinical course of patients with intractable diarrhea of infancy. Gastroenterology
1986;90:527.
133. Orenstein SR. Enteral versus parenteral therapy for intractable diarrhea of
infancy: a prospective, randomized trial. J Pediatr 1986;109:277.
134. Schmitz J, Ginies JL, Arnaud-Battandier F, et al. Congenital microvillous

atrophy, a rare cause of neonatal intractable diarrhoea. Pediatr Res
1982;16:1014.
135. Patey N, Scoazec JY, Cuenod-Jabri B, et al. Distribution of cell adhesion

molecules in infants with intestinal epithelial dysplasia (tufting enteropathy).
Gastroenterology 1997;113:833.
P.964
136. Girault D, Goulet O, L-Ldeist F, et al. Intractable infant diarrhea associated
with phenotypic abnormalities and immunodeficiency. J Pediatr 1994;125:36.
137. Unsworth J, Hutchins P, Mitchell J, et al. Flat small intestinal mucosa and
autoantibodies against the gut epithelium. J Pediatr Gastroenterol Nutr
1982;1:503.
138. Holcomb GW III, Gheissari A, O'Neill JA Jr, et al. Surgical management of

alimentary tract duplications. Ann Surg 1989;209:167.
139. Davis WS, Campbell JB. Neonatal small left colon syndrome. Occurrence in
asymptomatic infants of diabetic mothers. Am J Dis Child 1975;129(9):1024.
140. Martin LW, Torres Am. Hirschsprung's disease. Surg Clin North Am
1985;65:1171.
141. Bill AJ, Chapman ND. The enterocolitis of Hirschsprung's disease: its
natural history and treatment. Am J Surg 1962;103:70.
142. Swenson O. Hirschsprung's disease: A review. Pediatrics 2002; 109(5):914.
143. Heikkinen M, Rintala R, Luukkonen P. Long-term anal sphincter

performance after surgery for Hirschsprung's disease. J Pediatr Surg
1997;32:1443.
144. Granata C, Puri P. Megacystis-microcolon-intestinal hypoperistalsis

syndrome. J Pediatr Gastroenterol Nutr 1997;25:12.
145. Durie PR, Forstner GG. Pathophysiology of the exocrine pancreas in cystic
fibrosis. J R Soc Med 1989;18[Suppl 16]:2.
146. Adrian TE, McKiernan J, Johnstone DI, et al. Hormonal abnormalities of the
pancreas and gut in cystic fibrosis. Gastroenterology 1980;79:460.
147. Aggett PJ, Cavanagh NPC, Matthew DJ, et al. Schwachman's syndrome.
148. Ip WE, Dupuis A, Ellis L, et al. Serum pancreatic enzymes define the
pancreatric phenotype in patients with Shwachman-diamod syndrome. J Pediatr
2002;141:259.
149. Alagille D. Management of chronic cholestasis in childhood. Semin Liver Dis

1985;5:254.
150. Balistreri WF. Neonatal cholestasis. In: Lebenthal E, ed. Textbook of

gastroenterology and nutrition in infancy. New York: Raven Press, 1981:1081.
151. Sokol RJ. Medical management of neonatal cholestasis. In: Balistreri WF,
Stocker JT, eds. Pediatric hepatology. New York: Hemisphere Publishing,
1990:41.
152. Balistreri WF. Neonatal cholestasis: medical progress. J Pediatr

1985;106:171.
153. Balistreri WF. Neonatal cholestasis: lessons from the past, issues for the
future. Semin Liver Dis 1987;7:61.
154. Morecki R, Glaser JH, Cho S, et al. Biliary atresia and reovirus type 3
infection. N Engl J Med 1982;307:481.
155. Morecki R, Glaser J. Reovirus 3 and neonatal biliary disease: discussion of

divergent results. Hepatology 1989;10:515.
156. Fischler B, Haglund B, Hjern A. A population-based study on the incidence

and possible pre- and perinatal etiologic risk factors of biliary atresia. J Pediatr
2002;141:217.
157. Maggiore G, Bernard O, Hadchouel M, et al. Diagnostic value of serum

gamma-glutamyl transpeptidase activity in liver diseases in children. J Pediatr
Gastroenterol Nutr 1991;12:21.
158. Faweya AG, Akinyinka OO, Sodeinde O. Duodenal intubation and aspiration
test: utility in the differential diagnosis of infantile cholestasis. J Pediatr
159. Rosenthal P, Miller JH, Sinatra FR. Hepatobiliary scintigraphy and the
string test in the evaluation of neonatal cholestasis. J Pediatr Gastroenterol Nutr
1989;8:296.
160. Raffensperger JG. A long-term follow-up of three patients with biliary

atresia. J Pediatr Surg 1991;26:176.
161. Alagille D, Odievre M, Gautier M, et al. Syndromic paucity of interlobular

bile ducts (Alagille syndrome or arteriohepatic dysplasia): review of 80 cases. J
Pediatr 1987;110:195.
162. Tajiri H, Nose O, Shimizu K, et al. Prevention of neonatal HBV infection

with the combination of HBIG and HBV vaccine and its long-term efficacy in
infants born to HBeAg positive HBV carrier mothers. Acta Paediatr Jpn
1989;31:663.
163. Hart MH, Kaufman SS, Vanderhoof JA, et al. Neonatal hepatitis and
extrahepatic biliary atresia associated with cytomegalovirus infection in twins.
Am J Dis Child 1991;145:302.
164. Povey S. Genetics of alpha-1-antitrypsin deficiency in relation to neonatal

liver disease. Mol Biol Med 1990;7:161.
165. Pittschieler K. Liver disease and heterozygous alpha-1-antitrypsin

deficiency. Acta Paediatr Scand 1991;80:323.
166. Egawa H, Berquist W, Garcia-Kennedy R, et al. Rapid development of

hepatocellular siderosis after liver transplantation for neonatal
hemochromatosis. Transplantation 1996;62:1511.
167. Barnard JA, Manci E. Idiopathic neonatal iron-storage disease.

Gastroenerology 1991;101:1420.
168. Witzleben CL, Uri A. Perinatal hemochromatosis: entity or end result? Hum
Pathol 1989;20:335.
169. Bell RL, Ferry GD, Smith EO, et al. Total parenteral nutrition-related
cholestasis in infants. J Parenter Enteral Nutr 1986;10:356.
170. Merritt RJ. Cholestasis associated with total parenteral nutrition. J Pediatr
171. Balistreri WF, Novak DA, Farrell MK. Bile acid metabolism, total parenteral
nutrition, and cholestasis. In: Lebenthal E, ed. Total parenteral nutrition:
indications, utilization, complications and pathophysiological considerations.
New York: Raven Press, 1986: 319.
172. Cohen C, Olsen MM. Pediatric total parenteral nutrition, liver

histopathology. Arch Pathol Lab Med 1981;105:152.
173. Vanderhoof JA, Langnas AN, Pinch LW, et al. Short bowel syndrome: a
review. J Pediatr Gastroenterol Nutr 1992;14:359.
Chapter 41
Inherited Metabolic Disorders
Barbara K. Burton
Major advances in the recognition and treatment of inborn errors of metabolism have made it more essential
than ever that the neonatologist be familiar with the clinical presentation of these disorders. Many of the
diseases in this group are associated with symptoms in the neonatal period, and many affected infants find their
way into neonatal intensive care units. The likelihood of establishing a diagnosis often is directly related to the
level of awareness of the neonatologist responsible for the infant's care. Although many of the individual inborn
errors of metabolism occur infrequently, collectively, they are not rare. There is no doubt that a significant
number of children with these disorders are undiagnosed. Every geneticist has had the experience of diagnosing
an inborn error of metabolism in a child and discovering that the parents have had one or more other children
who died in early infancy of vague or undetermined causes. In such cases, it is reasonable to assume that the
other children were similarly affected but undiagnosed. Autopsy findings in such cases are often nonspecific and
unrevealing unless special biochemical studies are done. Infection often is suspected as the cause of death, and
sepsis is a common accompaniment of inherited metabolic disorders.
The significance of the precise diagnosis of metabolic disease cannot be overemphasized. Increasingly, these
disorders are lending themselves to successful medical management. If treatment means the prevention of
significant mental retardation or death, even when the numbers are small, the diagnosis is clearly worth
pursuing. However, the success of most treatment regimens depends on the earliest possible institution of
therapy, stressing the importance of early clinical diagnosis. Even when no effective therapy exists or an infant
cannot be salvaged, diagnosis is critical for purposes of genetic counseling.
Inborn errors of metabolism are all genetically transmitted, typically in an autosomal recessive or X-linked
recessive fashion, and there is usually a substantial risk of recurrence. Prenatal diagnosis is available for many
conditions in this group. Awareness of the diagnosis before birth of an at-risk infant can lead to earlier therapy
and an improved prognosis.
This chapter defines the constellation of findings in the newborn that should alert the clinician to the possibility
of inherited metabolic disease. The discussion is confined to the disorders for which manifestations are observed
in the first few months of life and does not include the many disorders (e.g., most lysosomal storage diseases)
that typically present in later infancy or childhood. The laboratory tools used to evaluate infants suspected of
having inherited metabolic disease are discussed. Treatment of important groups of metabolic disorders are
addressed, focusing on the stabilization and acute management of patients with these conditions.
A list of the major inborn errors of metabolism that have been described clinically in early infancy is shown in
Table 41-1. This table cannot be considered complete because it includes only the disorders for which
manifestations in the first few months of life have been documented in the literature. It is likely that disorders
typically occurring later in childhood occasionally may present as early as the first month of life. New disorders
causing neonatal disease undoubtedly will continue to be described. A single literature reference is listed for
each disorder in the table, and detailed information about most of the disorders listed can be found in recent
editions of reference textbooks 1 (1,2).
CLINICAL MANIFESTATIONS OF INBORN ERRORS OF METABOLISM

Acute Metabolic Encephalopathy
Several groups of inherited metabolic disorders, most notably the organic acidemias, urea cycle defects, and
certain disorders of amino acid metabolism, typically present with acute life-threatening symptoms in the
neonatal period. Because they are associated with protein intolerance, symptoms usually begin after feedings
have been instituted.
P.966
P.967
Affected infants are typically full term and usually appear normal at birth. The interval between birth and clinical
symptoms ranges from hours to weeks. The initial findings are usually those of lethargy and poor feeding, as
seen in almost any sick infant. Although sepsis is often the first consideration in infants who present in this way,
these symptoms in a full-term infant with no specific risk factors strongly suggest a metabolic disorder. Infants
with inborn errors of metabolism may rather quickly become debilitated and septic; therefore, it is important
that the presence of sepsis not preclude consideration of other possibilities. The lethargy associated with these
conditions is an early symptom of a metabolic encephalopathy that may progress to coma. Other signs of central
nervous system (CNS) dysfunction, such as seizures and abnormal muscle tone, may also exist. Evidence of
cerebral edema may be observed, and intracranial hemorrhage occasionally occurs (3).
TABLE 41-1 INBORN ERRORS OF METABOLISM THAT PRESENT IN EARLY INFANCY
Disorder
Reference
Disorders of Carbohydrate Metabolism

Galactosemia (i.e., galactose-1-phosphate uridyl transferase deficiency)
39
Hereditary fructose intolerance (i.e., fructose-1-phosphate aldolase deficiency)

Fructose-1,6-bisphosphatase deficiency
40
41
Glycogen storage disease type I (i.e., von Gierke disease, glucose-6-phosphate deficiency)
Glycogen storage disease type II (i.e., Pompe disease, alpha-glucosidase deficiency)
42
43
Glycogen storage disease type III (i.e., limit dextrinosis, debrancher deficiency)
Glycogen storage disease type IV (i.e., amylopectinosis, brancher deficiency)
Disorders of Amino Acid Metabolism
42
44
Maple syrup urine disease
45
Homocystinuria
Nonketotic hyperglycinemia
Phenylketonuria
Hereditary tyrosinemia
46
47
48
49
Hyperornithinemia-hyperammonemia-homocitrullinuria syndrome
Lysinuric protein intolerance
50
51
Methylene tetrahydrofolate reductase deficiency

Pyridoxine dependency with seizures (i.e., presumed glutamic acid decarboxylase
deficiency)
Organic Acidemias
52
53
Methylmalonic acidemia
Methylmalonic acidemia with homocystinuria
Propionic acidemia
Isovaleric acidemia
3-Methyl crotonyl CoA carboxylase deficiency
Holocarboxylase synthetase deficiency (i.e., early-onset multiple carboxylase deficiency)
Biotinidase deficiency (i.e., late-onset multiple carboxylase deficiency)
Glutaric acidemia type I
Glutaric acidemia type II (i.e., multiple acyl CoA dehydrogenase deficiency, severe)
Ethylmalonic-adipic aciduria (i.e., later-onset glutaric acidemia type II, multiple acyl CoA
dehydrogenase deficiences, mild)
3-Hydroxy-3-methylglutaric acidemia
2-Methylacetoacetyl-CoA thiolase deficiency
Mevalonic aciduria
Pyroglutamic aciduria
3-Hydroxyisobutyric aciduria
3-Methylglutaconic aciduria
2-Methylbutyryl CoA dehydrogenase deficiency
Urea Cycle Disorders
54
55
56
57
58
59
60
61
62
63
Carbamyl phosphate synthetase deficiency

Ornithine transcarbamylase deficiency
Citrullinemia
Argininosuccinic aciduria
71
71
71
71
64
65
66
67
68
69
70
Arginase deficiency
72
N-Acetylglutamate synthetase deficiency

Fatty Acid Oxidation Defects
73
Short-chain acyl CoA dehydrogenase deficiency

Medium-chain acyl CoA dehydrogenase deficiency
74
75
Very-long-chain acyl CoA dehydrogenase deficiency

Long-chain 3-hydroxyacyl-CoA dehydrogenase deficiency
76
77
Carnitine-acylcarnitine translocase deficiency

Carnitine transporter deficiency (i.e., primary systemic carnitine deficiency)
78
79
Carnitine palmitoyl transferase II deficiency

Isobutyryl-CoA dehydrogenase deficiency
Lactic Acidemias
80
81
Pyruvate dehydrogenase deficiency
32
Pyruvate carboxylase deficiency

Phosphoenolpyruvate carboxykinase deficiency
82
83
Mitochondrial encephalomyopathies
Transport Disorders
84
Cystic fibrosis
Infantile free sialic acid storage disease
85
86
Hartnup disease
Lysosomal Storage Disorders
87
GM1 gangliosidosis type I (i.e., generalized gangliosidosis, beta-galactosidase deficiency)
88
Gaucher disease type II (i.e., glucocerebrosidase deficiency)

Niemann-Pick disease types A and B (i.e., sphingomyelinase deficiency)
Niemann-Pick disease type C
Mannosidosis (i.e., a-mannosidase deficiency)
Fucosidosis (i.e., a-fucosidase deficiency)
Farber disease (i.e., acid ceramidase deficiency)
Wolman disease (i.e., acid lipase deficiency)
Krabbe disease (i.e., galactocerebrosidase deficiency)
Mucopolysaccharidosis type VI (i.e., Maroteaux-Lamy syndrome, arylsufatase B
deficiency)
Mucopolysaccharidosis type VII (i.e., beta-glucuronidase deficiency)
Mucolipidosis type II (i.e., I-cell disease)
Mucolipidosis type IV
Multiple sulfatase deficiency
Sialidosis type II (i.e., neuraminidase deficiency)
Peroxisomal Disorders
89
90
91
92
93
94
95
96
97
98
99
100
101
102
Zellweger syndrome
Neonatal adrenoleukodystrophy
Rhizomelic chondrodysplasia punctata
Disorders of Metal Metabolism
103
103
103
Menke kinky hair disease

Molybdenum cofactor deficiency
Sulfite oxidase deficiency
Neonatal hemochromatosis
Other Disorder
104
105
106
107

Carbohydrate deficient glycoprotein syndrome
Hereditary orotic aciduria
Hypophosphatasia
Crigler-Najjar syndrome
A1-antitrypsin deficiency
108
109
110
111
112
113
Canavan disease (i.e., aspartoacylase deficiency)
114
Steroid sulfatase deficiency

Senger syndrome
115
116
Smith-Lemli-Opitz syndrome
Lowe syndrome
32
117
An infant with an inborn error of metabolism who presents more abruptly or in whom the lethargy and poor
feeding go unnoticed may first come to attention because of
P.968
apnea or respiratory distress. The apnea is typically central in origin and a symptom of the metabolic
encephalopathy, but tachypnea may be a symptom of an underlying metabolic acidosis, as occurs in the organic
acidemias. Infants with urea cycle defects and evolving hyperammonemic coma initially exhibit central
hyperventilation, which leads to respiratory alkalosis. Indeed, the finding of respiratory alkalosis in an infant
with lethargy is virtually pathognomonic of hyperammonemic encephalopathy.
TABLE 41-2 LABORATORY STUDIES FOR AN INFANT SUSPECTED OF HAVING AN INBORN ERROR OF
METABOLISM
Complete blood count with differential

