Get ebook downloads in full at ebookname.

com

Forfar and Arneil s Textbook of Pediatrics 7th

Edition Neil Mcintosh

https://ebookname.com/product/forfar-and-arneil-s-textbook-
of-pediatrics-7th-edition-neil-mcintosh/

OR CLICK BUTTON

DOWNLOAD EBOOK

Explore and download more ebook at https://ebookname.com

Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

IAP Textbook of Pediatrics 5th Edition A. Parthasarathy

https://ebookname.com/product/iap-textbook-of-pediatrics-5th-
edition-a-parthasarathy/

Clinical Ethics in Pediatrics A Case Based Textbook 1st

Edition Douglas S. Diekema

https://ebookname.com/product/clinical-ethics-in-pediatrics-a-
case-based-textbook-1st-edition-douglas-s-diekema/

Shaw s Textbook of Operative Gynaecology 7e 7th Edition

Marcus E. Setchell

https://ebookname.com/product/shaw-s-textbook-of-operative-
gynaecology-7e-7th-edition-marcus-e-setchell/

Historical Linguistics 2005 Selected Papers from the

17th International Conference on Historical Linguistics
Madison Wisconsin 31 July 5 August 2005 1st Edition
Joseph C. Salmons
https://ebookname.com/product/historical-
linguistics-2005-selected-papers-from-the-17th-international-
conference-on-historical-linguistics-madison-
Rethinking College Student Retention 1st Edition John M
Braxton

https://ebookname.com/product/rethinking-college-student-
retention-1st-edition-john-m-braxton/

Numeracy and Clinical Calculations for Nurses 1st

Edition Neil Davison

https://ebookname.com/product/numeracy-and-clinical-calculations-
for-nurses-1st-edition-neil-davison/

Systems Engineering A 21st Century Systems Methodology

Derek K. Hitchins

https://ebookname.com/product/systems-engineering-a-21st-century-
systems-methodology-derek-k-hitchins/

Enhancing the Performance of Ad Hoc Wireless Networks

with Smart Antennas 1st Edition Somprakash
Bandyopadhyay

https://ebookname.com/product/enhancing-the-performance-of-ad-
hoc-wireless-networks-with-smart-antennas-1st-edition-somprakash-
bandyopadhyay/

Engineering Global E Commerce Sites The Morgan Kaufmann

Series in Data Management Systems 1st Edition James
Bean

https://ebookname.com/product/engineering-global-e-commerce-
sites-the-morgan-kaufmann-series-in-data-management-systems-1st-
edition-james-bean/
Measuring Society 1st Edition Chaitra H. Nagaraja

https://ebookname.com/product/measuring-society-1st-edition-
chaitra-h-nagaraja/
 Seventh Edition
Forfar & Arneil’s
TEXTBOOK
of PEDIATRICS
Book Donated by,
http:\\vip.persianss.ir

Released by:
tahir99 - UnitedVRG

Donate to keep us alive.....

0000732993.INDD i 3/22/2008 5:23:27 AM

 Seventh Edition

Forfar & Arneil’s

TEXTBOOK
of PEDIATRICS
Edited by
Neil McIntosh DSc(Med) FRCP(Edin) FRCP(Lond) FRCPCH
Professor of Child Life and Health, University of Edinburgh;
Honorary Consultant Paediatrician, Lothian University Hospitals NHS Trust, Edinburgh, UK

Peter J. Helms MB BS PhD FRCP(Edin) FRCPCH

Professor of Child Health, University of Aberdeen, Aberdeen, UK

Rosalind L. Smyth MA MB BS MD FRCPCH F Med Sci

Brough Professor of Paediatric Medicine, University of Liverpool, Liverpool, UK

Stuart Logan MB ChB MSc(Epidemiology) MSc(politics) MRCP FRCPCH

Professor of Paediatric Epidemiology; Director, Institute of Health Research, Peninsula Medical Service School,
St Luke’s Campus, Exeter, UK

Edinburgh London New York Oxford Philadelphia St Louis Sydney Toronto 2008

0000732996.INDD iii 3/22/2008 5:37:12 AM

 An imprint of Elsevier Limited

© 2008, Elsevier Limited. All rights reserved.

© 2003, Elsevier Limited.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical,
including photocopying, recording, or any information storage and retrieval system, without permission in writing
from the publishers. Permissions may be sought directly from Elsevier’s Rights Department: Phone: (+1) 215 239
3804 (US) or (+44) 1865 843830 (UK); fax: (+44) 1865 853333; e-mail: [email protected].
You may also complete your request on-line via the Elsevier website at http://www.elsevier.com/permissions.

First edition 1973

Second edition 1978
Third edition 1984
Fourth edition 1992
Fifth edition 1998
Sixth edition 2003
Seventh edition 2008

ISBN 978–0–443–10396–4
IE ISBN 978–0–443–10397–1

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloging in Publication Data

A catalog record for this book is available from the Library of Congress

Note
Knowledge and best practice in this field are constantly changing. As new research and experience broaden our
knowledge, changes in practice, treatment and drug therapy may become necessary or appropriate. Readers are
advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each
product to be administered, to verify the recommended dose or formula, the method and duration of administration,
and contraindications. It is the responsibility of the practitioner, relying on their own experience and knowledge of
the patient, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take
all appropriate safety precautions. To the fullest extent of the law, neither the Publisher nor the Authors assumes
any liability for any injury and/or damage to persons or property arising out or related to any use of the material
contained in this book.
The Publisher

The
publisher's
policy is to use
paper manufactured
from sustainable forests
Printed in China

0000732992.INDD iv 3/22/2008 5:21:40 AM

 10001

Preface

One of the most important developments in the 21st century publishing and consequently another important new feature for
has been the acknowledgment that clinical practice should be this edition is a website of the text in which references listed
embedded where possible in a sound evidence base. The editors in PUBMED can be accessed via the internet. We would like
of this new edition of Forfar & Arneil’s Textbook of Pediatrics to acknowledge the outstanding contributions of the many
continue to set contributors the challenging task of ensuring individual authors and our chapter editors, without whom this
that the statements and recommendations made are based on ambitious project would not have been possible. It has been a
the most robust clinical research available. The objective assess- great privilege to work with them. We would also like to thank
ment of evidence and the formulation of secure recommenda- our own secretaries for their unstinting help (Elaine Forbes,
tions from the evidence are gathering pace. In Child Health the Flora Buthlay, Collette Lorne and Stella Taylor) and of course
evidence base, with the possible exception of disciplines such as our respective families for sharing the burden with us.
oncology and neonatology, is not as comprehensive or as well
defined as we would wish. This edition of Forfar & Arneil con- As with previous editions the book is produced by the Elsevier
tinues to flag up the most important evidence sources for the Organization. There have been many highly professional people
content of each chapter. The whole book has been revised and involved. We hope that this new 7th edition with its expanded
updated and the chapter authors and editors again incorporate formal evidence base will be well received by you, the reader,
where possible secure evidence of therapy, diagnosis, etiology and become as trusted a friend and guide as have previous
and prognosis. This evidence is incorporated into the individ- editions.
ual chapters and in the chapter reference list is starred for easy
reference. We believe that the result is the most comprehensive Neil McIntosh
evidence-based general textbook of pediatrics available. Our Peter J. Helms
ability to access best evidence has been greatly enhanced by Rosalind L. Smyth
the rapid advances of information technology and electronic Stuart Logan