Urinalysis
Blood gases
Electrolytes
Blood glucose
Plasma ammonia
Urine reducing substances
Urine ketones if acidosis or hypoglycemia is present
Plasma and urine amino acids, quantitative
Urine organic acids
Plasma lactate
Vomiting is a striking feature of many of the inborn errors of metabolism associated with protein intolerance,
although it is less common in the newborn than in the older infant. If persistent vomiting occurs in the neonatal
period, it usually signals significant underlying disease. Inborn errors of metabolism should always be
considered in the differential diagnosis. It is common for an infant to be diagnosed as having a metabolic
disorder after having undergone surgery for suspected pyloric stenosis (4). Formula intolerance frequently is
suspected, and many affected infants have numerous formula changes before a diagnosis finally is established.
The basic laboratory studies that should be obtained for an infant who has acute life-threatening symptoms
consistent with an inborn error of metabolism are listed in Table 41-2.
HYPERAMMONEMIA
Among the most important laboratory findings associated with inborn errors of metabolism presenting with an
acute encephalopathy is hyperammonemia. A plasma ammonia level should be obtained for any infant with
unexplained vomiting, lethargy, or other evidence of an encephalopathy. Significant hyperammonemia is
observed in a limited number of conditions. Inborn errors of metabolism, including urea cycle defects and many
of the organic acidemias, are at the top of the list. Also in the differential diagnosis is a condition referred to as
transient hyperammonemia of the newborn (THAN) (5). Ammonia levels in these conditions frequently exceed
1,000 mol/L. The finding of marked hyperammonemia provides an important clue to diagnosis and indicates
the need for urgent treatment to reduce the ammonia level. The degree of neurologic impairment and
developmental delay subsequently observed in infants with urea cycle defects depends on the duration of the
neonatal hyperammonemic coma (6).
A flow chart for the differentiation of conditions producing significant hyperammonemia in the newborn is shown
in Fig. 41-1. The timing of the onset of symptoms may provide an important clue. Infants with urea cycle
defects typically do not become symptomatic until after 24 hours of age. Patients with some of the organic
acidemias, such as glutaric acidemia type II, or with pyruvate carboxylase deficiency may exhibit symptomatic
hyperammonemia during the first 24 hours. Symptoms in the first 24 hours are characteristic of THAN, a
condition that is poorly understood but apparently not genetically determined. The typical patient with this
disorder is a large premature infant (mean gestational age of 36 weeks) who has symptomatic pulmonary
disease, often from birth, and severe hyperammonemia. Survivors do not have recurrent episodes of
hyperammonemia and may or may not exhibit neurologic sequelae, depending on the extent of the neonatal
insult. There are some affected infants who survive with normal intelligence despite extraordinarily high
ammonia levels (5). The disorder has become extremely rare in recent years for unknown reasons.
Infants who develop severe hyperammonemia after 24 hours of age usually have a urea cycle defect or an
organic acidemia; infants with organic acidemias typically exhibit a metabolic acidosis and ketonuria as well.
Urine organic acids should always be obtained, regardless of whether or not acidosis is present. Metabolic
acidosis is not a feature of the urea cycle defects. Plasma amino acid analysis is helpful in the differentiation of
the specific defects in this group. Characteristic amino acid abnormalities provide a definitive diagnosis of
citrullinemia and argininosuccinic aciduria. Although no diagnostic amino acid elevations are observed in
carbamyl phosphate synthetase deficiency or ornithine transcarbamylase deficiency, a low or undetectable level
of plasma citrulline is observed in both of these conditions. This finding is helpful in differentiating these two
conditions from THAN, in which the plasma citrulline level is normal. However, plasma citrulline is not accurately
measured in all laboratories performing amino acid analysis, probably because it is important in few other
clinical settings. In clinical situations in which this is a critical diagnostic test, samples should be sent to
laboratories with expertise in the differentiation of urea cycle defects. Carbamyl phosphate synthetase deficiency
and ornithine transcarbamylase deficiency may be differentiated by measuring urine orotic acid, which is low in
the former and elevated in the latter. The pattern of inheritance of the two may also help to differentiate them;
ornithine transcarbamylase deficiency, an X-linked disorder, rarely produces severe hyperammonemia in a
female infant, whereas carbamyl phosphate synthetase deficiency, an autosomal recessive disorder, occurs with
equal frequency in the two genders.
Although the clinical and laboratory evaluation outlined should lead to a specific tentative diagnosis for virtually
P.969
all patients, liver biopsy may be indicated for enzymatic confirmation of the diagnoses of carbamyl phosphate
synthetase and ornithine transcarbamylase deficiencies, because these diagnoses dictate rigid lifelong therapy
or consideration of hepatic transplantation. Acute treatment should be based on the presumptive diagnosis, with
biopsy considered only after the infant is stabilized.
Figure 41-1 Differentiating between conditions that produce severe neonatal hyperammonemia. ASA,
argininosuccinic acid; CPS, carbamyl phosphate synthetase; OTC, ornithine transcarbamylase; PC, pyruvate
carboxylase; THAN, transient hyperammonemia of the newborn.
Less significant elevations of plasma ammonia than those associated with inborn errors of metabolism and THAN
can be observed in a variety of other conditions associated with liver dysfunction, including sepsis, generalized
herpes simplex infection, and perinatal asphyxia. Liver function studies should be obtained in evaluating the
significance of moderate elevations of plasma ammonia. However, even in cases of severe hepatic necrosis, it is
rare for ammonia levels to exceed 500 mol/L (7). Mild transient hyperammonemia with ammonia levels as high
as twice normal is relatively common in the newborn, especially in the premature infant, and is usually
asymptomatic. It appears to be of no clinical significance, and there are no long-term neurologic sequelae (8).
METABOLIC ACIDOSIS
The second important laboratory feature of many of the inborn errors of metabolism during acute episodes of
illness is metabolic acidosis with an increased anion gap, readily demonstrable by measurement of arterial blood
gases or serum electrolytes and bicarbonate. A flow chart for the evaluation of infants with this finding is shown
in Fig. 41-2. An increased anion gap (greater than 16) is observed in many inborn errors of metabolism and in
most other conditions producing metabolic acidosis in the neonate. The differential diagnosis of metabolic
acidosis with a normal anion gap essentially is limited to two conditions, diarrhea and renal tubular acidosis.
Among the inborn errors, the largest group typically associated with overwhelming metabolic acidosis in infancy
is the group of organic acidemias, including methylmalonic acidemia, propionic acidemia, and isovaleric
acidemia. The list of disorders in this group has expanded dramatically as new disorders have been defined
through the use of organic acid analysis.
In addition to specific organic acid intermediates, plasma lactate often is elevated in organic acidemias as a
result of secondary interference with coenzyme A (CoA) metabolism. Neutropenia and thrombocytopenia
commonly are observed and further underscore the clinical similarity of these disorders to neonatal sepsis.
Hyperammonemia, sometimes as dramatic as that associated with urea cycle defects, is seen commonly but not
uniformly in critically ill neonates with organic acidemias.
The metabolic acidosis associated with organic acidemias and certain other inborn errors of metabolism may
have
P.970
significant adverse impact on many different organ systems, which may lead to the erroneous diagnosis of a
wide variety of seemingly unrelated disorders. I had the experience of caring for an infant with isovaleric
acidemia who presented at 10 days of age with respiratory distress, severe metabolic acidosis, a dilated heart,
and poor cardiac output. The infant was suspected of having the hypoplastic left heart syndrome or other severe
congenital heart disease. Cardiac catheterization was performed, even though members of the nursing staff had
observed that the infant had a strong unpleasant odor, reminiscent of sweaty feet. Personnel in the
catheterization laboratory also noticed that the blood had a strong peculiar odor, but it was not until 18 hours
later, long after significant heart disease had been ruled out, that the diagnosis of metabolic disease was first
considered. Despite attempts at therapy with dialysis and other measures, the child succumbed to the disease.
In this case, the metabolic acidosis led to poor function of the myocardium and not the reverse.
Figure 41-2 Evaluating metabolic acidosis in the young infant. fructose-1,6-DP, fructose-1,6-bisphosphatase;
GSD, glycogen storage disease; L/P, lactate/pyruvate.
Another child subsequently found to have methylmalonic acidemia was admitted through the emergency room
with severe metabolic acidosis and a tight, distended abdomen with evidence of multiple air-fluid levels on X-ray
films. The history revealed that the child had fed poorly since birth and had repeated episodes of vomiting
despite several formula changes. Intestinal obstruction was suspected, and the child was taken to the operating
room, in which most of the small intestine was found to be infarcted, presumably secondary to the acidosis and
poor tissue perfusion. No anatomic abnormalities were found. Postoperatively, metabolic disease was
considered, and the diagnosis of a vitamin B12-responsive form of methylmalonic acidemia was made. The
infant died of complications of the disease even though the early diagnosis and treatment of this disorder,
before the terminal episode, should have been associated with a good prognosis.
Defects in pyruvate metabolism or in the respiratory chain may lead to primary lactic acidosis presenting as
severe metabolic acidosis in infancy (9,10). Unlike most of the other conditions presenting acutely in the
newborn, the clinical features of these disorders are unrelated to protein intake. Disorders in this group should
be considered in patients with lactic acidosis who have normal or nondiagnostic urine organic acids.
Differentiation of the various disorders in this group can be facilitated by measuring plasma pyruvate and
calculating the lactate/pyruvate ratio. A normal ratio (less than 25) suggests a defect in pyruvate
dehydrogenase (PDH) or in gluconeogenesis, and an elevated ratio (greater than 25) suggests pyruvate
carboxylase deficiency or a mitochondrial respiratory chain defect.
Not all infants with life-threatening metabolic disease have metabolic acidosis or hyperammonemia. For
example, patients with nonketotic hyperglycinemia typically present in the neonatal period with evidence of
severe and progressive CNS dysfunction, but do not exhibit metabolic acidosis or hyperammonemia (11). Even
patients with galactosemia rarely may present with symptoms of acute CNS toxicity, which may progress to
cerebral edema, when galactose-1-phosphate levels rise precipitously. Therefore, a series of laboratory studies
designed to screen for inborn errors of metabolism should be obtained for any infant with clinical findings
suggesting an inborn error of metabolism, even if metabolic acidosis and hyperammonemia are not present.
These studies are listed in Table 41-2. Most are self-explanatory. Although not available in many hospital
laboratories, amino acid and organic acid analysis can be obtained in any part of the country through reference
P.971
laboratories or through referral of samples to medical center genetics units. It is important to insist that any
reference laboratory used for this purpose provide prompt test results and reference ranges and provide
interpretation of abnormal results.
TABLE 41-3 DISORDERS ASSOCIATED WITH NONGLUCOSE REDUCING SUBSTANCES IN URINE
Disorder
Galactosemia
Hereditary fructose intolerance
Galactokinase deficiency
Essential fructosuria
Pentosuria
Severe liver disease with secondary galactose intolerance
Compound
Galactose
Fructose
p-Hydroxy-phenylpyruvic acid
Galactose
Fructose
Xylulose
Galactose
Urine testing for reducing substances should be performed using Benedict reagent (Clinitest tablets, Miles,
Elkhart, IN). If the result is positive, the urine should be tested for glucose by dipstick. A nonglucose reducing
substance in the urine is probably galactose, but there are other possibilities (Table 41-3).
Several disorders associated with an acute metabolic encephalopathy in the neonate deserve special mention
because they typically are not associated with hyperammonemia or metabolic acidosis. One of these is
nonketotic hyperglycinemia, a condition that typically results in severe and progressive CNS dysfunction,
including obtundation, seizures, and altered muscle tone. Routine laboratory studies all yield normal findings.
The first diagnostic clue is usually the finding of elevated glycine on plasma amino acid analysis. The diagnosis is
confirmed by measurement of cerebrospinal fluid (CSF) glycine and demonstration of an elevated CSF to plasma
glycine ratio. Although therapy of infants with nonketotic hyperglycinemia has been attempted with dietary
protein restriction, sodium benzoate, and a variety of other drugs, the results have been disappointing. Most
infants with this disorder die or exhibit significant neurologic impairment.

TABLE 41-4 MAJOR INBORN ERRORS OF METABOLISM PRESENTING IN THE NEONATE AS AN ACUTE
ENCEPHALOPATHY
Disorders
Characteristic Laboratory Findings
Organic acidemias (includes MMA, PA, IVA, MCD and
Metabolic acidosis with increased anion gap;
many less common conditions)
elevated plasma and urine ketones; variably

elevated plasma ammonia and lactate;
Urea cycle defects
abnormal urine organic acids

Respiratory alkalosis, no metaboalic acidosis;
markedly elevated plasma ammonia; elevated
urine orotic acid in OTCD; abnormal plasma
Maple syrup urine disease
amino acids
Metabolic acidosis with increased anion gap;
elevated plasma and urine ketones; abnormal
plasma amino acids
Non-ketotic hyperglycinemia
No acid-base or electrolyte abnormalities;

normal ammonia; abnormal plasma amino acids
No acid-base or electrolyte abnormalities;

normal ammonia; normal amino and organic
acids; low serum uric acid; elevated sulfites in

urine
IVA, isovaleric acidemia; MCD, multiple carboxylase deficiency; MMA, methylmalonic acidemia; OTCD,
ornithine transcarbamylase deficiency; PA, propionic acidemia.
A second disorder that produces a progressive en-cephalopathy with no clues on routine laboratory studies is
molybdenum cofactor deficiency. The neurologic findings in the affected infant are virtually indistinguishable
from those associated with hypoxic-ischemic encephalopathy. Surviving infants exhibit similar neurologic
sequelae including cerebral palsy, mental retardation, and seizures. The diagnosis may be suggested by the
finding of hypouricemia or, beyond the neonatal period, by ectopia lentis noted on ophthalmologic examination.
If it is suspected, urine should be screened for the presence of sulfites, a finding attributable to a deficiency of
the enzyme sulfite oxidase, which accompanies the disorder.
The inborn errors of metabolism most likely to be associated with an acute encephalopathy in the newborn are
summarized in Table 41-4. The typical laboratory findings in each condition or group of conditions also are listed.
EMERGENCY TREATMENT OF THE INFANT WITH AN ACUTE

METABOLIC ENCEPHALOPATHY
When an inborn error of metabolism, such as an organic acidemia or urea cycle defect, is suspected in a
critically ill infant, immediate treatment should be initiated even if a definitive diagnosis may not yet be
established. Within 48 to 72 hours, the results of amino acid and organic acid analyses should be available,
allowing diagnostic confirmation in
P.972
most cases. Appropriate and aggressive treatment before the confirmation of a diagnosis may be life saving and
may avert or reduce the neurologic sequelae of some of these disorders. The immediate treatment of infants
with disorders in this group has two primary goals. The first is the removal of accumulated metabolites, such as
organic acid intermediates or ammonia. At the first suspicion of a disorder associated with protein intolerance,
protein intake in the form of breast milk, infant formula, or hyperalimentation should be discontinued
immediately. In critically ill infants with hyperammonemia, arrangements should be made for hemodialysis.
Although peritoneal dialysis, continuous arteriovenous hemoperfusion, and exchange transfusion all have been
used in the past to lower plasma ammonia levels, all are substantially less effective than hemodialysis (12). In
infants who are comatose, ventilator dependent, or exhibit evidence of cerebral edema, dialysis should be
instituted immediately without waiting to see if there is a response to dietary manipulation, medication, or other
less aggressive therapy. Maximal supportive care should be provided simultaneously. In patients suspected of
having a urea cycle defect because of significant hyperammonemia without acidosis, an infusion of 6 cc/kg of
10% arginine hydrochloride (HC1) (0.6 g/kg) can be given intravenously over 90 minutes. In patients with
citrullinemia and argininosuccinic aciduria, this often results in a precipitous drop in the plasma ammonia level.
An intravenous arginine preparation is commercially available and should be readily accessible from any hospital
pharmacy.
If an organic acidemia is suspected, vitamin B12 (1 mg) should be given intramuscularly in case the patient has
a vitamin B12-responsive form of methylmalonic acidemia. Biotin (10 mg) should be given orally or by
nasogastric tube, because some patients with multiple carboxylase deficiency are biotin responsive. If acidosis
exists, intravenous bicarbonate should be administered liberally. Calculations of bicarbonate requirements
appropriate for the treatment of other conditions rarely are adequate in these disorders because of ongoing
production of organic acids or lactate. The acid-base status should be monitored frequently, with therapy
adjusted accordingly.
After removing toxic metabolites, the second major goal of therapy in infants with inborn errors of metabolism
should be to prevent catabolism. Ten percent glucose should be liberally administered intravenously, because it
is important to provide as many calories as possible. Intravenous lipids can be given to infants with urea cycle
defects and other disorders in which dietary fat plays no role. Protein should not be withheld indefinitely. If
clinical improvement is observed and a final diagnosis has not been established, some amino acid intake should
be provided after 2 to 3 days of complete protein restriction. Essential amino acids or total protein can be
provided orally or intravenously at an initial dose of 0.5 g protein/kg body weight/24 hours. This should be
increased incrementally to 1.0 g/kg/24 hours and held at that level until the diagnostic evaluation is complete
and plans can be made for definitive long-term therapy. Therapy should be planned in conjunction with a
geneticist or specialist in metabolic disease. Until then, supplemental calories and nutrients can be provided
orally using protein-free diet powder (Product 80056, Mead Johnson, or Prophree, Ross Laboratories).
The chronic therapy of urea cycle defects and most of the organic acidemias involves restriction of dietary
protein. Depending on the specific diagnosis, this may be accomplished by simple restriction of total protein
intake in breast milk or standard infant formula or by use of special formulas designed for individual inborn
errors of metabolism. Formulas have been developed for many of the more common metabolic disorders and are
commercially available. These specialized formulas typically are deficient in one or several specific amino acids.
Dietary treatment alone may be effective in management of some patients with organic acidemias and in several
disorders of amino acid metabolism, such as maple syrup urine disease.
In several of the vitamin-responsive disorders, such as methylmalonic acidemia, multiple carboxylase deficiency,
and homocystinuria, dietary protein restriction may be combined with specific cofactor therapy. In the organic
acidemias and certain other disorders, l-carnitine, usually beginning with a dose of 100 mg/kg/d, may be given.
Acyl-CoAs accumulating in these disorders combine with carnitine to produce acylcarnitine that are water soluble
and excreted in the urine. Without treatment, many patients with these disorders develop a secondary carnitine
deficiency. Treatment with exogenous carnitine prevents the development of symptoms of carnitine deficiency
and provides a measure of protection against recurrent episodes of metabolic decompensation by providing an
augmented mechanism for excretion of accumulated meta-bolites.
Patients with urea cycle defects require supplementation with oral arginine or, in some cases, citrulline, which is
converted to arginine in the body. In normal persons, adequate amounts of arginine are synthesized via the
urea cycle. Patients with a defect in urea synthesis have deficient arginine production and must depend on
dietary supplementation. In the case of carbamyl phosphate synthetase and ornithine transcarbamylase
deficiencies, the most severe of the urea cycle defects, drug therapy also is required. These disorders were
formerly almost uniformly lethal in the neonatal period. The development of novel drugs that provide an
alternate pathway for waste nitrogen excretion has allowed survival of many affected infants (13). Sodium
benzoate and sodium phenylacetate were the agents originally used, but these have been replaced largely for
oral use by sodium phenylbutyrate.
Despite rigorous therapy and intensive surveillance, patients with urea cycle defects remain at risk for
intercurrent episodes of hyperammonemia, which may result in death or neurologic sequelae. The risk appears
to be greatest for patients with carbamyl phosphate synthetase and ornithine transcarbamylase deficiency. Liver
transplantation should be considered seriously for patients with these disorders, if they can be stabilized.
P.973
HYPOGLYCEMIA
Hypoglycemia and its associated symptoms occasionally may be seen in infants with disorders of protein
intolerance, but it more commonly is seen in disorders of carbohydrate metabolism or of fatty acid oxidation.
Among the best known inborn errors of metabolism associated with hypoglycemia are the glycogen storage
diseases, of which types I and III are the most likely to be associated with manifestations in the neonatal
period. The hypoglycemia in these disorders is related to the inability of the liver to release glucose from
glycogen, and it is most profound during periods of fasting. Hypoglycemia, hepatomegaly, and lactic acidosis are
prominent features of these disorders. Hypoglycemia is not a feature of glycogen storage disease type II
(Pompe disease) because cytoplasmic glycogen metabolism and release are normal in this disorder, in which
glycogen accumulates within lysosomes as a result of the deficiency of the lysosomal enzyme a-1,4-glucosidase.
The clinical manifestations of this disorder include macro-glossia, hypotonia, cardiomegaly with congestive heart
failure, and hepatomegaly. Cardiomegaly is the most striking and may be apparent in the neonatal period.
Congestive heart failure is the cause of death in most cases.
A disorder that presents clinically with findings virtually indistinguishable from the hepatic glycogen storage
diseases types I and III is fructose-1,6-bisphosphatase deficiency, a disorder of gluconeogenesis. Several other
disorders of gluconeogenesis have been described. The basic immediate treatment of all of these disorders is
frequent feedings and glucose administration. The definitive diagnosis is made by liver biopsy and assay of
appropriate hepatic enzymes. In some cases, enzymatic assays can be performed using lymphocytes or cultured
skin fibroblasts.
A number of inherited defects in fatty acid oxidation have been identified in infants presenting with
hypoglycemia. Although many of the disorders in this group typically present after 2 months of age, neonatal
manifestations may be observed. These disorders are important because of their apparent frequency and
because of the variability of the initial presentation. Affected infants have an impaired capacity to use stored fat
for fuel during periods of fasting and readily deplete their glycogen stores. Despite the development of
hypoglycemia, acetyl CoA production is diminished, and ketone production is impaired. The hypoglycemia
occurring in these conditions typically is characterized as nonketotic, although small amounts of ketones may be
produced. Hypoglycemia may occur as an isolated finding or may be accompanied by many of the other
biochemical derangements typically associated with Reye's syndrome, such as hyperammonemia, metabolic
acidosis, and elevated transaminases. Hepatomegaly may or may not be present. Any infant presenting with
findings suggesting Reye's syndrome should be evaluated for fatty acid oxidation defects. As the incidence of
true Reye's syndrome has decreased, most children presenting at any age with this constellation of findings
have an inherited metabolic disorder.
The most common of the fatty acid oxidation defects is medium-chain acyl-CoA dehydrogenase deficiency, which
is estimated to occur in 1 of 15,000 births, an incidence similar to that observed for phenylketonuria (PKU) (14,
15). It is among the most common inborn errors of metabolism. In addition to presenting as nonketotic
hypoglycemia or a Reye's-like syndrome, it may present as sudden death or an acute life-threatening event.
Many infants diagnosed as having medium-chain acyl-CoA dehydrogenase deficiency have a history of a sibling
who died of sudden infant death syndrome (16). Fat accumulation in the liver or muscle of any infant who dies
unexpectedly should strongly suggest the possibility of this or a related disorder of fatty acid oxidation. Verylong-chain fatty acyl-CoA dehydrogenase deficiency is associated with similar clinical findings, although there
may be evidence of a significant cardiomyopathy. Infants with this defect may present with cardiac arrhythmias
or unexplained cardiac arrest. Defects in the carnitine cycle or in carnitine uptake also may lead to a profound
defect in fatty acid oxidation and result in sudden neonatal death.
The accumulation of fatty acyl-CoAs in patients with fatty acid oxidation defects leads to a secondary carnitine
deficiency, probably as a result of excretion of excess acylcarnitine in the urine (17,18). Urine organic acid
analysis and measurement of serum carnitine and analysis of the plasma acylcarnitine profile are the most
helpful laboratory studies in the initial screening for defects in fatty acid oxidation. These studies are sufficient
to establish the diagnosis of medium-chain acyl-CoA dehydrogenase deficiency, which is associated with the
presence of a characteristic metabolite, octanoylcarnitine, on the acylcarnitine profile. Enzymatic assays may be
necessary for the definitive diagnosis of some of the fatty acid oxidation defects. As is true for the defects in
carbohydrate metabolism leading to hypoglycemia, treatment of the fatty acid oxidation defects involves
avoidance of fasting and provision of adequate glucose. Restriction of dietary fat intake and supplemental lcarnitine therapy at a dose of 50 to 100 mg/kg/d is recommended. With appropriate therapy, patients with
medium-chain acyl-CoA dehydrogenase deficiency appear to have an excellent prognosis. The prognosis for the
other fatty acid oxidation defects is more variable.
JAUNDICE AND LIVER DYSFUNCTION
Jaundice or other evidence of liver dysfunction may be the presenting finding in a number of inherited metabolic
disorders in the neonatal period. These are listed in Table 41-5, along with the laboratory studies useful in
diagnosis. For most of the inborn errors of metabolism associated with jaundice, the elevated serum bilirubin is
of the direct-reacting type. This generalization does not include those inborn errors of erythrocyte metabolism,
such as
P.974
glucose-6-phosphate dehydrogenase deficiency or pyruvate kinase deficiency, which occasionally are responsible
for hemo-lytic disease in the newborn. The best-known metabolic disease associated with jaundice is
galactosemia, in which the deficiency of the enzyme galactose-1-phosphate uridylyl transferase results in an
accumulation of galactose-1-phosphate and other metabolites, such as galactitol, which are thought to have a
direct toxic effect on the liver and on other organs. Jaundice and liver dysfunction in this disorder are
progressive and usually appear at the end of the first or during the second week of life, with vomiting, diarrhea,
poor weight gain, and eventual cataract formation if the infant is receiving breast milk or a galactose-containing
formula. Hypoglycemia may be observed. The disease may present initially with indirect hyperbilirubinemia
resulting from hemolysis secondary to high levels of galactose-1-phosphate in erythrocytes. Alternatively, the
effects of acute galactose toxicity on the brain rarely may cause the CNS symptoms to predominate and, in
some cases, Escherichia coli sepsis is the presenting problem.
TABLE 41-5 INBORN ERRORS OF METABOLISM ASSOCIATED WITH NEONATAL LIVER DISEASE AND
LABORATORY STUDIES USEFUL IN DIAGNOSIS
Disorder
Galactosemia
a1-antitrypsin deficiency
Neonatal hemochromatosis
Zellweger syndrome
Niemann-Pick disease type C
Glycogen storage disease type IV (brancher
deficiency)
Laboratory Studies
Urine reducing substances; red blood cell

galactose-1-phosphate uridyl transferase
Plasma quantitative amino acids; urine
succinylacetone
Quantitative serum alpha1-antitrypsin; protease
inhibitor (Pi) typing
Serum ferritin; liver biopsy; buccal biopsy
Plasma very long chain fatty acids
Skin biopsy for fibroblast culture; studies of
cholesterol esterification and accumulation
Liver biopsy for histology and biochemical
analysis or skin biopsy with assay of branching
enzyme in cultured fibroblasts
If galactosemia is suspected, the urine should be tested simultaneously with Benedict reagent and with a
glucose oxidase method. The glucose oxidase method is specific for glucose, and Benedict reagent can detect
any reducing substance. A negative dipstick for glucose with a positive Benedict reaction means that a
nonglucose reducing substance is present. With appropriate clinical findings, this is most likely to be galactose.
Paper or thin-layer chromatography can be used to identify positively the reducing substance. If a child with
galactosemia has been on intravenous fluids and recently has not been receiving galactose in the diet, galactose
may not be present in the urine.
If the diagnosis of galactosemia is suspected, whether or not reducing substances are found in the urine,
galactose-containing feedings should be discontinued immediately and replaced by soy formula or other lactosefree formula, pending the results of appropriate enzyme assays on erythrocytes to confirm the diagnosis.
Untreated galactosemics, if they survive the neonatal period, have persistent liver disease, cataracts, and severe
mental retardation. Many affected infants die of E. coli sepsis in the neonatal period, and the early onset of
sepsis may alter the presentation of the disorder (19).
Treatment of the disorder by maintenance of strict dietary restriction of galactose, if started early, results in
complete reversal of liver disease and enables many affected individuals to develop normal or near-normal
intelligence. Unfortunately, there continues to be an increased incidence of mental retardation even among
treated patients. Additionally, there are some late sequelae of the disorder that appear to be unaffected by
current therapy. These include premature ovarian failure in females and a late-onset neurologic syndrome
involving ataxia and tremors in both genders (20,21). Many states have newborn screening programs for
galactosemia, but clinical manifestations of the disorder often appear before the results of screening studies are
available; therefore, it is critical that physicians remain alert to this possibility.

Another inborn error of metabolism that occasionally presents in the newborn period with jaundice,
hepatomegaly, and the presence of reducing substances in the urine is hereditary fructose intolerance, which is
characterized by episodes of profound hypoglycemia, vomiting, and metabolic acidosis. This disorder is seen
uncommonly in the neonate, because most newborns are not exposed immediately to a fructose-containing diet
unless they have been given a soy formula with sucrose as the carbohydrate source. In the uncommon event
that an infant who has been receiving fructose should present with these findings, this diagnosis should be
considered. Analysis of the urine reveals the presence of a nonglucose reducing substance that can be
demonstrated by chromatography to be fructose. Treatment involves elimination of fructose from the diet and
results in complete resolution of all clinical signs and symptoms. Confirmation of the diagnosis is by assay of the
deficient enzyme fructose-1-phosphate aldolase in liver tissue, but this rarely is necessary.
P.975
a1-Antitrypsin deficiency, a puzzling disorder that is among the most common of all inherited metabolic
diseases, also may present with neonatal jaundice (22). The clinical manifestations of this disorder may be
identical to those of traditional neonatal or giant cell hepatitis, and a determination of serum a1-antitrypsin
should be a part of the initial evaluation of all children presenting with this syndrome. Infants with deficient
levels of a1-antitrypsin on quantitative analysis should have protease inhibitor typing performed to confirm the
diagnosis. There is no specific treatment for the liver disease associated with a1-antitrypsin deficiency, but
approximately one-half of all affected infants eventually exhibit complete resolution of the liver dysfunction.
Others may progress to end-stage disease and require liver transplantation. A history of chronic pulmonary
disease in adult family members may be obtained.
Hereditary tyrosinemia is another disorder that presents with liver disease in early infancy. The biochemical
hallmarks of this disorder include marked elevations of plasma tyrosine and methionine and generalized aminoaciduria with a disproportionate increase in the excretion of tyrosine. However, these findings are relatively
nonspecific and may be observed as a secondary phenomenon in other forms of liver disease. Hereditary
tyrosinemia once was among the most difficult of inborn errors of metabolism to diagnose clinically. The finding
of succinylacetone in the urine of patients with this disease has led to a helpful diagnostic test for the disorder
(23). It also has become possible to establish the diagnosis definitively by demonstrating a deficiency of the
enzyme fumarylacetoacetate fumarylhydrolase in lymphocytes and cultured skin fibroblasts of affected
individuals (24).
Neonatal hemochromatosis, a recently described and poorly understood disorder, may be the most common
cause of congenital cirrhosis. Its fulminating course distinguishes it from many of the other metabolic disorders
associated with neonatal liver disease. In addition to being associated with severe liver failure from birth, the
disorder is characterized by distinctive hepatic morphology and hepatic and extrahepatic parenchymal iron
deposition. Serum ferritin and iron typically are elevated, whereas total transferrin is low, but these findings are
not diagnostic. The definitive diagnosis is established by liver biopsy or autopsy. If liver biopsy is contraindicated
because of a secondary coagulopathy, biopsy of the salivary glands or buccal mucosa is a useful alternative.
Most affected infants succumb to the disorder during the early weeks of life. The only clearly effective mode of
therapy is liver transplantation; intensive chelation therapy may be attempted but largely has been
unsuccessful. At present, it is not clear whether iron storage is the primary defect or is secondary to fetal liver
disease that may be causally heterogeneous. There are familial recurrences, so at least some cases appear to
have a genetic basis, although the mode of inheritance is not well established.
Less common metabolic causes of neonatal liver dysfunction include Niemann-Pick disease type C and glycogen
storage disease type IV. Infants with Niemann-Pick disease type C exhibit cholestatic jaundice, which typically
resolves by several months of age. They then are clinically normal for a period of months to years before
developing findings of a degenerative neurologic disorder. Infants with glycogen storage disease type IV
accumulate an abnormal form of glycogen in the liver as a result of a deficiency of the glycogen branching
enzyme. This leads to progressive cirrhosis and generalized hepatic dysfunction. Hypoglycemia is not a
prominent feature, as it is in some other forms of glycogen storage disease.
Zellweger syndrome, formerly referred to as the cerebrohepatorenal syndrome, is another cause of neonatal
jaundice and hepatic dysfunction, but it usually is recognizable clinically because of the associated hypotonia
and dysmorphic features. It is the prototype of the peroxisome assembly disorders and is associated with
generalized peroxisomal dysfunction.
In contrast to disorders in which there is an elevation of the direct-reacting bilirubin, a persistent elevation of
indirect bilirubin beyond the limits of physiologic jaundice, without evidence of hemolysis, suggests the diagnosis
of the Crigler-Najjar syndrome. The hyperbilirubinemia in this disorder is related to a partial or complete
deficiency of glucuronyl transferase, the liver enzyme responsible for the normal conjugation of bilirubin to
bilirubin diglucuronide. There is no effective long-term therapy for all patients with this disorder, but the
standard modalities of phototherapy and exchange transfusion may prevent the development of kernicterus in
the neonatal period (25,26). Hepatic transplantation has been performed successfully in patients with this
disorder. Patients with a partial deficiency of the enzyme may respond to phenobarbital therapy (26).
FINDINGS SUGGESTIVE OF A STORAGE DISEASE

Many of the well-known lipid storage diseases typically do not present in the neonatal period. Among those that
occasionally may be associated with hepatosplenomegaly in the neonatal period are GM1-gangliosidosis type I,
Gaucher disease, Niemann-Pick disease, and Wolman disease. The glycogen storage diseases that are
associated with hepa-tomegaly in the newborn have previously been discussed in reference to hypoglycemia.
Infants with the most common mucopolysaccharidoses, such as the Hurler and Hunter syndromes, uncommonly
exhibit clinical abnormalities in the first month of life. Newborns with the typical features of these syndromes,
such as coarse facial features, hepatosplenomegaly, skeletal abnormalities, and hernias, are more likely to have
GM1-gangliosidosis or a mucolipidosis, such as I-cell disease. Beta-glucuronidase deficiency, also classified as
mucopolysaccharidosis type VII, may present in the neonatal period with features virtually indistinguishable
clinically from those seen later in the Hurler and Hunter syndromes. An infantile form of
P.976
sialidosis (i.e., neuraminidase deficiency) typically is associated with findings at birth. The clinical manifestations
of several of these conditions may be so severe in utero that fetal hydrops develops.
If one of these disorders is suspected, urine screening tests for mucopolysaccharides and oligosaccharides
should be performed. These can be helpful diagnostically, but negative results do not rule out the possibility of a
storage disorder. False-positive mucopolysaccharide spot tests are commonly observed in neonates. The
definitive diagnosis of most disorders of lipid or mucopolysaccharide metabolism is made by appropriate
biochemical studies on leukocytes or cultured skin fibroblasts.
ABNORMAL ODOR
Abnormal body or urinary odor, more commonly observed by nurses or mothers rather than physicians, is an
important but often overlooked clue to the diagnosis of several of the inborn errors of metabolism and may be
the most specific clinical finding in these patients. It is best described for PKU, for which the urine was found to
have a peculiar musty odor years before the biochemical basis of the disease was understood. In the acutely ill
neonate with an abnormal odor, isovaleric acidemia, glutaric acidemia type II, and maple syrup urine disease
are the most likely entities to be encountered. In maple syrup urine disease, the urine has a distinctive sweet
odor, said to be reminiscent of maple syrup or burnt sugar. The odor associated with isovaleric acidemia and
glutaric acidemia type II is pungent and unpleasant and similar to that of sweaty feet.
DYSMORPHIC FEATURES
There formerly appeared to be a clear distinction between inborn errors of metabolism and dysmorphic
syndromes, both of which may be inherited in a similar fashion. Infants with inherited metabolic disease were
thought to be phenotypically normal at birth, with no evidence of major or minor structural anomalies. It is
becoming increasingly apparent that inherited metabolic disorders may be associated with consistent patterns of
birth defects, suggesting that metabolic derangements in utero may disrupt the normal process of fetal
development.
This phenomenon is illustrated clearly by the group of disorders associated with multiple defects in peroxisomal
enzymes, including those involved in fatty acid oxidation and plasmalogen synthesis (27,28). These include
Zellweger syndrome, neonatal adrenoleukodystrophy, and several variant conditions, all of which are associated
with congenital hypotonia and dysmorphic features, such as epicanthal folds, Brushfield spots, large fontanels,
simian creases, and renal cysts. Patients with glutaric acidemia type II, one of the organic acidemias, have a
characteristic phenotype, including a high fore head, hypertelorism, low-set ears, abdominal wall defects,
palpably enlarged kidneys, hypospadias, and rocker bottom feet (29,30). An energy-deficient mechanism,
referred to as fuel-mediated teratogenesis, similar to that postulated for maternal diabetes mellitus, has been
suggested to explain these findings. Several of the other organic acidemias, such as mevalonic aciduria, and 3-
hydroxyisobutyric aciduria, have been associated with multiple dysmorphic features.