 10001

Contributors

Ishaq Abu-Arafeh, MBBS, MD, MRCP, FRCPCH Paul L.P. Brand, MD Andrew Gavin Cleary, MB ChB BSc MSc MRCPCH
Consultant in Paediatrics and Paediatric Consultant Paediatrician and Director of Consultant Paediatric Rheumatologist, Royal
Neurology, Stirling Royal Infirmary, Stirling; Postgraduate Education, Princess Amalia Liverpool Children’s Hospital, Liverpool, UK
Honorary Consultant Paediatrician, Fraser of Children’s Clinic, Isala klinieken, Zwolle,
J. Brian S. Coulter, MD FRCP(I) FRCPCH
Allander Neurosciences Unit, Royal Hospital The Netherlands
Honorary Clinical Lecturer in Tropical
for Sick Children, Glasgow, UK
Colin E. Bruce, MB ChB FRCS FRCS(ORTH) Child Health, Liverpool School of Tropical
N. Archer, MA FRCP FRCPCH DCH Consultant Paediatric Orthopaedic Medicine, Liverpool, UK
Consultant Paediatric Cardiologist, Oxford Surgeon, Alder Hey Children’s Hospital,
Jonathan Coutts, MB ChB FRCPCH FRCP(Glas)
Children’s Hospital, Oxford; Honorary Liverpool, UK
Consultant Neonatal and Respiratory
Clinical Senior Lecturer, University of
Helen Budge, BA BM BCh PhD MRCP FRPCH Paediatrician, The Queen Mother’s Hospital
Oxford, Oxford, UK
Associate Professor of Neonatology, and Royal Hospital for Sick Children,
Peter D. Arkwright, FRCPCH DPhil University of Nottingham, Nottingham, Glasgow, UK
Lecturer in Child Health, Department of UK
Richard Coward, MB ChB MRCPCH PhD
Child Health, University of Manchester,
Neil R.M. Buist, MB ChB FRCPE DCH MRC Clinician Scientist and Consultant
Manchester, UK
Professor Emeritus, Departments of Paediatric Nephrologist, Bristol Royal
Ian A. Auchterlonie, MB ChB FRCPCH FRCP Paediatrics and Medical Genetics, Oregon Hospital for Children, Bristol, UK
Honorary Consultant Paediatrician (Retired), Health and Science University, Portland,
Finella Craig, BSc MBBS MRCP
Royal Aberdeen Children’s Hospital, Oregon, USA
Consultant in Paediatric Palliative Care,
Aberdeen, UK
Nigel Peter Burrows, MD FRCP Great Ormond Street Hospital for Children,
Alastair J. Baker, MB ChB FRCP FRCPCH Consultant Dermatologist and Associate London, UK
Consultant Paediatric Hepatologist, King’s Lecturer, Department of Dermatology,
Timothy J. David, MB ChB MD PhD FRCP FRCPCH
College Hospital, London, UK Addenbrooke’s Hospital, Cambridge, UK
DCH
Susan V. Beath, BSc MB BS MRCP DTM&H FRCPCH Andrew Bush, MD FRCP FRCPCH Professor of Child Health and Paediatrics,
Consultant Paediatric Hepatologist, The Professor of Paediatric Respirology, Imperial Honorary Consultant Paediatrician, The
Liver Unit, Birmingham Childrens Hospital, College, London; Honorary Consultant University of Manchester, Manchester, UK
Birmingham, UK Paediatric Chest Physician, Royal Brompton
Joyce E. Davidson, BSc MBChB MRCP (UK) FRCPCH
Hospital, London, UK
Thomas F. Beattie, MB MSc FRCSEd(A&E), FCEM, Consultant Paediatric Rheumatologist, Royal
FRCPE, FFSEM, DCH Gary Butler, MB BS MD FRCPCH Hospital for Sick Children, Glasgow, and
Consultant in Emergency Medicine, Royal Professor of Paediatrics, Institute of Health Royal Hospital for Sick Children,
Hospital for Sick Children, Edinburgh; Sciences, University of Reading and Edinburgh, UK
Honorary Senior Lecturer in Paediatrics, Department of Paediatrics, Royal Berkshire
E. Graham Davies, MB BChir MA FRCPCH
University of Edinburgh, Edinburgh, UK Hospital, Reading, UK
Consultant Paediatric Immunologist, Great
Julie-Clare Becher, MB ChB MRCPCH MD Catherine M. Cale, MB ChB PhD MRCP MRCPCH Ormond Street Hospital, London, UK
Consultant Neonatologist, Simpson Centre MRCPath
Jan Dudley, BM MRCP(UK) FRCPCH PhD
for Reproductive Health, Royal Infirmary of Consultant in Paediatric Immunology
Consultant Paediatric Nephrologist, Bristol
Edinburgh, Edinburgh, UK and Immunopathology, Department of
Royal Hospital for Children, Bristol, UK
Immunology, Great Ormond Street Hospital,
W. Michael Bisset, BSc MB ChB DCH MSc MD
London, UK P. Barton Duell, MD
FRCPE FRCPCH
Associate Professor of Medicine, Division
Consultant Paediatric Gastroenterologist, Harry Campbell, MD FRCPE FFPH FRSE
of Endocrinology, Diabetes, and Clinical
Department of Medical Paediatrics, Royal Professor of Genetic Epidemiology and
Nutrition; director, Lipid-Atherosclerosis
Aberdeen Children’s Hospital, Aberdeen, UK Public Health Sciences, College of Medicine
Laboratory and Metabolic Disorders Clinic,
and Vet Medicine, University of Edinburgh,
Peter S. Blair, PhD Oregon Health and Science University,
Edinburgh, UK
Medical Statistician, Senior Research Fellow, Portland, Oregon
University of Bristol, Division of Child Health, Andrew J. Cant, BSc MD FRCP FRCPH
Heather Elphick, MB ChB MRCP MRCPCH MD
Royal Hospital for Children, Bristol, UK Consultant in Paediatric Immunology and
Specialist Registrar in Respiratory
Infectious Diseases, Newcastle General
Paula H.B. Bolton-Maggs, FRCPCH FRCP FRCPath Paediatrics, Sheffield Children’s Hospital,
Hospital, Newcastle upon Tyne, UK
Consultant Haematologist, Manchester Sheffield, UK
Comprehensive Care Haemophilia Centre, Michael Clarke, BSc MRCP DCH FRCPCH
Nicholas D. Embleton, BSc MD FRCPCH
Department of Clinical Haematology, Consultant Paediatric Neurologist, Leeds
Consultant in Neonatal Medicine and
Manchester Royal Infirmary, Manchester, UK Teaching Hospitals NHS Trust, Leeds, UK xv
xvi Contributors