Some infants with pyruvate dehydrogenase (PDH) deficiency have dysmorphic facial features resembling those
observed in the fetal alcohol syndrome (FAS) (31). The specific findings observed include a narrow forehead
with frontal bossing, a broad nasal bridge, short nose with anteverted nostrils, and a long philtrum. The
resemblance to FAS has been explained by suggesting that there is a common mechanism in the two disorders,
involving a deficiency of PDH activity. It has been postulated that, in FAS, acetaldehyde from the maternal
circulation inhibits fetal PDH, which leads to malformations.
The Smith-Lemli-Opitz syndrome is an autosomal recessive disorder associated with a wide range of
malformations, including dysmorphic facies, cleft palate, congenital heart disease, hypospadias, polydactyly, and
2-3 syndactyly of the feet. Recent observations have revealed that this disorder is an inborn error of cholesterol
biosynthesis. Affected infants have decreased levels of plasma cholesterol accompanied by markedly elevated
levels of the cholesterol precursor 7-dehydrocholesterol (32).
Isolated malformations may be even more commonly associated with inherited metabolic disorders than are
specific malformation patterns. Patients with nonketotic hyperglycinemia frequently have agenesis of the corpus
callosum and may have gyral malformations related to defects in neuronal migration as well (33). Patients with
PDH deficiency also may exhibit agenesis of the corpus callosum (34). It is not uncommon for patients with
almost any of the inborn errors of metabolism to exhibit one or more dysmorphic features or anomalies that are
nonspecific. The observation of dysmorphic features in an infant in no way should preclude consideration of an
inherited metabolic disorder. In selected circumstances, it may heighten the clinical suspicion.
ABNORMAL EYE FINDINGS

Abnormal eye findings typically are associated with many of the inborn errors of metabolism, although they are
not always found at the time of initial presentation. Cataracts classically are associated with galactosemia and
other disorders of galactose metabolism, but also are observed in disorders such as Zellweger syndrome and
Lowe syndrome. Dislocated lenses, seen in homocystinuria, molybdenum cofactor deficiency, and sulfite oxidase
deficiency, may be found as early as the first month of life and are an important clue to diagnosis. Retinal
degenerative changes are typical of the peroxisomal disorders, including Zellweger syndrome
P.977
and neonatal adrenoleukodystrophy, and are observed in several other conditions. Other abnormalities that may
be associated with inborn errors of metabolism include corneal clouding and congenital glaucoma. A careful eye
examination, preferably by an ophthalmologist, should be performed whenever an inherited metabolic disorder
is suspected. A summary of some of the inherited metabolic disorders associated with specific ocular
abnormalities is shown in Table 41-6
TABLE 41-6 EYE ABNORMALITIES ASSOCIATED WITH INBORN ERRORS OF METABOLISM
Eye Finding
Cataracts
Ectopia lentis
Cherry red spot
Associated Disorders
Galactosemia
Homocystinuria
Lowe syndrome
Zellweger syndrome
Rhizomelic chondrodysplasia punctata
Senger syndrome
Hypophosphatasia
Homocystinuria
Sulfite oxidase deficiency
Niemann-Pick disease types A and B
Gaucher disease type II
GM2 gangliosidosis (Tay-Sachs; Sandhoff)
Sialidosis type II
Farber disease
Corneal clouding
Mucopolysaccharidoses
Mucolipidoses
Lowe syndrome
Pigmentary retinopathy
Homocystinuria
Zellweger syndrome
Neonatal adrenoleukodystrophy
Long-chain 3-hydroxyacyl-CoA dehydrogenase deficiency
SAMPLES TO OBTAIN FROM A DYING CHILD WITH A SUSPECTED

INBORN ERROR OF METABOLISM
If death appears imminent in a child suspected of having an inborn error of metabolism, it is important to obtain
the appropriate samples for postmortem analysis. This is critical for resolution of the cause of death and is
essential for subsequent genetic counseling and prenatal diagnosis. The following samples should be collected
and stored: urine, frozen; plasma, separated from whole blood and frozen; and a small snip of skin obtained
using sterile technique and stored at room temperature or 37C in tissue culture medium, if available, or sterile
saline. The latter sample can be obtained by slipping a 25-gauge needle under the skin, lifting the skin, and
snipping a 2- to 3-mm ellipse with a sterile scissors, which can be found in a suture removal tray. The skin
should be cleansed with alcohol. If an autopsy is performed, a sample of unfixed liver tissue should be obtained
as soon as possible after death and frozen at -20C for subsequent biochemical studies. Additional tissue should
be preserved for electron micro-scopy. If consent for autopsy is denied, consent for a postmortem needle biopsy
of the liver should be requested. The liver tissue should be frozen in total or in part if histologic studies appear
to be indicated. As soon as possible after death, the case should be reviewed with a metabolic specialist and
plans made for the transport of samples to the appropriate laboratory.
NEWBORN SCREENING FOR INHERITED METABOLIC DISORDERS

All 50 states and the District of Columbia in the United States and many other countries have newborn screening
programs in place for genetic disorders. However, there are significant state-to-state differences in the disorders
that are included, the methods of screening, and follow-up. Therefore, physicians must become familiar with the
newborn screening protocol in the state or country in which they are practicing. The recent development of the
technology of tandem mass spectroscopy has led many states to reexamine their newborn screening practices
and many are now moving toward implementation of expanded newborn screening using this technology.
Tandem mass spectroscopy (MS/MS) allows for the performance of a complete acylcarnitine profile and amino
acid profile on the filter paper blood specimen and is being used in an increasing number of states and countries
to screen for organic acid and fatty acid oxidation disorders and an increased number of amino acid disorders,
including maple syrup urine disease, homocystinuria, and hereditary tyrosinemia. Phenylketonuria is also
detected through MS/MS, replacing testing methods previously used for this disorder which was routinely
screened for in all states and the District of Columbia.
The only other disorder that is a routine component of newborn screening in all 50 states and the District of
Columbia is congenital hypothyroidism. Hemoglobinopathies and galactosemia are included in the newborn
screening panel of most states. Testing methods vary from state to state for each of these individual disorders.
In the case of many metabolic disorders such as PKU, the levels of the abnormal metabolite do not begin to rise
in the infant's circulation until the umbilical cord is cut and therefore testing is not recommended until the infant
is at least 24 hours of age. In the case of infants discharged before 24 hours of age, most states recommend
that a sample be obtained at discharge and subsequently repeated. In the case of sick infants who are not being
fed, the initial newborn screening sample should not be withheld indefinitely but should be submitted at the
recommended time, often 7 days, because screening for some other disorders, such as hypothyroidism, is not
affected by the feeding history. For PKU screening and screening for some other disorders, however, another
specimen should be submitted
P.978
after feedings are instituted. Confirmation of positive newborn screening test results is always necessary to
establish a definitive diagnosis regardless of the type of screening method utilized. Even with tandem mass
spectroscopy, false positive test results occur. Each state or country has developed a system for the evaluation
of infants with positive results which may or may not include submission of repeat specimens to a central
laboratory.
As newborn screening for metabolic disorders is expanded, an increasing number of inborn errors of metabolism
will be diagnosed presymptomatically. This should not in any way discourage the neonatologist from considering
the diagnosis of an inborn error of metabolism in an infant presenting with appropriate signs and symptoms.
Many infants with conditions such as galactosemia, maple syrup urine disease and many of the organic
acidemias will become symptomatic before the results of newborn screening tests are available. Additionally,
there are many important inherited metabolic disorders, such as ornithine transcarbamylase deficiency that will
not be detected by newborn screening, even with tandem mass spectroscopy. Therefore, it will be important to
continue to have a high index of suspicion for metabolic disease in any infant who presents with findings
suggestive of such a disorder.
MATERNAL METABOLIC DISORDERS

With advances in therapy for inborn errors of metabolism, it is now common for patients with many of these
disorders to reach adult life with normal or near-normal intelligence and the desire to have families of their own.
This has led to serious concerns about the potential adverse effects of maternal metabolic derangements on
fetal growth and development. The real potential for adverse consequences is illustrated by the experience that
has accumulated with maternal PKU. In the past, patients with PKU were severely retarded and did not
reproduce. This changed completely with the initiation of newborn screening programs and early dietary
management. Dietary therapy was once maintained until 5 to 6 years of age and then discontinued. Treatment
now is continued indefinitely in most cases, because it has been demonstrated that some patients exhibit
neurologic deterioration and loss of intelligence quotient (IQ) points after discontinuation of the diet.
Nonetheless, many patients with PKU who are now adults have been off the diet for years and have high
phenylalanine levels. After women with PKU began reproducing, it became clear that the maternal metabolic
environment in this condition had extremely harmful effects on fetal development. A spectrum of findings
referred to as maternal PKU syndrome is observed in a large percentage of exposed infants, most of whom do
not themselves have PKU (35,36). More than 90% of exposed infants exhibit mental retardation, and
microcephaly occurs in 72%, growth retardation in 40%, and congenital heart disease in 12%. Altered facial
features, similar to those observed in FAS, may be observed. Mothers with hyperphenylalaninemia, a condition
that is associated with lower phenylalanine levels than classical PKU and does not always require treatment,
may also be at increased risk for fetal abnormalities.
There is evidence that dietary treatment of pregnant women before conception and throughout pregnancy, with
careful control of phenylalanine levels, reduces the risk to the fetus (38). This is a difficult goal to achieve,
however, because the phenylalanine-restricted diet is an onerous one to patients who have ever been on a
normal diet, and some adult patients, despite early therapy, may have borderline intellectual functioning. There
has been no evidence for an increased risk of birth defects or any other problems in infants born to fathers with
PKU.
Pregnancies have been reported in mothers with a variety of other inherited metabolic disorders, including
several forms of glycogen storage disease, propionic acidemia, isovaleric acidemia, homocystinuria, hereditary
orotic aciduria, and several others, with no adverse outcomes clearly attributable to the maternal disorder. The
collaborative experience with many disorders, however, is limited to single cases or small numbers of patients.
It is probable that other maternal metabolic disorders will be identified that adversely affect fetal development.
REFERENCES
1. Scriver CR, Beaudet AL, Valle D, et al, eds. The metabolic and molecular basis of inherited disease, 8th ed.
New York: McGraw-Hill, 2001.
2. Rimoin DL, Conner JM, Pyeritz RE, et al, eds. Emery and Rimoin's principles and practice of medical
genetics, 4th ed. New York: Churchill Livingston, 2002.
3. Fischer AQ, Challa VR, Burton BK, et al. Cerebellar hemorrhage complicating isovaleric acidemia: a case
report. Neurology 1981;31:746.
4. Nyhan WL. Patterns of clinical expression and genetic variation in the inborn errors of metabolism. In:
Nyhan WL, ed. Heritable disorders of amino acid metabolism. New York: John Wiley and Sons, 1974.
5. Ballard RA, Vinocur B, Reynolds JW, et al. Transient hyperammonemia of the preterm infant. N Engl J Med
1978;299:920.
6. Msall M, Batshaw ML, Suss R, et al. Neurologic outcome in children with inborn errors of urea synthesis. N
Engl J Med 1984;310:1500.
7. Goldberg RN, Cabal LA, Sinatra FR, et al. Hyperammonemia associated with perinatal asphyxia. Pediatrics
1979;64:336.
8. Batshaw ML, Wachtel RC, Cohen L, et al. Neurologic outcome in premature infants with transient
asymptomatic hyperammonemia. J Pediatr 1986;108:271.
9. Robinson BH, Taylor J, Sherwood WG. The genetic heterogeneity of lactic acidosis: occurrence of
recognizable inborn errors of metabolism in a pediatric population with lactic acidosis. Pediatr Res
1980;14:956.
10. Robinson BH, Glerum DM, Chow W, et al. The use of skin fibroblast cultures in the detection of respiratory
chain defects in patients with lactic acidemia. Pediatr Res 1990;28:549.
11. Dalla Bernardina B, Aicardi J, Goutieres F, et al. Glycine encephalopathy. Neuropadiatrie 1979;10:209.
12. Wiegand C, Thompson T, Bock GH, et al. The management of life-threatening hyperammonemia: a
comparison of several therapeutic modalities. J Pediatr 1980;96:142.
13. Batshaw ML, Brusilow SW, Waber L, et al. Treatment of inborn errors of urea synthesis: activation of
alternative pathways of waste nitrogen synthesis and excretion. N Engl J Med 1982;306:1387.
P.979
14. Matsubara Y, Narisawa K, Tada K, et al. Prevalence of K329E mutation in medium-chain acyl-CoA
dehydrogenase gene determined from Guthrie cards. Lancet 1991;1:552.
15. Ziadeh R. Medium chain acyl-CoA dehydrogenase deficiency in Pennsylvania: neonatal screening shows
high incidence and unexpected mutation frequencies. Pediatr Res 1995;37:675.
16. Duran M, Hofkamp M, Rhead WJ, et al. Sudden child death and healthy affected family members with
medium-chain acyl coenzyme A dehydrogenase deficiency. Pediatrics 1986;78: 1052.
17. Stanley CA, Hale DE, Coates PM, et al. Medium chain acyl-CoA dehydrogenase deficiency in children with
non-ketotic hypoglycemia and low carnitine levels. Pediatr Res 1983;17:877.
18. Engel AG, Rebouche CJ. Carnitine metabolism and inborn errors. J Inherit Metab Dis 1984;1[Suppl 7]:38.
19. Levy HL, Sepe SJ, Shih VE, et al. Sepsis due to Escherichia coli in neonates with galactosemia. N Engl J
Med 1977;297:823.
20. Kaufman FR, Kogut MD, Donnell GN, et al. Hypergonadotropic hypogonadism in female patients with
galactosemia. N Engl J Med 1981;304:994.
21. Friedman JH, Levy HL, Boustany RM. Late onset of distinct neurologic syndromes in galactosemic siblings.
Neurology 1989;39:741.
22. Cutz E, Cox DW. Alpha1-antitrypsin deficiency: the spectrum of pathology and pathophysiology. Perspect
Pediatr Pathol 1979;5:1.
23. Lindbland B, Lindstedt S, Stein G. On the enzymic defects in hereditary tyrosinemia. Proc Natl Acad Sci U
S A 1977;74:4641.
24. Kvittingen EA, Halvorsen S, Jellum E. Deficient fumarylacetoacetate fumarylhydrolase activity in

lymphocytes and fibroblasts from patients with hereditary tyrosinemia. Pediatr Res 1983;14:541.
25. Karon M, Imach D, Schwartz A. Effective phototherapy in congenital non-obstructive non-hemolytic