Honorary Lecturer in Child Health, Newcastle Cellular Medicine, University of Newcastle and Science University, Portland,
Neonatal Service, Royal Victoria Infirmary, upon Tyne, and Children’s Bone Marrow Oregon, USA
Newcastle upon Tyne, UK Transplant Unit, Newcastle General Hospital,
Newcastle upon Tyne, UK C. Anthony Hart, BScMB BS PhD FRCPCH FRCPath
Gregory M. Enns, MB ChB Professor, Division of Medical Microbiology,
Associate Professor of Pediatrics, Director, Diana Gibb, MD FRCPCH MSc School of Infection and Host Defence,
Biochemical Genetics Program, Division Professor of Paediatric Epidemiology, University of Liverpool, Liverpool , UK
of Medical Genetics, Stanford University, MRC Clinical Trials Unit, University College
Stanford, California, USA London, London, UK Paul T. Heath, MB BS FRACP FRCPCH
Senior Lecturer and Honorary Consultant,
Jeremy Farrar, FRCP DPhil OBE K. Michael Gibson, PhD, FACMG Paediatric Infectious Diseases, Division
Professor of Tropical Medicine Oxford Professor, Pediatrics, Pathology and Human of Child Health, St George’s University of
University; Honorary Professor of Genetics, Division of Medical Genetics, London, London, UK
International Health, London School of Children’s Hospital of Pittsburgh, University
Hygiene and Tropical Medicine, London; of Pittsburgh School of Medicine, Pittsburgh, Peter J. Helms, MB BS PhD FRCP(Edin) FRCPCH
Director, Oxford University Clinical Research PA, USA Professor of Child Health, University of
Unit, The Hospital for Tropical Diseases, Aberdeen, Aberdeen, UK
Barbara Golden, BSc MB BCh BAO MD FRCPI
Ho Chi Minh City, Viet Nam John Henderson, MD FRCP FRCPCH FRCPEd
FRCPCH DCH RNutr
Colin D. Ferrie, BSc MB ChB MD MRCP FRCPCH Associate Dean (Medicine), Clinical Senior Reader in Paediatric Respiratory Medicine,
Consultant Paediatric Neurologist, Lecturer, Honorary Consultant International University of Bristol and Bristol Royal
Department of Paediatric Neurology, Leeds Child Health, Child Health Institute, Hospital for Children, Bristol, UK
General Infirmary, Leeds, UK School of Medicine, University of Aberdeen, Tom Hilliard, MD FRCPCH
Royal Aberdeen Children’s Hospital, Consultant Respiratory Paediatrician, Bristol
Alistair R. Fielder, FRCP FRCS FRCOphth
Aberdeen, UK Children’s Hospital, Bristol, UK
Professor of Ophthalmology, Department
of Optometry and Visual Science, City John M. Goldsmid, BSc MSc PhD FRCPath FACTM Peter Hoare, DM FRCPsych
University, London, UK FASM CBiol FIBiol FAIBiol HonFRCPA HonFACTM Senior Lecturer, University of Edinburgh;
Emeritus Professor, Discipline of Pathology, Honorary Consultant Psychiatrist, Royal
Brian W. Fleck, MD FRCOph
University of Tasmania, Hobart, Tasmania, Hospital for Sick Children, Edinburgh, UK
Consultant Ophthalmologist, Edinburgh
Australia
Royal Infirmary and Edinburgh Royal Richard F. Howard, BSc MBChB FRCA
Hospital for Sick Children, Edinburgh, UK John R.W. Govan, BSc PhD DSc Consultant in Anaesthesia and Pain
Professor of Microbial Pathogenicity, Management, Great Ormond Street Hospital
Peter J. Fleming, PhD MB ChB FRCPCH FRCP
Centre for Infectious Diseases, University of for Children, London, UK
FRCP(C)
Edinburgh Medical School, Edinburgh, UK
Professor of Infant Health and Imelda Hughes, FRCPCH
Developmental Physiology, University of Stephen M. Graham, FRACP Consultant Paediatric Neurologist,
Bristol; Consultant Paediatrician, Royal Paediatric Research Fellow and Deputy Department of Paediatric Neurology,
Hospital for Children, Bristol, UK Director Malawi-Liverpool-Wellcome Trust Royal Manchester Children’s Hospital,
Clinical Research Programme, University Manchester, UK
Caroline Foster, MB BS MRCP
of Malawi College of Medicine, Blantyre,
Honorary Clinical Research Fellow, Imperial Robert Hume, BSc MB ChB PhD FRCPEdin FRCPCH
Malawi; Senior Clinical Lecturer, Liverpool
College London; SPR Paediatric HIV, Professor of Developmental Medicine,
School of Tropical Medicine, Liverpool, UK
St Mary’s Hospital, London, UK Maternal and Child Health Sciences,
Anne Green, BSc MSc PhD FRCPath FRCS FIBiol
University of Dundee, Dundee, UK
Yvonne Freer, RGN RM RSCN BSc PhD
FRCPCH.
Research and Practice Development Carol Inward, MBBCh MD FRCPCH
Consultant Clinical Scientist, Birmingham
Facilitator, Neonatal Intensive Care, Simpson Consultant Paediatric Nephrologist, Bristol
Children’s Hospital NHS Trust,
Centre for Reproductive Health, Royal Royal Hospital for Children, Bristol, UK
Birmingham, UK
Infirmary of Edinburgh, Edinburgh, UK
Marcus Grompe, MD David Isaacs, MB BChir MD FRACP FRCPCH
Vijeya Ganesan, MRCPCH, MD Senior Staff Specialist at Children’s Hospital
Director, Oregon Stem Cell Center; Professor,
Senior Lecturer in Paediatric Neurology, at Westmead; Clinical Professor in Paediatric
Department of Molecular and Medical
Institute of Child Health, University College Infectious Diseases, University of Sydney,
Genetics, Oregon and Health Science University,
London, London, UK Sydney, Australia
Portland, Oregon, USA
Anne S. Garden, FRCOG Omar Ismayl, MD MSc MRCPCH
Henry L. Halliday, FRCPE FRCP FRCPCH
Head of Department of Medicine and Consultant Neurologist, Damascus University
Chair of Child Health and Welfare Research
Director of the Centre for Medical Education, Children’s Hospital, Mezza, Damascus, Syria
Group, Department of Child Health, Queen’s
Lancaster University, Lancaster; Honorary
University Belfast, Northern Ireland, UK
Consultant in Paediatric and Adolescent Huw R. Jenkins, MA MD FRCP FRCPCH
Gynaecology, Royal Liverpool Children’s Christina Halsey, BM BCh BA MRCP MRCPath Consultant Paediatric Gastroenterologist,
Hospital, Liverpool; Honorary Consultant Leukaemia Research Fund Clinical Research Department of Child Health, University
Gynaecologist, University Hospitals Fellow, Faculty of Medicine, Glasgow Hospital of Wales, Cardiff, Wales, UK
Morecambe Bay Trust, Morecambe, UK University, Glasgow, UK
Cheryl A. Jones, MB BS PhD FRACP
Andrew R. Gennery, MD MRCP MRCPCH Cary O. Harding, MD Associate Professor, and Sub Dean
DCH DipMedSci Assistant Professor, Molecular and Medical Postgraduate Studies, Discipline of
Senior Lecturer and Honorary Consultant Genetics; Director of the Biochemical Paediatrics and Child Health, University of
in Paediatric Immunology, BMT Institute of Genetics Clinical Laboratory, Oregon Health Sydney; Sydney, New South Wales, Australia
 Contributors xvii

Christopher J.H. Kelnar, MA MD FRCP FRCPCH Alison Leaf, BSc MB ChB MD MRCP FRCPCH Ian McKinley, OBE BSc MB ChB DCH FRCP FRCPCH
Professor of Paediatric Endocrinology, Consultant Neonatologist, Southmead Senior Lecturer in Child Health, McKay-
Section of Child Life and Health, Division of Hospital, Bristol, UK Gordon Centre, Royal Manchester Children’s
Reproductive and Developmental Sciences, Hospital, Manchester, UK
Malcolm I. Levene, MD FRCPCH FMedSc
University of Edinburgh, Edinburgh, UK
Professor of Paediatrics and Child Health, Sheila A.M. McLean, LLB MLitt PhD LLD FRSE
Alison M. Kemp, MB BCH DCH MRCP University of Leeds, Leeds General Infirmary, FRCGP FMedSci FRCP(Edin) FRSA
FRCPCH FRCP(Edin) Leeds, UK International Bar Association Professor of
Reader in Child Heath, Department of Child Law and Ethics in Medicine, University of
Leesa M. Linck, MD MPH
Health, Cardiff University, Cardiff, Wales Glasgow, UK
Consultant Paediatrician,
Alastair I.G. Kerr, MBChB FRCS(Edin & Glas) The Connty Clinic, Winthrop, Sabine Maguire, MB BCh BAO MRCPI FRCPI FRCPCH
Consultant Otolaryngologist, Ear, Nose and Washington State, USA. Senior Lecturer in Child Health, Cardiff
Throat Department, Royal Infirmary for Sick University, Department of Child Health,
John H. Livingston, MB ChB FRCP FRCPCH
Children, Edinburgh, UK Cardiff, UK
Consultant Paediatric Neurologist,
Alison M. Kesson, MB BS PhD FRACP FRCPA Department of Paediatric Neurology, Leeds Janice Main, FRCP(Edin) FRCP(Lond)
Virologist, Microbiologist and Paediatric General Infirmary, Leeds, UK Reader and Consultant Physician in
Infectious Disease Physician, The Children’s Stuart Logan, MB ChB MSc(Epidemiology)
Infectious diseases and General Medicine,
Hospital, Westmead, Sydney New South MSc(Econ) MRCP FRCPCH
Department of Medicine, Imperial College,
Wales; Conjoint Associate Professor, Professor of Paediatric Epidemiology; St Mary’s Hospital, London, UK
Paediatrics and Child Health, University of Director, Institute of Health and Social Guy Makin, BA BM BCh PhD MRCP FRCPCH
Sydney, Sydney New South Wales, Australia Care Research, Peninsula Medical School, Senior Lecturer in Paediatric Oncology,
Maurice A. Kibel, FRCP(Edin) DCH(Lond). St Luke’s Campus, Exeter, UK University of Manchester; Honorary
Emeritus Professor of Child Health, School Diana N.J. Lockwood, BSc MD FRCP Consultant Paediatric Oncologist,
of Child and Adolescent Health,University of Consultant Physician and Leprologist, Royal Manchester Children’s Hospital,
Cape Town, South Africa Hospital for Tropical Diseases, London, UK Manchester, UK
Denise Kitchiner, MD FRCP FRCPCH Andrew J. Lyon, MA MB FRCP FRCPCH Timothy Martland, MB ChB MRCP MRCPCH
Consultant Paediatric Cardiologist, Consultant Neonatologist, Simpson Centre Consultant Paediatric Neurologist,
Royal Liverpool Children’s NHS Trust, for Reproductive Health, Royal Infirmary of Department of Paediatric Neurology,
Liverpool, UK Edinburgh, Edinburgh, UK Royal Manchester Children’s Hospital,
Manchester, UK
Nigel J. Klein, BSc MB BS MRCP PhD FRCPCH Ronan Lyons, MD FFPH FFPHMI MPH DCH
Professor of Infection and Immunity, Professor of Public Health, Centre for Health Amha Mekasha, MD MSc
University College London; Head, Infectious Information, Research and Evaluation, Associate Professor of Paediatrics and Child
Diseases and Microbiology Unit, Institute of School of Medicine, Swansea University, Health, Department of Paediatrics and
Child Health; Head, Department of Infection, Swansea, UK Child Health, Addis Ababa University,
University College London; Consultant in Adis Ababa, Ethiopia
Infectious Diseases and Microbiology, Anita Macdonald, PhD BSc SRD
Consultant Dietician in Inherited Metabolic Craig Mellis, MB BS MPH MD FRACP
Great Ormond Street Hospital NHS Trust,
London, UK Disorders, Birmingham Children’s Hospital, Head of School, Associate Dean, Central
Birmingham, UK, Head of Dietetics, Clinic School, Faculty of Medicine, The
David M. Koeller, MD Birmingham Children’s Hospital NHS Trust, University of Sydney, Sydney, Australia
Associate Professor, Pediatrics and Molecular Birmingham, UK
Stefan Meyer, MD PhD MRCPCH
and Medical Genetics; Chief, Division
Mary McGraw, FRCP FRCPCH CRUK Clinician Scientist and Consultant
of Metabolism, Department of Pediatrics,
Consultant Paediatric Nephrologist, Paediatric Oncologist, Univeristy of
Doernbecher Children’s Hospital, Oregon
Bristol Royal Hospital for Children, Manchester, Royal Manchester Children’s
Health and Science University, Portland,
Bristol, UK Hospital and Young Oncology Unit, Christie
Oregon, USA
Hospital, Manchester, UK, Stefan Meyer MD
Sailesh Kotecha, MA FRCPCH PhD Keiran McHugh, FRCR FRCPI DCH PhD MRCPCH
Professor of Child Health, Department of Consultant Paediatric Radiologist, Great
Ormond Street Hospital for Children, Alan E. Mills, MA MB DCP DPath FRCPA.
Child Health, Wales College of Medicine,
Cardiff University, Cardiff, Wales, UK London, UK FFPathRCPI FACTM
Consultant Pathologist, Dorevitch Pathology,
Karen Kotloff, MD Neil McIntosh, DSc(Med) Bendigo, Victoria, Australia
Professor, Pediatrics and Medicine, Center Professor of Child Life and Health,
Department of Child Life and Health, Elizabeth Molyneux, OBE FRCP FRCPCH FEM
for Vaccine Development, University of
University of Edinburgh; Honorary Professor and Head of Department of
Maryland School of Medicine, Baltimore,
Consultant Paediatrician, Lothian University Paediatrics, College of Medicine, Blantyre,
Maryland, USA
Hospitals NHS Trust, Edinburgh, UK Malawi, C Africa.
Ian A. Laing, MA MD FRCPE FRCPCH
G.A. Mackinlay, MB BS LRCP FRCSEd Malcolm E. Molyneux, MD FRCP FMedSci
Consultant Neonatologist, Simpson Centre
FRCSEng FRCPCH Director, Malawi-Liverpool-Wellcome Trust
for Reproductive Health, Royal Infirmary of
Senior Lecturer, Department of Clinical Clinical Research Programme, College of
Edinburgh, Edinburgh, UK
Surgery, University of Edinburgh; Consultant Medicine, University of Malawi, Malawi;
David G. Lalloo, MB BS MD FRCP Paediatric Surgeon, Department of Paediatric Professor of Tropical Medicine, School of
Reader In Tropical Medicine, Liverpool Surgery, Royal Hospital for Sick Children, Tropical Medicine, University of Liverpool,
School of Tropical Medicine, Liverpool, UK Edinburgh, UK Liverpool, UK
xviii Contributors