jaundice. N Engl J Med 1970;282:377.
26. Gorodischer R, Levy G, Krasner J, et al. Congenital non-obstructive non-hemolytic jaundice: effect of
phototherapy. N Engl J Med 1970;282:375.
27. Schutgens RB, Heymans HS, Wanders RJ, et al. Peroxisomal disorders: a newly recognized group of
genetic diseases. Eur J Pediatr 1986;144:430.
28. Wilson GN, Holmes RD, Hajra AK. Peroxisomal disorders: clinical commentary and future prospects. Am J
Med Genet 1988;30:771.
29. Sweetman L, Nyhan WL, Trauner DA, et al. Glutaric aciduria type II. J Pediatr 1980;96:1020.
30. Chalmers RA, Tracy BM, King GS, et al. The prenatal diagnosis of glutaric acidemia type II using
quantitative gas chromatography-mass spectroscopy. J Inherit Metab Dis 1985;2:145.
31. Robinson BH, McMillan H, Petrova-Benedict R, et al. Variable clinical presentation in patients with
deficiency of pyruvate dehydrogenase complex. A review of 30 cases with a defect in the E component of the
complex. J Pediatr 1987;111:525.
32. Opitz JM, de la Cruz F. Cholesterol metabolism in the RSH/Smith-Lemli-Opitz syndrome: summary of an
NICHD conference. Am J Med Genet 1994;50:326.
33. Dobyns WB. Agenesis of the corpus callosum and gyral malformations are frequent manifestations of
nonketotic hyperglycinemia. Neurology 1989;39:817.
34. Wick H, Schweizer KK, Baumgartner R. Thiamine dependency in a patient with congenital lactic acidemia
due to pyruvate dehydrogenase deficiency. Agents Actions 1977;7:405.
35. Wilcken B, Wiley V, Hammond J, et al. Screening newborns for inborn errors of metabolism by tandem
mass spectrometry. N Engl J Med 2003;348:2304.
36. Lenke RR, Levy HL. Maternal phenylketonuria and hyperphenylalaninemia. An international survey of
untreated and treated pregnancies. N Engl J Med 1980;303:1202.
37. Levy HL, Waisbren SE. Effects of untreated maternal phenylketonuria and hyperphenylalaninemia on the
fetus. N Engl J Med 1983;309:1269.
38. Koch R, Friedman E, Azen C, et al. The International Collaborative Study of Maternal Phenylketonuria:
status report 1998. Eur J Pediatr 2000;[Suppl 2]:S156.
39. Fishler K, Koch R, Donnell GN, et al. Developmental aspects of galactosemia from infancy to childhood.
Clin Pediatr 1980;19:38.
40. Baerlocher K, Gitzelmann R, Steinmann B, et al. Hereditary fructose intolerance in early childhood: a
major diagnostic challenge. Survey of 20 symptomatic cases. Helv Paediatr Acta 1978;33:465.
41. Pagliara AS, Karl IE, Keating JP, et al. Hepatic fructose-1,6-diphosphatase deficiency. A cause of lactic
acidosis and hypoglycemia in infancy. J Clin Invest 1972;51:2115.
42. Chen Y-T. Glycogen storage diseases. In: Harrison's principles of internal medicine, 14th ed. New York,
McGraw Hill, 1998:2176.
43. Huijing F, van Creveld S, Losekoot G. Diagnosis of generalized glycogen storage disease (Pompe's
disease). J Pediatr 1963;63:984.
44. Levin B, Burgess EA, Mortimer PE. Glycogen storage disease type IV: amylopectinosis. Arch Dis Child
1968;43:548.
45. Clow CL, Reade TM, Scriver CR. Outcome of early and long-term management of classical maple syrup
urine disease. Pediatrics 1981;68:856.
46. Mudd SH, Skovby F, Levy HL, et al. The natural history of homocystinuria due to cystathionine betasynthase deficiency. Am J Hum Genet 1985;37:1.
47. Baumgartner R, Ando T, Nyhan WL. Nonketotic hyperglycinemia. J Pediatr 1969;75:1022.
48. Smith I, Wolff OH. Natural history of phenylketonuria and influence of early treatment. Lancet 1974;2:540.
49. Kvittingen EA. Hereditary tyrosinemia type Ian overview. Scand J Clin Lab Invest 1986;46:27.
50. Fell V, Pollitt RJ, Sampson GA, et al. Ornithinemia, hyperammonemia and homocitrullinuria. A disease
associated with mental retardation and possibly caused by defective mitochondrial transport. Am J Dis Child
1974;127:752.
51. Simell O, Perheentupa J, Rapola J, et al. Lysinuric protein intolerance. Am J Med 1975;59:229.
52. Fowler B. Genetic defects of folate and cobalamin metabolism. Eur J Pediatr 1998;157:S60.
53. Scriver CR, Whelan DT. Glutamic acid decarboxylase (GAD) in mammalian tissue outside the central
nervous system, and its possible relevance to hereditary B6 dependency with seizures. Ann N Y Acad Sci
1969;166:83.
54. Matsui SM, Mahoney MJ, Rosenberg LE. The natural history of the inherited methylmalonic acidemias. N
Engl J Med 1983;308:857.
55. Mitchell GA, Watkins D, Melancon SB, et al. Clinical heterogeneity in cobalamin C variant of combined
homocystinuria and methylmalonic aciduria. J Pediatr 1986;108:410.
56. Wolf B, Hsia YE, Sweetman L, et al. Propionic acidemia: a clinical update. J Pediatr 1981;99:835.
57. Newman CGH, Wilson BDR, Callaghan P, et al. Neonatal death associated with isovaleric acidemia. Lancet
1967;2:439.
58. Finnie MDA, Cottrall K, Seakins JWT, et al. Massive excretion of 2-oxoglutaric acid and 3-hydroxyisovaleric
acid in a patient with a deficiency of 3-methylcrotonyl-CoA carboxylase. Clin Chim Acta 1976;73:513.
59. Burri BJ, Sweetman L, Nyhan WL. Heterogeneity of holocarboxylase synthetase in patients with biotinresponsive multiple carboxylase deficiency. Am J Hum Genet 1985;37:326.
60. Wolf B, Heard GS, Weissbecker KA, et al. Biotinidase deficiency: initial clinical features and rapid
diagnosis. Ann Neurol 1985;18:614.
61. Leibel RL, Shih VE, Goodman SI, et al. Glutaric acidemia: a metabolic disorder causing progressive
choreoathetosis. Neurology 1980;30:1163.
62. Goodman SI, Stene DO, McCabe ERB, et al. Glutaric acidemia type II: clinical, biochemical and
morphologic considerations. J Pediatr 1982;100:946.
63. Mantagos S, Genel M, Tanaka K. Ethylmalonic-adipic aciduria: in vivo and in vitro studies indicating
deficiency of activities of multiple acyl-CoA dehydrogenases. J Clin Invest 1979;64: 1580.
64. Wysocki SJ, Hahnel R. 3-Hydroxy-3-methylglutaryl-CoA lyase deficiency: a review. J Inherit Metab Dis
1986;9:225.
65. Robinson BH, Sherwood WG, Taylor J, et al. Acetoacetyl CoA thiolase deficiency: a cause of severe
ketoacidosis in infancy simulating salicylism. J Pediatr 1979;95:228.
66. Hoffmann G, Gibson KM, Brandt IK, et al. Mevalonic aciduriaan inborn error of cholesterol and nonsterol
isoprene biosynthesis. N Engl J Med 1986;314:1610.
67. Hagenfeldt L, Larsson A, Zetterstrom R. Pyroglutamic aciduria. Studies of an infant with chronic metabolic
acidosis. Acta Paediatr Scand 1974;63:1.
P.980
68. Ko FJ, Nyhan WL, Wolff J, et al. 3-Hydroxyisobutyric aciduria: an inborn error of valine metabolism.
69. Kelley RI, Cheatham JP, Clark BJ, et al. X-linked dilated cardiomyopathy with neutropenia, growth
retardation, and 3-methylglutaconic aciduria. J Pediatr 1991;119:738.
70. Gibson KM, Burlingame TG, Hogema B, et al. 2-methylbutyryl-Coenzyme A dehydrogenase deficiency: a
new inborn error of L-isoleucine metabolism. Pediatr Res 2000;47:830.
71. Hudak ML, Jones MD Jr, Brusilow SW. Differentiation of transient hyperammonemia of the newborn and
urea cycle enzyme defects by clinical presentation. J Pediatr 1985;107:712.
72. Cederbaum SD, Shaw KNF, Valente M. Hyperargininemia. J Pediatr 1977;90:569.
73. Bachmann C, Krahenbiihl S, Colombo JP, et al. N-acetylglutamate synthetase deficiency: a disorder of
ammonia detoxification. N Engl J Med 1981;304:543.
74. Amendt BA, Greene C, Sweetman L, et al. Short chain acyl-CoA dehydrogenase deficiency: clinical and
biochemical studies in two patients. J Clin Invest 1987;79:1303.
75. Stanley CA. New genetic defects in mitochondrial fatty acid oxidation and carnitine deficiency. Adv Pediatr
1987;34:59.
76. Strauss AW. Molecular basis of human mitochondrial very-long-chain acyl-CoA dehydrogenase deficiency
causing cardiomyopathy and sudden death in childhood. Proc Natl Acad Sci U S A 1995;92:10496.
77. Sewell AC. Long chain 3-hydroxyacyl-CoA dehydrogenase deficiency: a severe fatty acid oxidation
disorder. Eur J Pediatr 1994;153:745.
78. Chalmers RA, Stanley CA, English N, et al. Mitrochondrial carnitine-acylcarnitine translocase deficiency
presenting as sudden neonatal death. J Pediatr 1997;131:220.
79. Rinaldo P, Stanley CA, Hsu BYL, et al. Sudden neonatal death in carnitine transporter deficiency. J Pediatr
1997;131:304.
80. Hug G, Bove KE, Soukup S. Lethal neonatal multiorgan deficiency of carnitine palmitoyl-transferase II. N
Engl J Med 1991;325:1862.
81. Nguyen TV, Andresen BS, Corydon TJ, et al. Identification of isobutyryl-CoA dehydrogenase and its
deficiency in humans. Mol Genet Metab 2002;77:68.
82. Robinson BH, Oei J, Sherwood WG, et al. The molecular basis for the two different clinical presentations of
classical pyruvate carboxylase deficiency. Am J Hum Genet 1984;36:283.
83. Clayton PT, Hyland K, Brand M, et al. Mitochondrial phosphoenolypyruvate carboxikinase deficiency. Eur J
Pediatr 1986;145:46.
84. Munnich A, Rustin P. Clinical spectrum and diagnosis of mitochondrial disorders. Amer J Med Genet
2001;106:4.
85. The Cystic Fibrosis Genotype-Phenotype Consortium. Correlation between genotype and phenotype in
patients with cystic fibrosis. N Engl J Med 1993;329:1308.
86. Stevenson RE, Lubinsky M, Taylor HA, et al. Sialic acid storage disease with sialuria: clinical and
biochemical features in the severe infantile type. Pediatrics 1983;72:441.
87. Scriver CR, Mahon B, Levy HL, et al. The Hartnup phenotype: mendelian transport disorder, multifactorial
disease. Am J Hum Genet 1987;40:401.
88. O'Brien JS. Generalized gangliosidosis. J Pediatr 1969;75:167.
89. Barranger JA, Murray GJ, Ginns EI. Genetic heterogeneity of Gaucher's disease. In: Barranger JA, Brady
RO, eds. Molecular basis of lysosomal storage disorders. New York: Academic Press, 1984:311.
90. Besley GT, Elleder M. Enzyme activities and phospholipid storage patterns in brain and spleen samples
from Niemann-Pick disease variants: a comparison of neuropathic and non-neuropathic forms. J Inherit Metab
Dis 1986;9:59.
91. Funk JK, Filling-Katz MR, Sokol J, et al. Clinical spectrum of Niemann-Pick disease type C. Neurology
1989;39:1040.
92. Autio S, Louhimo T, Helenius M. The clinical course of mannosidosis. Ann Clin Res 1982;14:93.
93. Dawson G, Spranger JW. Fucosidosis: a glycosphingolipidosis. N Engl J Med 1971;285:122.
94. Antonarakis SE, Valle D, Moser HW, et al. Phenotypic variability in siblings with Farber disease. J Pediatr
1984;104:406.
95. Young LW, Sty JR, Babbitt JP. Wolman's disease. Am J Dis Child 1979;133:959.
96. Clarke JTR, Ozere RL, Krause VW. Early infantile variant of Krabbe globoid cell leukodystrophy with lung
involvement. Arch Dis Child 1981;8:640.
97. Spranger JW, Koch F, Mekusick VA, et al. Mucopolysaccharidosis VI (Maroteaux-Lamy's disease). Helv
Paediatr Acta 1970;25:337.
98. Nelson A, Peterson L, Frampton B, et al. Mucopolysaccharidosis VII (beta-glucuronidase deficiency)

presenting as nonimmune hydrops fetalis. J Pediatr 1982;101:574.
99. Leroy JG, Spranger JW, Feingold M, et al. I-cell disease: a clinical picture. J Pediatr 1971;79:360.
100. Amir N, Zlotogora J, Bach G. Mucolipidosis type IV: clinical spectrum and natural history. Pediatrics
1987;79:953.
101. Burk RD, Valle D, Thomas GH, et al. Early manifestations of multiple sulfatase deficiency. J Pediatr
1984;104:574.
102. Aylsworth AS, Thomas GH, Hood JL, et al. A severe infantile sialidosis: clinical, biochemical and
microscopic features. J Pediatr 1980;96:662.
103. Wilson GN, Holmes RD, Hajra AK. Peroxisomal disorders: clinical commentary and future prospects. Am J
Med Genet 1988; 30:771.
104. Danks DM, Stevens BJ, Campbell PE, et al. Menkes kinky hair syndrome: an inherited defect in copper
absorption with widespread effects. Pediatrics 1972;50:188.
105. Arnold GL, Greene CL, Stout JP, et al. Molybdenum cofactor deficiency. J Pediatr 1993;4:595.
106. Mudd SH, Irreverre F, Laster L. Sulfite oxidase deficiency in man: demonstration of the enzymatic defect.
Science 1967;156:1599.
107. Verlos A, Temple IK, Hubert A-F, et al. Recurrence of neonatal haemochromatosis in half sibs born of
unaffected mothers.J Med Genet 1996;33:444.
108. White PC, New MI, Dupont B. Congenital adrenal hyperplasia. N Engl J Med 1987;316:1580.
109. Jaeken J, Stibler H, Hagberg B. The carbohydrate-deficient glycoprotein syndrome: a new inherited
multisystemic disease with severe central nervous system involvement. Acta Paediatr Scand Suppl
1991;375:1.
110. McClard RW, Black MJ, Jones ME, et al. Neonatal diagnosis of orotic aciduria: an experience with one
family. J Pediatr 1983;102:85.
111. Kozlowski K, Sutcliffe J, Barylak A, et al. Hypophosphatasia: review of 24 cases. Pediatr Radiol
1976;5:103.
112. Berk PD, Jones EA, Howe RB, et al. Disorders of bilirubin metabolism. In: Bondy, PK, Rosenberg LE, eds.
Metabolic control and disease, 8th ed. Philadelphia: WB Saunders, 1980:1009.
113. Sveger T. Liver disease in alpha1-antitrypsin deficiency detected by screening of 200,000 infants. N Engl
J Med 1976;294:1316.
114. Matalon R, Michals K, Sebesta D, et al. Aspartoacylase deficiency and N-acetylaspartic aciduria in
patients with Canavan disease. Am J Med Genet 1988;29:463.
115. Shapiro LJ, Weiss R, Buxman MM, et al. Enzymatic basis of typical X-linked ichthyosis. Lancet
1978;2:756.
116. Cruysberg JRM, Sengers RCA, Pinckers A, et al. Features of a syndrome with congenital cataracts and
hypertrophic cardiomyopathy. Am J Ophthalmol 1986;102:740.
117. Charmas LR, Bernardini I, Rader D, et al. Clinical and laboratory findings in the oculocerebrorenal
syndrome of Lowe, with special reference to growth and renal function. N Engl J Med 1991;324:1318.
Chapter 42
RENAL DISEASE
Suhas M. Nafday M.D., M.r.C.P.I., F.A.A.P. Lisa M. Satlin M.D. Corinne Benchimol D.O.
Luc P. Brion M.D.Chester M. Edelmann Jr. M.D.
Developmental Physiology
The kidneys play a central role in the physiologic transition from fetal to postnatal
life. Whereas homeostasis in utero is maintained largely by the placenta,
adaptation to the extra uterine environment requires that the kidneys assume the
responsibility of regulating water and solute balance. Although the neonatal kidney
traditionally has been characterized as dysfunctional, closer analysis indicates that
this organ functions at a level that is appropriate to the physiologic needs of the
growing infant, except in the very-low-birth-weight (VLBW) infant.
Embryology
The metanephros, the definitive mammalian kidney, first appears at 5 weeks
postconceptional age and begins to produce urine by 10 weeks postconceptional
age (1,2 and 3). The ureteric bud, an offshoot of the mesonephric duct, induces
formation of nephrons within the metanephric blastema. The nephroblastic cells of
the blastema, on contact with the ureteric bud, differentiate into the glomerulus,
proximal convoluted tubule, loop of Henle, and distal convoluted tubule. The
ureteric bud ultimately forms the ureter, pelvis, calyces, and collecting ducts of the
metanephros.
The processes of induction, morphogenesis and differentiation of the metanephros
occurs in a centrifugal pattern, proceeding from the center to the periphery (4,5
and 6). Thus, the first nephrons to develop are those residing in the
juxtamedullary region, whereas the youngest are located in the nephrogenic zone
just under the renal capsule.
The full complement of approximately 1 million neph-rons per kidney in the human
is achieved by about 35 weeks' gestational age (GA), or a body weight of about
2,300 g (6). When birth occurs before this age, the formation of new nephrons
(nephronogenesis) continues until a full complement is achieved. Once complete,
nephronogenesis is never resumed, even after extensive loss of renal tissue. Thus,
the full-term infant is born with as many nephrons as he or she will ever have.
Physiologic, biochemical, and enzymatic maturation of newly formed nephrons
may lag behind anatomic maturation by weeks or months. Thus, the immature
kidney is characterized by structural and functional heterogeneity arising from the
concurrent presence of nephrons in varied stages of differentiation.
Renal Physiology
Urine produced by the fetal metanephric kidneys contributes to the formation of
amniotic fluid (3,6). Although renal function is not necessary for long-term
regulation of fetal water and electrolyte homeostasis, it is important for normal

growth (7), lung development (8), and protection from acute changes in fetal
vascular volume (9). The hourly rate of urine production in normal fetuses is about
5 mL at 20 weeks, 10 mL at 30 weeks, increasing to up to 50 mL at 40 weeks
(3,10).
Renal Blood Flow