Kim Mulholland, MD FRACP Irene A.G. Roberts, MD FRCP FRCPath FRCPCH Children’s Hospital; Deputy Director,Oregon
Professor of Child Health, London School DRCOG Clinical and Translational Research Institute,
of Hygiene and Tropical Medicine, London, Professor of Paediatric Haematology and Oregon Health and Science University,
UK, Infectious Disease Epidemiology Unit, Consultant Paediatric Haematologist, Portland, Oregon, USA
London School of Hygiene and Tropical Hammersmith Hospital and St Mary’s
Ben Stenson, MD FRCPCH FRCPE
Medicine, London, UK Hospitals, Imperial College, London, UK
Consultant Neonatologist, Simpson Centre
Alan O. Mulvihill, FRCSI FRCSEd Peter T. Rudd, MD FRCPCH for Reproductive Health, Royal Infirmary of
Consultant Ophthalmic Surgeon, Princess Consultant Paediatrician, Royal United Edinburgh, Edinburgh, UK
Alexandra Eye Pavilion and Royal Hospital Hospital, Bath, UK
Terence Stephenson, DM FRCP FRCPCH
for Sick Children, Edinburgh, UK
S.C. Robson, MD MRCOG Professor of Child Health; Dean,
Richard W. Newton, MD FRCPCH FRCP Professor of Fetal Medicine, School of Faculty of Medicine and Health
Consultant Paediatric Neurologist, Surgical and Reproductive Sciences, Newcastle Sciences, Medical School, University of
Royal Manchester Children’s Hospital, University, Newcastle upon Tyne, UK Nottingham, Queen’s Medical Centre,
Manchester, UK Nottingham, UK
Jean-Baptiste Roullet, PhD
Angus Nicoll, CBE FFPHM FRCP FRCPCH Associate Professor (Research), Division Stephen Stick, PhD MB BChir MRCP FRACP
Director, Communicable Disease Surveilaance of Metabolism, Department of Paediatrics, Associate Professor, Department of
Centre, Centre for Infection, Health Oregon Health and Science University, Respiratory Medicine, Princess Margaret
Protection Agency, London, UK Portland, Oregon, USA; Adjunct Associate Hospital for Children, University of Western
Professor of Pharmacology and Therapeutics, Australia, Australia
James Y. Paton, MD, FRCPCH FRCPSG
University of Calgary, Calgary, Alberta, Stephen Sturgiss, MD MRCOG
Reader in Paediatric Respiratory Medicine, Canada
Division of Developmental Medicine, Consultant in Obstetrics and Fetal Medicine,
University of Glasgow, Royal Hospital for George Rylance, MB FRCPCH Royal Victoria Infirmary, Newcastle upon
Sick Children, Glasgow, UK Consultant in Paediatrics, Department of Tyne, UK
Child Health, Royal Victoria Infirmary, David Sweet, MD FRCPCH
Michael A. Patton, MA MSc MBChB FRCPCH
Newcastle upon Tyne, UK Consultant Neonatologist, Royal Maternity
Professor of Medical Genetics and Consultant
Medical Geneticist, Department of Medical Moin A. Saleem, MB BS FRCP PhD Hospital, Belfast, Northern Ireland, UK
Genetics, St George’s Hospital Medical Reader and Honorary Consultant Paediatric Angela E. Thomas, MB BS PhD FRCPE FRCPath
School, London, UK Nephrologist, Bristol Children’s Hospital, FRCPCH
University of Bristol, Bristol, UK Consultant Paediatric Haematologist,
Philip L. Pearl, MD
Associate Professor of Paediatrics and Alison Salt, MB BS MSc FRACP FRCPCH Royal Hospital for Sick Children, Edinburgh,
Neurology, Department of Neurology, Consultant Developmental Paediatrician, UK
Children’s National Medical Center, Neurodisability Service, Great Ormond Street James Tooley, MB BS MRCPCH
The George Washington University School Hospital for Children NHS Trust and Institute Consultant in Neonatal Medicine, Neonatal
of Medicine, Washington DC, USA of Child Health, London, UK Intensive Care Unit, St Michael’s Hospital,
Gale A. Pearson, MB BS MRCP FRCPCH Jenefer Sargent, MA MB BCh MSc MRCPCH Bristol, UK
Consultant in Paediatric Intensive Consultant Developmental Paediatrician, E. Jane Tizard, MB BS, FRCP FRCPCH
Care, Birmingham Children’s Hospital, Neurodisability Service, Great Ormond Consultant Paediatric Nephrologist, Bristol
Birmingham, UK Hospital for Children NHS Trust, London, UK Royal Hospital for Children, Bristol, UK
Stavros Petrou, BSc MPhil PhD Ayad Shafiq, FRCOphth MRCPPaeds DCH BM BCh Russell Viner, MB BS FRCPCH FRACP FRCP PhD
MRC Senior Non-Clinical Research Fellow, Consultant Ophthalmologist, Royal Victoria Consultant in Adolescent Medicine and
National Perinatal Epidemiology Unit, Hospital, Newcastle upon Tyne, UK Endocrinology, Department of Paediatrics,
University of Oxford, Oxford, UK Middlesex Hospital, London, UK
Jonathan R. Skinner, MB ChB FRCPCH FRACP MD
Ian J. Ramage, FRCPCH Paediatric Cardiologist, Green Lane Paediatric Rachel Webster, BSc PhD DipRCPath
Consultant Paediatric Nephrologist, Royal and Cardiac Services, Starship Children’s Clinical Scientist, Department of Clinical
Hospital for Sick Children, Glasgow, UK Hospital, Auckland, New Zealand Biochemistry, Birmingham Children’s
Hospital, Birmingham, UK
John J. Reilly, BSc PhD Alan Smyth, MA MB BS MRCP MD FRCPCH
Professor of Paediatric Energy Metabolism, Reader in Child Health, University of Philip D. Welsby, FRCP(Ed)
University Division of Developmental Nottingham; Honorary Consultant in Consultant in Infectious Diseases, Infectious
Medicine, Royal Hospital for Sick Children, Paediatric Respiratory Medicine, Disease Unit, Western General Hospital,
Glasgow, UK Nottingham University Hospitals NHS Edinburgh, UK
Trust, Nottingham, UK
J.M. Rennie, MD FRCP FRCPCH DCH Johannes Wildhaber, MD PhD
Consultant and Senior Lecturer in Neonatal Rosalind L. Smyth, MA MB BS MD FRCPCH Head, Division of Respiratory Medicine,
Medicine, Elizabeth Garrett Anderson and Brough Professor of Paediatric Medicine, University Children’s Hospital, Zürich,
Obstetric Hospital, University College London University of Liverpool, Liverpool, UK Switzerland
Hospitals, London
Robert D. Steiner, MD Anthony F. Williams, BSc DPhil MB BS
Sam Richmond, MB BS FRCP FRCPCH Professor, Pediatrics and Molecular and FRCP FRCPCH
Consultant Neonatologist, Sunderland Royal Medical Genetics; Vice Chair for Research, Reader in Child Nutrition, St George’s,
Hospital, Sunderland, UK Department of Pediatrics, Doernbecher University of London, London, UK
 Contributors xix

Bridget A. Wills, BmedSci BM BS MRCP DTM&H and Health, Department of Reproductive Christopher Wren, MB ChB FRCP FRCPCH
Clinical Senior Lecturer in Paediatrics, and Developmental Sciences, University of Consultant Paediatric Cardiologist, Freeman
University of Oxford – Wellcome Trust Edinburgh, Edinburgh, UK Hospital, Newcastle upon Tyne, UK
Clinical Research Unit, Ho Chi Minh City,
Rachael Wood, BSc MB ChB DCH MFH MFPH
Vietnam
Clinical Academic Fellow, Department
David C. Wilson, MD FRCP FRCPCH of Public Health Sciences and Centre for
Senior Lecturer in Paediatric International Public Health Policy University
Gastroenterology and Nutrition, Child Life of Edinburgh, Edinburgh, UK
 10001

1
Evidence-based child health
Rosalind L Smyth
Introduction 3 How textbooks ensure that they are evidence based 6
The practice of evidence-based medicine 4 Clinical evidence within the 7th edition of Forfar & Arneil 7
Systematic reviews 5 Summary 9
Evidence-based child health 6

Introduction justify suggesting to the parents that the child should undergo a small
intestinal biopsy but also too high to allow you to dismiss the possibil-
Evidence-based pediatrics and child health has been defined as ‘the inte- ity. You wish to know whether a test which measures antiendomysial
gration of clinical information obtained from a patient, with the best antibodies in the serum will, if positive, mean that the diagnosis is suffi-
evidence available from clinical research and experience and the appli- ciently likely to make the biopsy worth doing or, if negative, decrease the
cation of this knowledge to the prevention, diagnosis or management probability to a level where you can reasonably discount the diagnosis.
of disease in that child’.1 This definition has been adapted from an ear- A publication from 13 European Centres has evaluated both serum
lier one by Sackett and colleagues2 who have consistently argued that antitissue transglutaminase IgA (IgA-TTG) antigliadin antibodies and
there is an art to medicine, as well as objective scientific knowledge, serum IgA antiendomysial antibodies (IgA-EMA) and compared them
and that both are essential to the clinical encounter. Within the classi- with findings of subtotal or severe partial villous atrophy with crypt
cal clinical method, as taught in medical schools and beyond, the diag- hyperplasia on small intestinal biopsy (the reference standard).4 The
nostic approach starts with a history, leads on to a clinical examination blood samples were taken no more than 60 days before and not after the
and utilizes, if required, special investigations. In the process of history small intestinal biopsy, and were independently reviewed by two inves-
taking, the clinician integrates the case specific features of the patient’s tigators, who had no knowledge of the diagnosis. So it appeared that
own story with their accumulated case expertise. For example, the dif- the reference standard was applied independently of the diagnostic tests
ferential diagnosis and mode of history taking in two children present- being evaluated and the reference standard was applied objectively to all
ing with cough will be very different, if in one, it is elicited that she has patients. It is not clear from the paper whether the diagnostic tests were
experienced fever and night sweats and is recently returned from India, evaluated in an appropriate spectrum of patients or whether all patients
and in the other, recurrent wheeze and hay fever are accompanying with a possible diagnosis of celiac disease underwent small intestinal
symptoms. In taking a history, one constantly updates the differential biopsy, regardless of the results of the blood tests. However, the sensi-
diagnoses and redirects one’s questioning as each new item of informa- tivities of IgA-TTG and IgA-EMA in the detection of celiac disease were
tion is elicited. Indeed, it is this ability to take a history in a directive and 94% and 89%, with specificities of 99% and 98% respectively.
efficient manner that distinguishes the experienced clinician from the The likelihood ratio is the ratio of the probability of the test result in
new clinical student. patients with the disease to the probability of the test result in patients
Sackett and colleagues have long argued that the clinical examina- without the disease. It enables an updated ‘post-test’ probability to be
tion should be studied rigorously to establish which features are predic- calculated from the pretest probability, using the results of appropriate
tive of specific diagnoses. For example, they have reviewed studies that studies. The likelihood ratio associated with a positive test result can be
evaluated the accuracy of signs recommended for use in the diagnosis of calculated from the sensitivity and specificity as follows:
chronic obstructive pulmonary disease.3 They found that for all the signs
LR+ = sensitivity/(1 − specificity)
reviewed, the sensitivity, specificity and likelihood ratios varied consider-
ably between studies, and were far short of what would be required for The equation which relates post-test odds to pre-test odds, using the
an ideal diagnostic test.3 likelihood ratio is:
It is only the final part of the diagnostic process, the use of special
Post-test odds = pretest odds × likelihood ratio
investigations, for which there are many studies which meet established
criteria for valid assessment of clinical evidence.2 Conventionally, diag- From the example of IgA-TTG provided above, the study we have
nostic tests are judged by their sensitivity (proportion of patients with found would suggest the following:
the target disorder who have positive tests) and specificity (proportion
LR+ = 94%/1%, or 94
of patients without the target disorder who have negative tests). More
useful than sensitivity or specificity in understanding the performance and for IgA-EMA:
of a test is the likelihood ratio which can be regarded as a summary of
LR+ = 89%/2% or 44.5
sensitivity and specificity (although it can also be used to summarize the
performance of tests which have many categories where sensitivity and If this is applied to the estimated pretest probability, provided in the
specificity are unhelpful). For example, suppose you are assessing a child example above, of 10%, or odds of 1:9, a positive test result for IgA-TTG
with failure to thrive and malabsorption. You might estimate on the will convert these pretest odds to a post-test odds of 94:9, or a post-test
basis of the clinical picture that the probability of the child having celiac probability of 91%. The equivalent calculation for IgA-EMA is a post-
disease is around 10%. You might feel that this probability is too low to test odds of 44.5:9 or 83%. Most clinicians would consider that a small 
 Foundations of health and practice