Low rates of fetal and neonatal renal blood flow (RBF) exist in a variety of animal
species, whether normalized to body weight, surface area or kidney weight. RBF in
the human fetus, estimated by Doppler ultrasonography, increases from 20 mL/
min at 25 weeks GA to more than 60 mL/min at 40 weeks and reaches adult levels
by 2 years of age (11,12). The effective renal plasma flow (RPF) has traditionally
been calculated from the renal clearance of the organic acid para-aminohippurate
(PAH), corrected for the fraction of PAH extracted by the kidney (12); the value for
effective RPF divided by (100- hematocrit [%])/100 provides an estimate of RBF.
The effective RPF increases rapidly between 30 and 40 weeks GA, reaching adult
values by 24 months of postnatal life (12) (Fig. 42-1).
Because the renal extraction ratio for PAH varies with age, calculated values of
effective RPF must be interpreted with caution in the fetus and neonate. The rate
of extraction of PAH in the 1-week-old full-term newborn is only about 60%,
reaching the adult value of 90% by 5 months (13). PAH clearance averages 148
mL/min/1.73 M2 body surface
P.982
area (BSA) in full-term infants at 2 weeks of life, increasing to 200 mL/min/1.73
M2 by 3 months (14). PAH clearance, corrected for BSA, reaches adult levels
sometime between 1 and 2 years of age (12). The incomplete PAH extraction early
in life appears to be as a result of relatively greater blood flow to juxtamedullary
nephrons and efferent arteriovenous shunting in cortical nephrons, thereby
allowing blood to bypass PAH-transporting segments of the nephron. Additionally,
the organic acid secretory pathways in the proximal tubule are immature early in
development (15). Thus, RBF and RPF will be underestimated by PAH clearance
alone in neonates, particularly in premature infants, in the absence of
determination of the PAH extraction ratio.
Figure 42-1 Developmental changes in effective renal plasma flow (ERPF),

calculated from the renal clearance of para-aminohippurate (PAH) and corrected
for the fraction of PAH extracted by the kidney. The ERPF, used to estimate RBF,
increases rapidly between 30 and 40 weeks GA, reaching adult values by 24
months of postnatal life (From ref. 12, with permission).
RBF is determined by cardiac output and the ratio of renal to systemic vascular
resistance. Developmental changes in both hemodynamic variables contribute to
the postnatal increase in RBF. Adult kidneys (which comprise 0.5% of total body
mass) receive approximately 20% to 25% of the total cardiac output,
corresponding to a RBF of 4 mL/min/g kidney weight. In contrast, the previable
fetus receives only about 5% of the cardiac output, and the 1-week-old term
infant receives about 9% (16).
The perinatal increase in RBF is gradual (13,17,18), consistent with the premise
that clamping of the umbilical cord and the immediate imposition of functional
demands at birth do not in themselves account for the postnatal increase in RBF.
The maturational increase in RBF cannot be completely explained by increases in
renal mass (19) or by nephronogenesis, which is complete well before maximal
levels of RBF are achieved.
The intrarenal distribution of RBF in the newborn differs from that in the adult,
reflecting the relative size, number, and maturity of glomeruli present in the
different regions of the kidney at that stage of development. RBF in the newborn is
distributed primarily to the inner cortex and medulla. Maturation is accompanied
by a redistribution of blood flow toward the superficial cortex, so that the ratio of
inner cortical to outer cortical blood flow becomes progressively smaller than in the
fetus (18,20,21 and 22) (Fig. 42-2). At maturity, about 93% of RBF goes to the
cortex, which constitutes about 75% of the renal mass, whereas only 7% is
distributed to the renal medulla and perirenal fat.
Figure 42-2 Postnatal changes in the intrarenal distribution of blood flow.

Relative rates of blood flow per glomerulus in the four cortical zones of the canine
kidney. Zone I represents the most superficial region, and zone IV represents the
deepest. The total height of the bars in each age group is equal. At birth, the
blood flow to the superficial cortex was lowest, with most blood flow perfusing the
deep cortex. By 6 weeks of age, this pattern was reversed. Maturation is
accompanied by an increase in blood flow to the outer cortex, due primarily to a
decrease in renal vascular resistance (From ref. 17, with permission).
The primary factor responsible for the maturational increase and change in
intrarenal distribution in RBF is a decrease in renal vascular resistance (RVR) (23),
with the rise in cardiac output and mean arterial blood pressure (MAP) only
partially accounting for these postnatal changes in renal hemodynamics (23,24
and 25,27). The balance of afferent and efferent arteriolar resistances ultimately
determines the RVR, RBF, glomerular capillary pressure and glomerular filtration
rate (GFR). The intrarenal vascular resistance, localized both at the afferent and
efferent arterioles (26), is much higher in the newborn than in the adult (19,23).
The postnatal fall in RVR occurs at a time when the systemic vascular resistance
increases about 6-fold (23).
Anatomic factors also contribute to the developmental increase and redistribution
of RBF. The intrarenal vascular system distal to the afferent arteriole in the
neonate differs from that in the adult. Variability in the complexity of the
glomerular capillary network exists early in postnatal life (28). Inner cortical
glomeruli at this age generally have a smaller number of capillaries than the in
adult, although
P.983
they appear similar in overall structure. Few efferent arterioles descend into the
medulla; most connect directly to the venous system (arteriovenous shunting)
(29), bypassing the proximal tubules (26).
A variety of vasoactive substances have been implicated in the developmental
regulation of RBF and GFR. The most important factors are discussed below.
Changes in the balance of the vasoconstrictor renin-angiotensin system (RAS) and
renal sympathetic nervous systems, both of which are highly active in early
development, and vasodilatory humoral factors, including nitric oxide (NO), are
believed to mediate, at least in part, the developmental reduction in RVR and
increase in RBF.
Renin-Angiotensin System
All components of the RAS are present in the fetal metanephric kidney
(30,31,32,33,34) (Fig. 42-3). Renin, angiotensinogen and angiotensin-converting
enzyme (ACE) are first detected in fetal metanephros by ~6 weeks gestation. The
site of localization of renin within the kidney is developmental-stage specific.
Whereas the majority of renin-containing cells are located in the juxtamedullary
apparatus in both the newborn and adult, renin message and protein in the fetus
are also present in the arcuate and interlobular arteries and glomeruli (35).
The RAS is very active in the fetus and newborn. The fetus produces renin as early
as 17 weeks GA (36). Plasma renin activity (PRA) is inversely related to GA in the
fetus and newborn, decreasing from 60 ng/mL/h at 30 weeks to about 10 to 20 ng/
mL/h at term (37). PRA is 3-to-5 times higher in human infants
(38,39,40,41,42,43 and 44) than in adults. Changes in PRA in the fetus in
response to several stimuli (increase after volume depletion or hypoxia, and
decrease after volume expansion or after administration of a -adrenergic
antagonist or of a prostaglandin synthetase inhibitor [PGSI]) are similar to those
observed in adults.
The high levels of PRA in neonates are generally associated with circulating levels
of angiotensin II (ANG II) and aldosterone that exceed those measured in the
Figure 42-3 Relationship between the renin-angiotensin and kinin systems. See
text for details.
adult (45,46,47,48,49 and 50). These high levels may reflect high rates of
secretion or low metabolic clearance rates relative to body size. The reason for
these high levels may be a relative end-organ unresponsiveness to aldosterone
(51,52). Circulating levels of plasma ANG II decrease during postnatal life in
parallel with PRA (53).
The observation that administration of ACE inhibitors during late gestation can
result in anuria and oligohydramnios, renal tubular abnormalities, pulmonary
hypoplasia, growth retardation, and increased fetal loss (54,55,56) underscores
the pivotal role of the RAS in normal fetal growth and development. Genetic
disruption of the genes encoding angiotensinogen (57), ACE (58), or ANG II type 1
re-ceptor is also associated with profound abnormalities in kidney morphology.
SThe significance of the RAS in the maintenance of renal hemodynamics in the
perinatal period remains uncertain. However, as identified above, a number of
observations suggest that this axis may be important under conditions of stress in
the near-term fetus. An acute reduction in blood volume or onset of hypoxia in the
fetus or neonate significantly increases renin and ANG II levels (10,59,60). ACE
inhibition blocks the increase in RVR and reduction in RBF experienced by nearterm fetuses in response to hemorrhage (61,62).
Kinins
Bradykinins are potent vasodilator peptides generated from the protein precursor
kininogen by the proteolytic enzyme kallikrein (Fig. 42-3). The kinins are
inactivated by kininase I, which is a carboxypeptidase, and the peptidyl dipeptide
hydrolase kininase II, which is ACE. Thus, ACE inhibitors not only decrease ANG II
production but also prevent the breakdown of kinins.
Premature infants have undetectable levels of urinary kinins (63); urinary
excretion of kallikreins and kinins is lower in newborns than older children (64).
The role of these substances in modulating the function of the immature kidney
remains unclear.
Prostaglandins
Prostaglandins (PGs) contribute to the maintenance of RBF and GFR, especially
during conditions of enhanced
P.984
vasoconstrictor activity. Complex interactions between PGs and the RAS and kinin
systems exist, making it difficult to identify the specific effects of PGs on regulation
of blood pressure, RBF, and electrolyte and water homeostasis. PG synthesis from
arachidonic acid requires cyclooxygenase (COX), the inhibitory target of aspirin
and various non-steroidal antiinflammatory drugs (65). Two isoforms of COX have
been identified, each representing a different gene product and subject to
differential regulation. COX-1 has been proposed to participate in

glomerulogenesis (66), whereas COX-2 regulates renal perfusion and glomerular
hemodynamics (67). Differences in intrarenal COX-1 and -2 localization between
the adult and fetal human kidney may account for the variable renal responses to
PG inhibition observed between the two groups (66).
Urinary excretion of PGs is high in the fetus (68,69), presumably reflecting a high
rate of renal synthesis. Urinary excretion of PGE2 and prostacyclin metabolites in
the premature infant is five times that noted at term and 20 times that measured
in older children (70). Although rates of excretion decrease after birth, urinary PG
excretion remains significantly higher in neonates than in children or adults
(71,72).
PGs produced by the fetal and neonatal kidney may contribute to the regulation of
renal perfusion and glomerular hemodynamics. Maternal administration of
indomethacin, a nonselective COX inhibitor, reduces RBF in the fetus, and is
associated with an increase in RVR and MAP (73). Intrauterine exposure of the
human fetus to COX inhibitors can lead to a decrease in GFR with consequent
reduction in fetal urine output and oligohydramnios (74,75,76,77). Postnatal
administration of a PGSI to preterm infants (78,79) to close a patent ductus
arteriosus (PDA) may also compromise renal function, leading to a reduction in
RBF, GFR and urine volume.
Renal Nerves and the Adrenergic System

The renal vascular bed of the fetus is less reactive to renal nerve stimulation than
that of the newborn and adult (80,81). In contrast to observations made in adult
animals, several studies have demonstrated that renal nerve stimulation in the
fetal and newborn kidney activates noradrenergic pathways, norepinephrine
stimulates 2-adrenoceptors to produce renal vasodilation, and maturation of the
renal adrenergic system is associated with downregulation of the renal adrenoceptor response (80,81 and 82).
Circulating catecholamine levels, particularly norepinephrine, are very high just
before and immediately after birth (83), falling to adult values within a few days of
life. The high plasma levels of catecholamines act directly to increase afferent
arteriolar tone and indirectly, via stimulation of renin and ANG II release, to
increase efferent resistance, possibly contributing to the maintenance of the high
RVR characteristic of the neonatal kidney (84). The fetal and neonatal kidney
demonstrates enhanced sensitivity to catecholamines compared to the adult,
related in part to developmental differences in adrenergic receptor density (85,86).
Nitric Oxide
Endogenous NO is a major vasodilator in the developing kidney, contributing to the
maintenance of intrarenal vascular tone and serving to buffer the highly activated
endogenous vasoconstrictor state (87,88). The sensitivity of renal hemodynamics
in the fetal and neonatal kidney to inhibition of NO synthesis is significantly
greater than that of the adult kidney.
Dopamine and Dopamine Receptors

The fetal and neonatal kidneys are less sensitive than their adult counterparts to
the renal vasodilatory and the natriuretic effects of dopamine (89,90). Exogenous
administration of dopamine, which results in an increase in RBF in the adult, has
no effect in the young animal unless there also is an increase in systemic blood
pressure and cardiac output (81,91,92). This blunted response is considered to
reflect the limited generation of the vasodilatory second messenger cyclic
adenosine monophosphate (cAMP) (89,91) and a low density of renal dopamine-1like receptors in the neonatal kidney (93).
The pharmacological effects of dopamine are dose-dependent. Whereas the renal
response to low concentrations of dopamine generally reflects dopaminergic
receptor-mediated effects, high concentrations stimulate -adrenergic receptors.
Thus, the renal vasoconstriction observed following intrarenal dopamine infusion in
fetal and neonatal animals (94,95) likely reflects dopamine binding to adrenoceptors, which are abundant by term (96,97).
Arginine Vasopressin
The plasma concentration of arginine vasopressin (AVP) in neonates increases
abruptly after birth and is highest in infants whose mothers labored before vaginal
delivery (98). Although AVP has the potential to contribute to the high RVR
characteristic of the neonate through its action on vascular and glomerular V1
receptors, its role in regulating basal renal hemodynamics remains poorly defined.
Infusion of synthetic AVP does not alter RBF and RVR in fetal sheep (99). However,
AVP may play a role in certain stress-induced responses. For example, during
hemorrhage, but not during hypoxemia, the marked decrease in RBF and increase
in RVR correlate closely with the rise in plasma AVP (10,59,100).
Atrial Natriuretic Factor

ANF acts on the mature kidney to increase sodium excretion and GFR, antagonize
renal vasoconstriction, and inhibit renin secretion (101). ANF release from atrial
cardiocytes is stimulated in the fetus by an increase in intra-cardiac pressure and
atrial distention (102,103,104 and 105), and levels fall in response to a decrease
in central venous pressure, as during hemorrhage (106). The natriuretic and
diuretic
P.985
response to systemically infused ANF in newborns is attenuated compared to

adults (107,108,109,110 and 111). Whereas specific ANF receptors have been
identified on near-term fetal glomerular membranes, the ANF binding capacity is
age-dependent, increasing 7-fold between fetal and adult life (112). The blunted
response of the immature subject to ANF has been attributed to ineffective
production of the second messenger cyclic GMP (108,113), a rapid systemic
clearance of ANF (107,108), and a greater ANF-induced reduction in RBF in fetuses
and newborns than in adults (107).
Glomerular Filtration
Initiation of glomerular filtration in the human fetus occurs between 9 and 12
weeks of GA (114,115). It has been suggested that the functional demand placed
on the neonatal kidney by cessation of placental function at birth stimulates GFR
to increase. However, estimates of GFR correlate well with GA, a relationship that
persists whether the fetus remains in utero or is born prematurely (116,117,118
and 119). Specifically, GFR averages approximately 8 to 10 mL/min/ 1.73 M2 at 28
weeks of GA and increases only slightly before 34 weeks postmenstrual age,
although body size and kidney weight increase appreciably during this time. After
about 34 weeks postmenstrual age, at which time GFR averages 25 mL/min/1.73
M2 regardless of postnatal age, GFR increases rapidly, often by threefold to
fivefold within 1 week (116,118,120), coincident with completion of
nephronogenesis (Fig. 42-4). Thus, an infant born prematurely at 28 weeks of GA
shows little increase in GFR until the infant is about 6 weeks old, i.e., until a
postmenstrual age of 34 weeks is attained and nephrogenesis has been completed
(121). The GFR of neonates who are small for GA is similar to that of infants
whose body weight is appropriate for the same GA.
Figure 42-4 Changes in glomerular filtration rate (mL/min), estimated by

creatinine clearance, and nephrogenic activity in the kidney cortex (%) are
plotted as a function of postconceptional age in the human infant. There is a
temporal relationship between the accelerated rate of increase in glomerular
filtration rate and completion of nephrogenesis after 34 weeks of gestation (From
Arant BS. Neonatal adjustments to extrauterine life. In: Edelmann CM Jr, ed.
Pediatric kidney disease. Boston: Little, Brown and Company, 1992:1021).
During the first 4 months of life, GFR increases rapidly relative to body size, kidney
weight, and BSA (11). Thereafter, a slower rise is noted until adult values,
determined on the basis of BSA, are reached by 2 years of age (11,122,123). The
postnatal increase in GFR in premature infants may lag behind that observed in
full-term newborns; at 9 months of age, the GFR of VLBW infants is approximately
70% of that measured in full-term infants of identical postnatal age (121).
Glomeruli are formed in the nephrogenic zone of the kidney. With maturation,
these glomeruli migrate to deeper zones of the renal cortex. At birth, the more
mature glomeruli in the juxtamedullary cortex, which are nearly as large as
glomeruli in the adult kidney, have higher filtration rates than the most recently
formed glomeruli in the superficial cortex, some of which may not begin filtration
for some time. Thus, GFR, similar to RBF, matures centrifugally. As nephrons
enlarge and forces that regulate filtration permit, GFR increases, with most of the
surge as a result of enhanced perfusion of superficial nephrons (18,24,124) (Fig.
42-2), related temporally to the increase in total RBF and its centrifugal
redistribution within the renal cortex.
The addition of newly functioning nephrons is insufficient, however, to account for
the large increases in GFR observed during maturation. A substantial rise in single
nephron GFR occurs (125). The process of urine formation starts with
ultrafiltration of plasma through the glomerular capillary membrane. The rate of
filtration depends on filtration characteristics of the membrane and on the net
ultrafiltration pressure, represented by the difference between the mean
glomerular transcapillary hydraulic, i.e., glomerular capillary minus the Bowman's
space hydraulic pressures and mean colloid osmotic pressures over the length of
the glomerular capillary, and the glomerular plasma flow. The capillary hydrostatic
pressure promotes filtration, whereas both the colloid osmotic pressure within the
capillary and the hydrostatic pressure in the Bowman's space oppose it. The
postnatal increase in GFR may result from changes in all of these variables, in
addition to the surface area available for filtration, and the permeability of the
filtering membrane.
Autoregulation of Renal Blood Flow and