intestinal biopsy should be performed in a child with these probabilities The Practice of Evidence-Based
of having celiac disease. Medicine
What should the clinician do, though, if the test is negative? The like-
lihood ratio of a negative test result can be calculated as follows: Traditionally there are four basic steps described in this approach. The
first is to frame a clinically relevant question and to focus it in a way
LR− = (1 − sensitivity)/specificity
which can be answered. This is not as trivial a step as it may at first
For our example, with IgA-TTG this is: appear. The key elements of a well framed question include a descrip-
tion of:
LR− = 6%/99% or 0. 06
1. the patient or population;
and for IgA-EMA: 2. the intervention or exposure;
3. the comparison intervention or exposure (if relevant); and
11%/98% or 0.11
4. the clinical outcome(s) of interest.
So a negative test result for IgA-TTG and IgA-EMA will convert our Such an approach can be applied whether the issue is one of diagnosis,
pretest odds of 1:9 or 10% to a post-test probability of around 0.7% and prognosis or therapy.
1.2% respectively. Most clinicians would not perform intestinal biopsies For example, let us consider the following scenario. You are the
in a child with the probabilities of having celiac disease. ­clinical director of a pediatrics department in a large district hospital.
What this example shows is that it is not only how a test works that An audit has demonstrated that among children who attend the pediat-
is important in interpreting the result but also the initial likelihood of ric clinics with a diagnosis of asthma, there appears to be a high rate of
disease in the child who has the test. Imagine if you did a test for IgA- attendances at the accident and emergency department and admissions
TTG on a completely well child where the probability of having celiac to hospital. You are interested in interventions which may improve the
disease is very low (perhaps less than 0.1%). If the test was negative the management of children with asthma and prevent exacerbations severe
likelihood of the child having celiac disease would be vanishingly small enough to require A&E attendances and admissions. It was suggested by
but even a positive test would mean only about an 8.5% chance that one of your colleagues that to achieve this, the department should invest
she had the condition. Using this approach to thinking about diagnostic in an asthma nurse specialist, who would initiate asthma self-manage-
tests lets you plan the rational use of tests and minimizes the chance of ment programs and education programs for all children who attended
being led astray by false positive or false negative results. outpatient clinics with asthma. The resource to provide this service has
For those familiar with statistical methods, it will be evident that the to compete with other priorities for the department and so you need to
concept of likelihood ratios has been adopted from the Bayesian approach know whether or not it might be effective. To investigate this further you
to statistical inference. The first step in any Bayesian analysis is to estab- formulate the ­following question: ‘in children with asthma, who attend
lish a prior probability distribution for the value of interest (e.g. probability hospital ­outpatients (population), does a program of education and self-
of a child having a diagnosis of gluten sensitive enteropathy). The clinical management, delivered by an asthma nurse specialist (intervention),
evidence is then used to update the prior distribution to a posterior distri- compared with no such program (comparison intervention), reduce the
bution using Bayes theorem. This is analogous to the intuitive process in risk of attendances with acute asthma at the accident and emergency
clinical history taking, where the probabilities of having one of a number ­department and admissions to hospital (outcome)?’
of different diagnoses are updated as each new piece of information is elic- Having defined the question, the second step is to undertake a com-
ited. In an evidence-based medicine approach, we start with a prior view prehensive review of the best available clinical evidence. This is likely to
about whether, for example, a drug treatment will improve clinical out- involve searching electronic databases (e.g. Medline, EMBASE, CINAHL
comes for an individual patient. This view may be informed by knowledge and The Cochrane Library). The development of search strategies for
of pathophysiology and pharmacology, the patient’s preferences, cost of these databases often involves the expertise of information scientists
therapy and so on. This prior view is then updated by the results of a ran- who can adjust the search strategy to maximize finding all relevant
domized controlled trial of the drug treatment, in a group of patients simi- studies (which increases sensitivity), but at the expense of identifying
lar to our patient, which shows clear improvement in the outcomes we many irrelevant articles (decrease specificity).
consider important. This evidence is integrated with our prior beliefs, to Having retrieved all potentially relevant articles which address the
enable a clinical decision to be made. Many have argued that the natural question, the third step is to critically appraise them. Tools have been
statistical framework for evidence-based medicine is a Bayesian approach developed to appraise studies evaluating diagnostic tests, prognostic
to decision making,5 both for the individual patient and for health policy.6 markers, treatments, adverse effects and systematic overviews.2 In the
In an article entitled ‘Narrative-based medicine in an evidence- example provided above, you will have been searching for studies which
based world’,7 Greenhalgh quotes a scenario, referred to as ‘Dr Jenkins’s are used to evaluate the effectiveness of treatments. There are many dif-
hunch’, which goes as follows: ‘I got a call from a mother who said that ferent study designs which may address this type of question. A hierar-
her little girl had had diarrhea and was behaving strangely. I knew the chy of such study designs for questions about treatment effectiveness
family well and was sufficiently concerned to break off my Monday has been developed with the most rigorous at the top and the least rig-
morning surgery and visit immediately.’ Dr Jenkins’ subsequent actions orous at the bottom (Fig. 1.1). The primary research study, which is
resulted in the child recovering from meningococcal septicemia. To considered to be the gold standard in assessing treatments, is the ran-
many this may seem like a fortunate fluke, and indeed meningococcal domized controlled trial. The key element in this design is that the allo-
septicemia presents rarely in primary care. No guideline could have reli- cation of patients to the treatment group or to the comparison group is
ably prompted Dr Jenkins’ action. However the doctor was integrating
his clinical intuition (mothers rarely use the word ‘strangely’ to describe
their children’s behavior), his acquaintance with the family (known to
be sensible and uncomplaining), with current best evidence (prognosis Systematic Review of Randomized Controlled Trials Fig. 1.1 Research
of meningococcal septicemia, with and without early administration of
Confirmed Randomized Controlled Clinical Trials pyramid of study designs
Single Randomized Controlled Clinical Trial used to assess the
parenteral penicillin) to take a course of action which saved a child’s life. Non-Randomized Controlled Clinical Trial
Greenhalgh rejects the notion that the ‘narrative of illness experience’ efficacy of treatments.
Case Controlled Observational Studies
and the ‘intuitive and subjective’ aspects of clinical method run counter Analysis of Large Computer Databases
to evidence-based medicine.7 In making clinical decisions, either infor- Case Series with Historical Controls
mally, by integrating new evidence with our prior views, or formally, Case series, Literature Control
by using Bayesian analysis, the framework exists to ensure that clinical Uncontrolled Case Series
evidence can play an explicit part in the process. Anecdotal Case
 Evidence-based child health 