Glomerular Filtration Rate
Autoregulation in the mature kidney allows for constancy of RBF and GFR as the
MAP and renal perfusion pressure vary over a wide range, from about 80 to 150
mmHg. Autoregulation is accomplished primarily by changes in the RVR at the
level of the afferent and efferent arterioles. Although systemic blood pressure in
the fetus (i.e., 40-60 mmHg) and neonate is less than the lower limit of
autoregulatory range defined for adults, experimental evidence suggests that the
fetus and newborn are able to autoregulate RBF efficiently at their prevailing low
arterial pressure,
P.986
albeit at a lower perfusion set point than in the adult (126,127 and 128). In
support of this notion is the observation that a 16% increase in fetal arterial blood
pressure elicited by AVP infusion does not alter RBF (129). The autoregulatory
response is considered to be mediated in part by a myogenic response of the renal
afferent arteriole (130,131), which constricts to limit any rise in glomerular
hydrostatic pressure, and the tubuloglomerular feedback control mechanism
(132,133) (see below). As MAP falls, the renal afferent arteriole dilates and the
efferent arteriole constricts, the latter effect due, at least in part, to stimulation of
renin release and ANG II generation (134,135). The balance of theses two
responses maintains glomerular capillary hydrostatic pressure and preserves RBF
and GFR.
Tubuloglomerular Feedback
Tubuloglomerular feedback serves to maintain a constant rate of water and salt
delivery to distal segments of the nephron in which reabsorption is regulated to
maintain fluid balance. A stimulus (e.g., tubular flow rate, ion concentration) at
the macula densa is transmitted to vascular structures of the nephron that control
GFR to elicit a change of single nephron GFR. Tubuloglomerular feedback is
maximally sensitive in a range that corresponds to normal values of single
nephron GFR and tubular flow rate. As GFR increases with maturation, the
maximal response and flow range of maximum sensitivity also increase, so that
the relative sensitivity of the tubuloglomerular feedback mechanism is unaltered
during growth (136,137). An intact RAS appears to be critical for this signaling
pathway (157) and NO may play a modulatory role (138,139).
Tubular Function
The axial and polarized (apical vs. basolateral) distribution of transport proteins
along sequential segments of the nephron allows the kidney to reabsorb the bulk
of glomerular filtrate proximally and then, in more distal segments, adjust the
solute and water content of the urine to maintain homeostasis. The fully
differentiated kidney is generally a reabsorptive organ. However, transport of
some ions and solutes by individual nephron segments is bi-directional. Thus,
sodium, bicarbonate, phosphate, amino acids, and glucose are reabsorbed,

hydrogen ions are secreted, and potassium and organic acids are both reabsorbed
and secreted. In general, the kidney of the full term, but not necessarily preterm,
neonate is uniquely suited to meet the developmental stage-specific metabolic
demands.
Sodium
Term infants are in a state of positive sodium balance, a requisite for somatic
growth, particularly of bone. Although the sodium intake per unit of BSA is
generally smaller in the newborn than in the adult, the magnitude of this positive
balance remains relatively constant within a wide range of sodium intakes
(140,141). The tendency of the neonatal kidney to retain sodium may become
problematic under conditions of salt loading. Exogenous administration of a sodium
load to an adult is followed by immediate expansion of the extracellular fluid (ECF)
space. This signals the kidney to decrease tubular sodium reabsorption, resulting
in increased excretion of sodium and a relatively rapid return of the ECF space to
the baseline condition. Full-term newborn infants given a sodium load in excess of
12 mEq/kg/d experience a rise in serum sodium levels, abnormal increase in
weight, and generalized edema (142). The limited capacity of the neonatal kidney
to excrete a sodium load compared to its mature counterpart appears to be due
more to factors related to tubular sodium reabsorption than to the low GFR (143).
The fractional excretion of sodium (FENa) is the ratio of the sodium clearance to
the creatinine clearance, expressed as a percent. It is obtained from the following
formula:
in which UNa is the urinary concentration of sodium (mEq/L), Ucr is the urinary
concentration of creatinine (mg/dL), PNa is the plasma sodium concentration (mEq/
L), and Pcr is the plasma creatinine concentration (mg/dL). FENa, which may be as
high as 20% during early fetal life, decreases progressively during gestation
(144,145,146), so that the FENa in the full-term newborn generally averages
about 0.2% (146,147). After the first few hours of postnatal life, the FENa and
urinary sodium excretion decline rapidly, possibly secondary to contraction of the
ECF volume (144). Premature infants of less than 30 weeks of GA continue to
show elevated values of FENa, similar to those observed in the fetus, which may
exceed 5% during the first few days of life (146,148,149). In these infants,
excessive urinary sodium losses exceeding dietary sodium intake (e.g., breast
milk, low-salt formula) often create a state of negative sodium balance and loss of
body weight during the first 2 weeks of life (i.e., hyponatremia of prematurity).
These infants may require at least 2 mEq/kg/d of supplemental sodium to maintain
a normal sodium concentration and remain in positive balance (150,151).
Sodium is freely filtered at the glomerulus. The initial two-thirds of the proximal
tubule of the suckling rat absorbs ~50% of the filtered load of sodium and water
(143,152,153 and 154), values only slightly less that those re-ported in the adult
(152,153,155). Studies in several mammalian species ingesting a normal salt
intake demonstrate parallel and proportionate increases in GFR and the
reabsorptive capacity of the proximal tubule after birth, consistent with
maintenance of glomerulotubular balance during postnatal development
(125,156,157). Functional glomerulotubular imbalance may exist in the premature
infant in whom the reabsorptive capacity of the proximal tubule lags behind the
capacity for glomerular filtration (158,159).
The fractional reabsorption of sodium along the loop of Henle, expressed as a
percentage of load, increases by about 20% during postnatal development (153),
consistent with functional maturation of this nephron segment. Sodium is
P.987
absorbed in this segment through the furosemide- and bumetanide-sensitive Na-K2Cl cotransporter located in the urinary membrane and is extruded from the cell at
the basolateral membrane by the Na-K-ATPase transporter. In contrast, the
fractional reabsorption of sodium along the distal tubule is greater in younger than
in older animals, thereby explaining the sodium retention and blunted response to
sodium loading characteristic of the young animal (143,160).
Clearance studies in human infants (151,152,153,154 and 155,158,
159,161,162,163,164,165 and 166) suggest that the percent of filtered sodium
reabsorbed by the proximal tubule increases by about 5% between 28 and 34 wks
GA, whereas the percent of distal sodium reabsorption increases by more than
15% during this same period. However, because the proximal tubule reabsorbs
more than 70% of the filtered load of sodium, the small percentage increase in
fractional reabsorption in this segment contributes to the postnatal increase in
renal sodium retention as much as the larger percentage increase in the distal
tubule.
Distal sodium reabsorption occurs in the cortical collecting duct (CCD) by apical
sodium entry into principal cells through the amiloride-sensitive sodium channel
ENaC and its extrusion at the basolateral membrane by the Na-K-ATPase. The
avidity of the distal nephron for sodium reabsorption, may be related to the high
levels of aldosterone that prevail in early postnatal life (44). Cellular effects of
aldosterone within the fully differentiated distal nephron (specifically the CCD)
include increases in the density of ENaC channels and stimulation of Na-K-ATPase
activity (167,168 and 169). The net effect of these actions is the stimulation of
sodium absorption.
Plasma aldosterone concentrations in the premature infant and newborn are high
compared to those in the adult (50,170). However, clearance studies in premature
infants (50,171) and investigations in neonatal laboratory animals (172,173)
reveal a blunted responsiveness of the immature kidney to aldosterone. The
density of aldosterone binding sites, receptor affinity, and degree of nuclear
binding of hormone-receptor appear to be similar in mature and immature rats

(18), suggesting that the early hyposensitivity to aldosterone represents a
postreceptor phenomenon. Premature infants can augment their renin secretion in
response to negative sodium balance, but aldosterone production initially fails to
respond adequately to this stimulation (174). The resulting relative
hypoaldosteronism results in an inability to conserve sodium, manifested clinically
by weight loss and hyponatremia.
Paradoxically, preterm infants of about 34 to 36 weeks GA excrete a sodium load
more efficiently than do term newborns, but not as efficiently as adults
(118,148,171). The sodium wasting characteristic of the preterm infant may be as
a result of paucity of ENaC channels in the urinary membrane of the distal nephron
(175).
Urinary sodium excretion during maturation is regulated not only by the RAS, but
also by circulating catecholamines, renal sympathetic innervation, atrial natriuretic
factor (ANF), PGs, dopamine, and glucocorticoids. Direct renal nerve stimulation in
fetal and newborn sheep leads to sodium retention (176), a response qualitatively
similar to that observed in adult animals and attributed to norepinephrine acting
on -adrenergic receptors (177). Studies in newborns indicate a relatively poor
natriuretic response to ANF compared to their adult counterparts (108,178,179).
Renal PGE2, PGD2, and PGI2 reduce sodium and water absorption in the adult,
promoting a natriuresis and diuresis, responses that are blocked by administration
of indomethacin. In contrast, administration of this inhibitor to fetal sheep causes
a natriuresis, despite a decrease in fetal RBF (73). Dopamine inhibits sodium
reabsorption in the nephron by activation of D1-like dopamine receptors. The
natriuretic effect of dopamine is less pronounced in the fetal and neonatal kidney
of many species, including man, compared to the adult as a result of blunted renal
hemodynamic and tubular effects (171,173,180,181 and 182).
Circulating levels of glucocorticoids, including cortisol and corticosterone, surge in
many species during or just before the period of weaning (183,184,185 and 186).
Both endogenous gluco- and mineralocorticoids bind to the mineralocorticoid
receptor with equal affinity (187). Although blood glucocorticoid concentrations are
~100-fold higher than aldosterone concentrations, the metabolism of cortisol into
inactive derivatives by 11-hydroxysteroid dehydrogenase type 2 (11--HSD2)
protects the mineralocorticoid receptor from glucocorticoids (187). The presence of
ENaC, mineralocorticoid receptor and low levels of 11--HSD2 in the CCD suggest
that glucocorticoids may act as sodium retaining steroids during early postnatal life
(188,189).
Potassium
Potassium is transported actively across the placenta from mother to fetus (190).
Indeed, fetal potassium is maintained at levels exceeding 5 mEq/L even in the face
of maternal potassium deficiency (190,191).
Unlike adults, who are in zero balance, growing infants maintain a state of positive
potassium balance (50,192). The relative conservation of potassium early in life
generally is associated with higher plasma potassium values than in the adult
(153,159,192,193). These levels average 5.2 mEq/L from birth to age 4 months,
decreasing to 4.2 mEq/L by 3 years of age (193). The renal clearance of potassium
in the infant is less than in the older child, even when corrected for GFR (193).
Children and adults ingesting a regular diet containing sodium in excess of
potassium excrete urine with a sodium-to-potassium ratio greater than 1, as
expected. Although the sodium-to-potassium ratio of breast milk and commercial
infant formulas averages 0.5, the urinary sodium-to-potassium ratio of the
newborn also is greater than 1, consistent with significant potassium retention.
Infants, like adults, can excrete potassium at a rate that exceeds its glomerular
filtration when given a potassium load, indicating the capacity for net tubular
secretion (194); however, the rate of potassium excretion per unit
P.988
body weight in response to exogenous potassium loading is less in newborn than
older animals (195). Clearance studies in saline-expanded dogs also provide
indirect evidence for a diminished secretory and enhanced reabsorptive capacity of
the immature distal nephron to potassium (196). In general, the limited potassium
secretory capacity of the immature kidney becomes clinically relevant only under
conditions of potassium excess. Under normal circumstances, potassium retention
by the newborn kidney is appropriate and a requirement for growth.
Potassium is freely filtered at the glomerulus. Approx-imately 50% of the filtered
load is reabsorbed along the proximal tubule in both newborns and adults (153).
Up to 40% of the filtered load of potassium reaches the superficial distal tubule of
the newborn, in contrast to about 10% in mature animals, providing evidence for
functional immaturity of the loop of Henle (153,197).
Urinary potassium excretion is derived almost entirely from secretion in distal
segments of the nephron, including the CCD. The low rates of potassium excretion
characteristic of the newborn kidney are due, at least in part, to a low potassium
secretory capacity of this segment (198). In contrast to the high rates of
potassium secretion observed in adult CCDs, neonatal segments show no
significant baseline potassium transport (199). An increase in tubular fluid flow
rate does not stimulate potassium secretion in the neonatal CCD of the rabbit, as it
does in the fully differentiated segment, until after weaning (199,200). Baseline
and flow-stimulated potassium secretion appear to be limited early in life by a
paucity of small-conductance (SK) (201) and calcium-activated maxi-K channels
(200), respectively, in the urinary membrane of the CCD. The developmental
expression of immunodetectable ROMK, the molecular correlate of the SK channel
(202,203), and shortly thereafter, the maxi-K channel, immediately precedes the
appearance of baseline and flow-stimulated potassium secretion in the CCD.
As stated above, fetal and neonatal kidneys are relatively insensitive to

aldosterone. Chronic administration of exogenous aldosterone to fetal sheep
increases sodium absorption and decreases PRA, yet fails to increase urinary
potassium excretion (204). However, administration of the aldosterone antagonist
spironolactone or sodium channel blocker amiloride, both without consistent effect
on so-dium excretion, inhibit renal potassium excretion in fetal sheep, suggesting
that endogenous mineralocorticoids influence renal potassium excretion in utero
(205).
Acid-Base
The acid-base status of the fetus is maintained by placental function and maternal
mechanisms. The fetal kidney in the second one-half of pregnancy, however, is
able to acidify the urine (206,207). Immediately after birth, the acid-base state of
the full-term newborn is characterized by a metabolic acidosis (208); respiratory
compensation generally occurs within 24 hours in the full-term infant (209).
The normal range for serum bicarbonate is lower for preterm infants (16 to 20
mmol/L) and full-term infants (19 to 21 mEq/L) than for children and adults (24 to
28 mmol/L). The lower levels of buffer base concentration in the blood of infants
can be accounted for in part by the inability to completely excrete the byproducts
of growth and metabolism (210,211). The load of endogenous acid generated by
the metabolism of protein and deposition of calcium into the skeleton is larger in
the infant than in the adult (212,213,214,215).
The concentration of bicarbonate in plasma is determined predominately by the
renal bicarbonate threshold, which is lower in preterm and term infants than adults
(194,216,217,218 and 219). The low bicarbonate threshold characteristic of the
newborn is considered to reflect nephron heterogeneity and/or a low fractional
reabsorption of bicarbonate in the immature kidney (50,220), the latter attributed
in part to the relatively large proportion of total body water in the infant.
Postnatal maturation of the proximal tubular capacity for bicarbonate absorption
has been proposed to be as a result of increases in activity of the sodiumhydrogen (Na/H) exchanger (NHE) and H+-ATPase localized to the urinary
membrane of this segment (221,222). Experimental evidence further suggests
that neonatal bicarbonate reabsorption is limited by low activity of carbonic
anhydrase. Carbonic anhydrase facilitates renal acidification by catalyzing the
interconversion of carbon dioxide and water to carbonic acid. The two major renal
isoforms are cytosolic carbonic anhydrase II (95%) and membrane-bound carbonic
anhydrase IV (5%). Carbonic anhydrase activity is pres-ent in the early human
fetal kidney (223,224 and 225). The newborn kidneys of several experimental
animals (226,227 and 228) exhibit less carbonic anhydrase activity than do
mature kidneys. Postnatal increases in carbonic anhydrases II and IV likely
contribute to the developmental increase in bicarbonate absorption observed in the
maturing proximal tubule (226,229).
The renal response to acid loading increases with advancing gestational and
postnatal ages. When compared to adult subjects given a comparable acid load,
the infant exhibits a larger fall in blood potential of hydrogen (pH) and bicarbonate
concentration, a smaller and slower fall in urine pH, and much smaller increase in
urinary titratable acid and ammonium excretion per M2 (230,231,232 and 233).
Premature infants born at 34 to 36 weeks of GA and studied 1 to 3 weeks after
birth exhibit rates of excretion of net acid, titratable acid, and ammonium that are
about 50% lower than term babies of similar postnatal age; net acid excretion
increases to levels observed in term newborns only after 3 weeks of age
(219,233). In response to acid loading with ammonium chloride, urinary pH values
of less than 6 are rarely observed in premature infants until the second month of
life (234). In contrast, by the end of the second postnatal week, urinary pH values
of 5.0 or lower, comparable to those in the adult, are consistently observed in
term infants (235,236).
The capacity of the neonatal kidney for the excretion of net acid is blunted, due in
part to a limited excretion of urinary buffers, including phosphate and ammonium
ions.
P.989
The rates of ammonia synthesis and excretion are low in the neonate (237) and, in
response to acid loading, do not increase to mature values until 2 months of age
(218, 235,238). Phosphate loading, administration of cow milk that is rich in
protein and phosphate instead of breast milk, or high-protein feeding enhances the
ability of the newborn to excrete titratable acids and ammonia (236,239).
The final site of urinary acidification is the renal collecting duct. Functional
immaturity of this segment and the acid-base transporting intercalated cells
therein may further limit the ability of the neonate to eliminate an acid load
(240,241). Postnatal differentiation of intercalated cells has been shown to include
changes in the morphology and function of these specialized cells with an increase
in density.
Up to 10% of preterm infants develop a partially compensated hyperchloremic
metabolic acidosis during weeks 1 to 3 of life, i.e., late metabolic acidosis (218),
despite an otherwise healthy appearance. Typically, spontaneous remission occurs
in the subsequent 2 weeks. These infants are characterized further by an apparent
delay in postnatal weight gain despite ample dietary intake, suggesting a high rate
of endogenous acid formation in infants whose dietary intake exceeds their
anabolic capacity (212). Although infants provided supplemental sodium
bicarbonate to maintain acid-base homeostasis showed greater increases in length
than controls, there were no differences in weight gain between the two groups
(217,242).
Calcium
The state of positive calcium balance characteristic of growing individuals is