done ­randomly, and thus the observed differences between the treatment strategy will almost certainly include searching electronic databases
and the comparison group(s) will be due to the experimental treatment but may involve other methods such as those designed to access unpub-
alone rather than to biases which may be introduced by patient char- lished studies, to ensure that the studies accessed are representative of
acteristics, physician preferences, etc. However, even given this robust all the research conducted in a particular area. The protocol for the
study design, the methodological quality of randomized controlled trials review should state prospectively what studies will be considered eligi-
may be subject to biases in other ways and these can be evaluated appro- ble for inclusion, defined according to study design, type of participant,
priately using checklists.2 intervention, (or exposure) and comparison intervention or exposure.
The fourth step, which is to apply the evidence in clinical practice, The outcomes of interest are stated prospectively. The protocol should
is the point at which clinical expertise and patient values are integrated also state how the quality of the included studies will be assessed.
with the best available external evidence. To do this, clinicians need to Having determined which studies are to be included and assessed
ask two questions concerning the evidence that they have appraised. them for methodological quality, the reviewer then extracts the rele-
The first is ‘is this evidence sufficiently robust for me to be confident in vant data and analyzes it. This may involve statistical methods such as
its application?’ and the second is ‘is this study applicable to the patient meta-analysis. This enables the data from a number of different studies
(or population) about whose care I am deciding?’ to be aggregated so that a pooled effect size can be estimated. There are
Practitioners of evidence-based medicine have long argued that it many examples where individual studies, usually because they are too
should not be conducted from ivory towers, by individuals who have lit- small, fail to show that an intervention is either beneficial or harmful.
tle patient contact. If it is to be applicable it needs to be practiced by all By combining the results from all potentially relevant studies, such ben-
clinicians, not least because the questions or question posed need to be efit or harm has been clearly demonstrated. This enables questions to be
relevant to routine practice. However, as will be apparent from the brief answered such as ‘does this treatment have a beneficial effect?’ and also
description above, the second and third steps of this approach can be ‘what is the size of the effect?’
very laborious. This has led to the development of a number of short- There now exists a database of systematic reviews of randomized
cuts which will enable the clinician to move easily from step one to step controlled trials of interventions across the whole range of health care.
four. To be reliable, such shortcuts must use rigorous methods. One of This is found in The Cochrane Library,8 which, at the time of writing (Disk
the most attractive and reliable of these tools is the development of sys- Issue 4, 2006), contains 2905 systematic reviews. These reviews have
tematic reviews. As will be seen from Table 1.1, the systematic review is been prepared by the Cochrane Collaboration, an international body ded-
now regarded by many as the most rigorous study design for providing icated to producing systematic reviews of the effects of health care using
information about treatments. The term systematic or scientific review is randomized controlled trials as the primary study design. The Cochrane
used to distinguish them from more traditional or narrative type reviews Library is produced as a CD-ROM, is updated quarterly and is accessible
of topic areas conducted by experts in the field. on the Internet http://www.cochrane.org/ This means that systematic
reviews can also be updated to take account of new knowledge.
Systematic Reviews For the question described above, ‘in children with asthma, who
attend hospital outpatients (population), does a program of education
Narrative reviews have become a regular feature of many medical jour- and self-management, delivered by an asthma nurse specialist (interven-
nals. Such reviews may be very informative, lively and interesting and tion), compared with no such review (comparison intervention), reduce
are usually well illustrated. They are popular because clinicians are busy the risk of attendances with acute asthma at the accident and emergency
and have limited time to try and assimilate all primary research which department and admissions to hospital (outcome)?’, a Cochrane review,
is relevant in a particular area. However, if judged as scientific work entitled, ‘Educational interventions for asthma in children’9 addresses
which provides summaries of the evidence, which can be used to guide this question. It is a systematic review of randomized controlled trials,
diagnosis or treatment, they are very subject to bias and are therefore which included children, aged 2–18 years. Studies which assessed any
unreliable. educational interventions targeted to children or adolescents (and/or
These deficiencies led to the concept of systematic reviews, which their parents) designed to teach self-management strategies related
should adopt a rigorous scientific methodology to eliminate systematic to prevention, attack management, or social skills using any instruc-
bias or random error, as would be expected of investigators undertak- tional strategy or combination of strategies (problem-solving, role-play-
ing primary clinical research. The methodology for systematic reviews ing, videotapes, computer assisted instruction, booklets, etc.) presented
is outlined in Table 1.1. First the reviewer needs to state the hypothesis either individually or in group sessions were included in the review. The
that they wish to investigate and then prospectively define a compre- outcomes assessed in the included studies were health care utilization
hensive search strategy to identify all potentially relevant studies. This outcomes, which were the main interest in our scenario, but also physi-
ological measures, morbidity and functional status and self-perception.
Table 1.1 How to conduct a systematic review A number of the studies included self-management programs delivered
1. State objectives and hypotheses by a specialist nurse and a high proportion of the studies took place in
2. Outline eligibility criteria, stating types of study, types of participant, an outpatient setting. For the outcomes of interest in this scenario, A&E
types of intervention and outcomes to be examined attendences and hospital admissions, 12 trials reported A&E department
3. Perform a comprehensive search of all relevant sources for potentially visits as an outcome and when the results of these trials were pooled,
eligible studies self-management education programs were associated with a reduc-
4. Examine the studies to decide eligibility (if possible with two tion in this outcome (standardized mean difference −0.21, 95% confi-
independent reviewers) dence intervals −0.33 to −0.09). However, when the results, for hospital
5. Construct a table describing the characteristics of the included studies admissions, of the eight trials which reported this outcome were pooled,
6. Assess methodological quality of included studies (if possible with two there was no significant benefit of educational interventions (standard-
independent reviewers) ized mean difference −0.08, 95% CI −0.21 to 0.05). Self-management
7. Extract data (with a second investigator if possible), with involvement and education was associated with improvements in lung function, days
of investigators if necessary of school absence and days of restricted activity. All of this information
8. Analyze results of included studies, using statistical synthesis of data will help informed decision-making in planning the delivery of care for
(meta-analysis), if appropriate
children with asthma in an outpatient hospital setting.
9. Prepare a report of review, stating aims, materials and methods and
Although well conducted systematic reviews are regarded as being
describing results and conclusions
at the top of the hierarchy of study designs to evaluate the effectiveness
Courtesy of Craig JV, Smyth RL. Evidence based practice manual for nurses. of treatments, they can be misused or be badly conducted. Therefore
Edinburgh: Churchill Livingstone, 2002. systematic reviews need also to be appraised and there are checklists
 Foundations of health and practice