sustained by the coordinated interaction of bone, intestine and kidney. Urinary
excretion of calcium varies inversely with GA in the first week of life and varies
directly with urine flow and sodium excretion (243). High rates of calcium
excretion may contribute in part to early neonatal hypocalcemia, which occurs in
the first 24 to 48 hours of life (244). The urinary calcium-to-creatinine ratio in fullterm infants ranges from 0.05 to 1.2 mg/mg during the first week of life, but may
exceed 2 in premature neonates (243,245). In children more than 1 year of age,
the ratio is approximately 0.2, and in adults, the ratio is less than 0.11 (246). The
high fractional excretion of calcium in preterm infants may be related to
maturational changes in the tubular handling of calcium. Approximately 50% of
filtered calcium is reabsorbed along the superficial proximal tubule in mature rats,
yet only 1% of filtered calcium is excreted (153), suggesting that, in the adult,
almost one-half of the filtered calcium is reabsorbed at a site beyond the proximal
tubule or in deep nephrons (153). The fractional reabsorption of calcium in the
loop of Henle, like that for sodium, potassium, and chloride, is low in newborn
rats, increasing significantly with advancing postnatal age (153,247). Absorption in
both proximal tubule and thick ascending limb of the loop of Henle (TALH) is
predominantly coupled to sodium absorption and is a passive process through the
paracellular pathway. Calcium reabsorption within the distal nephron is active,
transcellular and is regulated independently of sodium (248). Furosemide,
methylxanthines, and dexamethasone increase urinary calcium excretion, thus
increasing the risk for nephrocalcinosis and nephrolithiasis (249,250).
The principal hormones that regulate renal calcium excretion in the adult are
parathyroid hormone (PTH), 1,25-dihydroxyvitamin D3 and calcitonin (251). A
reduction in the serum calcium concentration results in the secretion of PTH from
the parathyroid glands, a response leading to the conservation of calcium by the
kidney and the enhanced uptake of calcium by the intestine, which occurs
indirectly through the PTH-induced synthesis of the active vitamin D metabolite
1,25-dihydroxyvitamin D. The net effect is to return the serum calcium back
toward its normal baseline level. In mature animals and in adults, PTH decreases
urinary calcium excretion by stimulating calcium reabsorption across the cortical
TALH and collecting duct (252,253,254,255,256,257 and 258). PTH-responsive
adenylate cyclase has been found in renal cortical homogenates from preterm
rabbits (259) and in the TALH in 2-day-old puppies (260). It has been suggested
that neonatal hypocalcemia may be as a result of end-organ unresponsiveness to
PTH. In premature and full-term newborns, however, the administration of
exogenous PTH increases urinary excretion of cAMP (261,262), results in a
calcemic response, but affects calciuria and phosphaturia only minimally (263).
Vitamin D stimulates calcium reabsorption in the distal tubule via binding to
vitamin D receptors (VDRs). Renal production of 1,25-dihydroxyvitamin D
increases rapidly after birth, provided that the concentration of the substrate, 25hydroxyvitamin D, is adequate (264). VDR are present in cells of branching
ureteral buds by 15 days GA and at later developmental stages in glomerular and

tubular cells (265).
Systemic calcium homeostasis is controlled in large part by the extracellular G
protein-coupled calcium-sensing receptor (CaSR), located on parathyroid and
kidney cells in which it senses the extracellular calcium concentration and in turn
alters the rate of secretion of PTH and renal calcium (and sodium and chloride)
reabsorption in the TALH and early distal nephron (266,267,268 and 269). There is
little expression of CaSR in the fetal kidney (270). Steady state abundance of
CaSR messenger ribonucleic acid (mRNA) and protein increase significantly during
the first week of life, related principally to an increase in expression of the receptor
in the TALH and, to a lesser extent, in the collecting duct (270).
Phosphate
In contrast to most other transport processes, the capacity of the kidney of the
neonate to reabsorb phosphate is far greater than in the adult and progressively
declines with advancing age (271,272). The fractional reabsorption of phosphate
increases from 85% of the filtered load at 28 weeks GA to 99% at term,
decreasing thereafter to approximately 85% between 3 and 20 months of age
(273). The
P.990
high renal reabsorptive capacity for phosphate early in life allows the infant to
retain a large portion of phosphate absorbed from the gut and sustain a state of
positive balance (273), as is required for growth and development. Note that
phosphate is a major constituent of bone, muscle, and membrane phospholipids
and is critical for cellular processes involving ATP. Human neonates and older
infants have a higher serum phosphate concentration than adults.
Studies in experimental animals also demonstrate enhanced renal tubular
reabsorption of phosphate early in life (168,169). Younger subjects have a higher
maximal net reabsorption of phosphate per unit glomerular filtrate than adults
(274,275). The fractional reabsorption of phosphate in the newborn proximal
tubule and distal nephron is higher than that in the adult (276,277). The high
intrinsic rate of sodium phosphate transport measured in neonatal proximal
tubules has been attributed to an abundance of a growth-related sodiumphosphate cotransporter protein in the luminal membrane (278), a high
membrane fluidity of the immature nephron (279), a low intracellular phosphate
concentration (174), and the hormonal milieu prevailing in the perinatal period
(279,280). Nephron heterogeneity may also explain, in part, the limited urinary
phosphate excretion observed in the rapidly growing animal. Because deep
nephrons reabsorb more phosphate than cortical nephrons (281,282), and
nephronogenesis begins in the juxtamedullary region, the kidney of the immature
animal may contain a relatively greater number of functioning nephrons with a
high capacity for phosphate reabsorption.

Renal phosphate transport is regulated by a number of factors, including dietary
intake, PTH, growth hormone, insulin-like growth factor 1, and glucocorticoids
(283,284). The sensitivity of the renal response to these factors is developmentalstage specific. Growth hormone stimulates proximal tubular phosphate transport
(283). The high reabsorptive capacity of the juvenile kidney for phosphate persists
after parathyroidectomy (275,285). Serum levels of PTH are low immediately after
birth, increasing substantially thereafter to levels in infancy exceeding those
observed in later childhood (286,287). Although exogenous PTH administered to
humans or experimental animals results in a blunted response in the immediate
postnatal period (259,274), 1-wk-old full term infants exhibit a prompt
phosphaturic response and an increase in urinary cAMP excretion after PTH
administration (261,262 and 263).
Magnesium
Ninety-seven percent of the filtered magnesium is reabsorbed by the mature
nephron, largely (~60%) in the TALH (288). Micropuncture analysis of magnesium
transport in developing rats shows efficient renal tubular reabsorption of
magnesium in this segment early in life (153). The proximal tubule of the adult
animal reabsorbs only about 10% of the filtered magnesium, whereas that of the
young rat reabsorbs about 60% of the filtered load (153). Postnatal maturation is
associated with a decrease in the fractional reabsorption of magnesium in the
proximal tubule (153). The avid retention of magnesium by the immature kidney
likely contributes to the inverse relationship noted between plasma magnesium
and somatic maturity in early postnatal life (289).
Magnesium reabsorption is regulated by a number of hormones, including PTH,
calcitonin, glucagon, and AVP (290,291,292,293,294 and 295). Additionally,
dietary magnesium restriction or loading stimulates or inhibits magnesium
reabsorption, as appropriate, a response mediated by the CaSR in the cortical
TALH and distal tubule (296,297). Loop diuretics such as furosemide inhibit
magnesium absorption and increase magnesium excretion as a result of their
inhibition of sodium chloride transport and modification of the transepithelial
voltage in the TALH (298).
Glucose
Premature infants of less than 34 weeks of GA have a higher urinary glucose
concentration, higher fractional excretion of glucose, and lower maximal
reabsorption of glucose than full-term infants and older children (116). However,
the maximal reabsorption of glucose factored by GFR, the fractional reabsorption
of glucose, is similar in neonates and in adults (116,299,300). These results
provide additional evidence for preservation of glomerulotubular balance, at least
in term infants. The lower renal threshold for glucose in newborns compared to
their adult counterparts is believed to reflect a greater degree of nephron

heterogeneity early rather than later in life (299).
The neonatal proximal tubule possesses both high- and low-affinity sodiumcoupled glucose transporters that mediate reabsorption of filtered glucose;
interestingly, only the low-affinity high-capacity system is present in adults
(301,302 and 303). It is not clear when during maturation the high-affinity system
disappears, but its presence early in life may enable the anatomically immature
kidney to efficiently reabsorb sugar from the glomerular filtrate.
Organic Acids
Organic acids, including PAH (see RBF) and endogenously produced uric acid, are
eliminated by filtration and proximal tubular secretion. Organic acids are
transported from the peritubular circulation across the basolateral surface of the
proximal tubule to the tubular fluid. The renal clearance of organic acids is low in
the neonate, even when corrected for body size, and increases gradually with age
(17,304,305). As discussed previously, the limitation in tubular excretion of weak
acids may be due in part to the preponderance of blood flow to the juxtamedullary
region, bypassing tubular secretory sites. Additional variables that may account for
the limited clearance of organic acids include the low GFR, limited energy for
transport, and restricted expression of organic anion transporter proteins (306).
P.991
Amino Acids
The renal reabsorption of many amino acids, including threonine, serine, proline,
glycine, and alanine, is lower in newborn animals and humans than in adults, often
resulting in aminoaciduria (307,308). This does not appear to be a generalized
defect in amino acid reabsorption, because other filtered amino acids (e.g.,
methionine, isoleucine, leucine, tyrosine) are reabsorbed more completely. Specific
transporters for acidic, basic, and neutral amino acids have been identified in the
urinary membrane of proximal tubules in newborn kidneys (309,310,311 and
312). The transient limitation in net transtubular reabsorption of amino acids
characteristic of the neonate may arise from intrinsic differences in activity and
transport capacity of these discrete transport systems, and/or a lower rate of
amino acid efflux out of the cell into the peritubular circulation in the neonate
compared to the adult, a mechanism that also would account for the high
intracellular concentrations of amino acids observed early in life (309).
Urinary Concentration and Dilution

The fetal metanephric kidney of a variety of species, including man, produces large
volumes of hypotonic urine that contribute significantly to the volume and
composition of amniotic fluid (144,158,313,314 and 315). Yet, the fetal nephron is
able to concentrate urine under conditions of stress, such as that induced by
maternal water deprivation (316), hemorrhage (61), or infusion of AVP (317,318).
However, the maximum urine osmolality that can be achieved is only about 20%
of that in the adult (129,319).
Urinary Concentration
Urine voided at or shortly after birth generally is hypotonic with respect to plasma
(313,320). The maximal urinary concentrating ability (~1,000-1,200 mOsm/kg) of
children and adults is generally not attained by the neonate before 6 to 12 months
of age (321,322). After fluid deprivation for 12 to 24 hours, the maximal urine
osmolality achieved in premature and full-term newborns is 600 to 800 mOsm/kg,
respectively (321,322). A few 1- to 2-month-old infants may be able to generate a
urine osmolality as high as 1,000 mOsm/kg (323,324).
Urinary concentration requires a corticomedullary osmotic gradient, the pituitary
release of AVP and the ability of the collecting duct to increase its water
permeability in response to AVP. The limited urinary concentrating ability of the
infant appears to be due primarily to an inability to generate a corticomedullary
osmotic gradient and diminished responsiveness of the distal nephron to AVP
(323,324).
Corticomedullary Gradient
The capacity to concentrate the urine has been directly related to elongation of the
loops of Henle and their penetration into the medulla (325). The inner medulla and
renal papillae are poorly developed in the immature kidney. In the rat, the 1.6-fold
increase in length of the renal medulla correlates well with the 1.5-fold increase in
urine osmolality observed between 10 and 20 days of age (325). In addition to
anatomic maturation of the loops of Henle, urinary concentration requires the
generation of a high interstitial solute concentration gradient in the medulla, which
is underdeveloped early in life (321,326,327). Generation of the corticomedullary
osmotic gradient necessitates the postnatal maturation of several processes
involved in urinary concentration: sodium reabsorption and urea sequestration by
the TALH and functional activation of aldose reductase, an enzyme necessary for
generation of intracellular osmolytes, important for maintenance of cell function in
the concentrated milieu (326,328,329). These structural and functional limitations
of the countercurrent multiplication and exchange systems prevent build-up and
maintenance of a medullary gradient in the immature kidney.
Antidiuretic Hormone
The limited ability of the immature kidney to concentrate urine is not as a result of
an inability to synthesize and secrete AVP. Circulating levels of antidiuretic
hormone (ADH) are elevated in preterm and term infants and decrease rapidly in
term infants within 24 hours of birth (98,330,331). Studies in fetal and newborn
animals (59,332,333), and in human infants (331,334), indicate a qualitatively
appropriate response to osmolar or volume stimuli known to affect AVP release.
Furthermore, exogenous administration of AVP or 1-des-amino-8-d-AVP (DDAVP)
to healthy 1- to 3-week-old newborns leads to a response, albeit of shorter
duration and reduced magnitude than that observed at 4 to 6 weeks (335).
Cumulative evidence suggests that the blunted sensitivity of the fetal and neonatal
kidney to AVP and limited concentrating ability of the neonatal animal is not as a
result of a paucity of V2 receptors (receptor to which ADH binds in the collecting
duct) (336,337), aquaporin channels involved in water transport across renal
tubule epithelia (338), or efficiency of coupling to second messengers (adenylate
cyclase and protein kinase A activity) (339,340 and 341) after the first week of
postnatal life, but is limited primarily by the poorl

Avery's Neonatology PDF

Uploaded by

Copyright:

Available Formats

Avery's Neonatology PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Avery's Neonatology PDF

Uploaded by

Copyright:

Available Formats

Avery's Neonatology, 6th Edition

MacDonald, Mhairi G.;

Copyright 2005 Lippincott Williams & Wilkins

Editors: MacDonald, Mhairi G.; Seshia, Mary M. K.; Mullett, Martha

Includes bibliographical references and index.

II. MacDonald, Mhairi G. III. Seshia, Mary M. K. IV. Mullett, Martha D. V.

Mary M. K. Seshia M.B.Ch.B., F.R.C.P.(E.), F.R.C.P.C.H., D.C.H.

Martha D. Mullett M.D., M.P.H.

Thomas E. Adrian Ph.D., F.R.C.Path.

Ruben E. Alvaro M.D., F.A.A.P.

Robert J. Arceci M.D., Ph.D.

John H. Arnold M.D.

Gordon B. Avery M.D., Ph.D.

Emeritus Professor of Pediatrics; Retired Chair

Stephen Baumgart M.D.

David A. Beckman Ph.D.

Edward F. Bell M.D.

Joseph A. Bellanti M.D.

Corinne Benchimol D.O.

Department of Pediatrics, Mount Sinai School of Medicine; Clinical

Forrest C. Bennett M.D.

Judy Bernbaum M.D.

Victor Blanchette M.D., B.Ch., F.R.C.P.

Carl L. Bose M.D.

Division: plastic and reconstructive surgery, Children's National Medical

Robert J. Boyle M.D.

Robert L. Brent M.D., Ph.D., D.Sc.

Luc P. Brion M.D.

Richard L. Bucciarelli M.D.

Dorothy I. Bulas M.D.

University Medical Center; and Director, Program in Diagnostic Imaging,

Barbara K. Burton M.D.

Anthony J. Camerota M.D.

Joyce Campbell R.N., M.S.

Anthony Chan M.B.B.S., F.R.C.P.C.

Sukgi S. Choi M.D.

Otolaryngology, Children's National Medical Center, Washington, D.C.

John M. Costello M.D.

Gary L. Darmstadt M.D.

Jonathan M. Davis M.D.

James G. H. Dinulos M.D.

Yigal Dror M.D.

Director, Marrow Failure and Myelodysplasia Program, Scientist Track,

Arie Drugan M.D.

Chester M. Edelmann Jr. M.D.

Martin R. Eichelberger M.D.

Mark I. Evans M.D.

Lynda B. Fawcett Ph.D.

Assistant Professor of Pediatrics

Michael F. Flanagan M.D.

Alan W. Flake M.D.

John C. Fletcher Ph.D.

Mary Ann Fletcher M.D.

Joseph T. Flynn M.D., M.Sc.

Chapter 42: Renal Disease

Wayne H. Franklin M.D., M.P.H.

Bishara J. Freij M.D.

Michael Friedman M.D.

David M. Frim M.D., Ph.D.