available for doing this. Jadad et al,10 for example, have published an that new drugs brought to the market should be tested in children unless
evaluation of reviews and meta-analyses of treatments used in asthma. there were compelling reasons not to do so. In the European Union, the
Of the 50 reviews that they considered, 40 were found to have seri- European Regulation on medicinal products for pediatric uses, which
ous methodological flaws. Included within these 40 were six reviews became law in early 2007, provides a system of requirements and incen-
funded by the pharmaceutical industry. All but one of these six reviews tives for pharmaceutical companies to agree a pediatric investigation
had results and conclusions that favored the intervention related to plan for all new drugs which may have an indication in children. There
the company which sponsored the review. Reassuringly they found is a provision, within the Regulation, for a European research network,
that Cochrane Reviews had higher overall quality scores than reviews of existing national and European networks, investigators and centres,
­published in other scientific journals. The Cochrane Collaboration has with expertise in conducting clinical studies with children. In the UK, the
made strenuous efforts to ensure that its methodology reduces bias to a Medicines for Children Research Network was established in 2005, with
minimum and has also guarded against biases which may be introduced substantial government funding.16
by authors with conflicts of interest, e.g. significant support from the
pharmaceutical industry. This methodology and the external appraisal How textbooks ensure that they
by individuals such as Jadad provide reassurance that systematic are evidence based
reviews produced by the Cochrane Collaboration are reliable sources of
­information for the clinician. Most manuals written for evidence-based practitioners do not consider
textbooks to be a valuable source of information. Sackett et al2 in the
Evidence-Based Child Health relevant section in their textbook Evidence-based Medicine state ‘we begin
with textbooks only to dismiss them’ later stating ‘while we may find
The evidence-based medicine movement was initially advanced by some useful information in textbooks about the pathophysiology of
practitioners of internal medicine, but, more recently, pediatricians clinical problems it is best not to use them for establishing the cause,
have become leading proponents. The scope of a number of Cochrane diagnosis, prognosis, prevention or treatment of a disorder’. Part of the
Collaborative Reviews Groups (such as the Cystic Fibrosis and Genetic problem is that the material published in textbooks is written some time
Disorders Group) covers areas that deal mainly with illnesses of child- (often some years) before the book is published. Thus textbooks rap-
hood. In addition, there is a Cochrane Child Health Field which has the idly become out of date. Publishers have realized this and are trying to
goal of supporting the production of child-focused, clinically relevant, address this with strategies such as providing regular electronic updates
methodologically rigorous systematic reviews. However, one feature more frequently than the published paper editions.
that this discipline of evidence-based health care has highlighted is that However, some textbooks have been found not to include informa-
when children’s health is compared to adult health care, research ques- tion which is current at the time they are written. The Oxford Textbook of
tions in children may have been addressed either not at all or by small, Medicine has been criticized for including a statement in its 2nd edition
poorly designed studies.11 This was the case when all randomized trials concerning the clinical benefits of thrombolysis for patients who had
published during a 15-year period in a specialist pediatric journal were had a myocardial infarction, which stated that these benefits had not
examined. In this review sample sizes were found to be generally small been established. This statement was made some years after this therapy
and only a small proportion of studies were multicenter.12 Subspecialty had been shown, in a systematic review of randomized controlled trials,
areas within pediatrics have also been reviewed, such as cystic fibrosis,13 to reduce the risk of premature death after myocardial infarction.17 By
pediatric rheumatology14 and community pediatrics,15 and the conclu- way of further illustration Jefferson18 described his experience in con-
sions about the volume and quality of the research have been similar. ducting a Cochrane Review of the effects of cholera vaccine. Although
The 6th edition of Forfar & Arneil was the first to ask contributors to be much of the data available for this review concerned older, killed whole
more explicit about the evidence that they cite and led to contributors cell cholera vaccine, the reviewers were aware that the killed whole cell
expressing some frustration at the lack of level 1 evidence available for vaccine had been discouraged as it had become widely accepted that it
them to access in many different disease areas. had a low efficacy and short duration of effect, required multiple doses
There are a number of reasons why it is more difficult to undertake and was less effective in children under 5 than in adults. However, the
clinical trials and other rigorous studies in children than in adults. There systematic review found that the efficacy of the vaccine compared
are obvious ethical dilemmas. For example, research on a new ­therapy with placebo was over 50% in both the first 7 months and in the first
is often conducted first in adults before studies are undertaken in year and just under 50% in the second year and that most of the tri-
children. However, if data from a study of a therapy in adults show a als achieved this efficacy using a single dose. Vaccine efficacy in the first
clear advantage for a drug over placebo, does equipoise still exist and year was also as great in children under 5 as in older people. The review-
is it ­ethical to repeat such a study in children? Such ethical dilemmas ers concluded that the level and duration of efficacy of the killed whole
need to be considered clearly and dispassionately by individuals who are cell cholera vaccine had been underestimated in the literature and that
advocates for the interests of children. However, there are many exam- the incidence of adverse effects had been overestimated. A further sur-
ples of therapies which have different effects depending on the stage of vey of journal editors and authors of reviews of cholera vaccines con-
the disease process or the age of the patient, including, for example, the cluded that many narrative reviews had been written using the so called
sedative effects of phenobarbital in adults, but its frequent association ‘desk drawer method’. This involved including the evidence that was
with hyperactivity in children. known to reviewers, but not assembling it or evaluating it systemati-
There are other practical problems. Generally the proportions of cally. Now that nearly 3000 Cochrane Reviews and 479 462 controlled
children affected by chronic diseases are smaller than in adults, and trials of interventions are available on The Cochrane Library, there can
even for common childhood diseases such as asthma, the condition may be no excuse for the reviewer, or indeed textbook writer, not to consult
be more heterogeneous than in adults and diagnostic criteria less pre- The Cochrane Library when considering treatments for the condition or
cise. Outcome measures which have been developed and validated in conditions they are writing about.
adults, such as quality of life measures, are unlikely to be appropriate or Many textbooks now contain the term ‘evidence-based’ within their
feasible for young children and infants. title. These fall into two groups. Firstly the manuals that aim to instruct
These difficulties have meant that pediatricians, wishing to practice the clinician on how to practice evidence-based medicine and secondly
evidence-based child health, have relatively little high quality evidence textbooks which purport to present evidence-based information which
on which to base their decisions. However, there are encouraging signs has been synthesized in a rigorous manner and presented in an easily
that as well as identifying the deficiencies in the evidence, initiatives are accessible format. Such textbooks include Evidence-based Pediatrics and
being made to promote research that will meet these needs. For exam- Child Health19 and Clinical Evidence.20 Clinical Evidence, which states at the
ple, the Food and Drug Administration in the USA mandated in 1997 outset that it is ‘not a textbook of medicine’, is rather a handbook, or
 Evidence-based child health 

a reference guide, of topics of wide general interest in health care. In space’. He describes his practice of ‘in every case to read the subject up
the ‘Compendium of evidence’, at the start of the book, the methods carefully; to compare the knowledge thus acquired or renewed with the
used are clearly presented. However, its question and answer presenta- results of my own experience, in those cases in which I had any experi-
tion, limitation to interventions and incomplete coverage of all subjects ence: and then, having taken a more or less definite view of the whole
within its scope make it rather different from a textbook such as this one. subject and while my mind was still full of it, and its details, to write as
Evidence-based Pediatrics and Child Health describes itself as a ‘melding clear and as comprehensive an account as I was capable of ’. He states
of a textbook of evidence-based medicine and a clinical pediatric text that this method of procedure will partly explain ‘the prevailing absence
addressing common conditions’. The first section of this book provides of notes, quotations and references to authorities’. This textbook, which
readers with the skills needed to practice evidence-based child health, ran to over 1000 pages, was written entirely by Bristowe and contained
while the second and third sections address common pediatric condi- no references. Indeed it was unusual for textbooks to contain more than
tions. Again the format is in a question and answer framework, with a few references per chapter until the middle of the twentieth century.
illustration by scenarios. Within the second and third sections, only The method used by Bristowe is similar to that used by many authors
those specific questions for which evidence is available are considered today, and the depth of the individual’s experience contributes to the
and as a consequence much of clinical pediatrics is not discussed. richness of the narrative and thus its readability.
A large and complete textbook of pediatrics such as this one needs to So whilst retaining these traditional methods for the provision of
provide a detailed consideration of all aspects of a condition, not just diag- background information we have, however, asked all authors to search
nosis, prognosis, prevention and treatment (which are more amenable to for and use the best available evidence when presenting ‘foreground’
evidence-based approaches). Moyer and Elliott1 refer to two sorts of ques- information. The methods that we have used to do this will be discussed
tions which may be used by the reader in trying to elicit information about in the next section.
a topic, including background information such as: what is Cornelia de
Lange syndrome, how commonly is tracheoesophageal fistula associated Clinical evidence within the 7th
with esophageal atresia and what is meant by the term apoptosis? These edition of forfar & arneil
are distinguished from ‘foreground’ questions relating to, for example,
benefits or adverse effects of therapy. Textbooks therefore need to consider For the 6th and 7th editions of Forfar & Arneil the editors gave specific
the definition, pathophysiology and clinical presentation of conditions as instructions to contributors to ensure that their contributions were
well as diagnosis, prognosis, prevention or treatment of the disorder. as evidence-based as possible. The contributors were asked to iden-
In considering how textbook writers put this information together in tify, where possible, ‘level 1 evidence’. This system of grading evidence
a readable format it is helpful to consider how this process has evolved and recommendations was chosen because it is widely used and has
over decades or even centuries. Bristowe in the preface to the 1876 first been developed over the last 20 years. The table which contributors
edition of his Theory and Practice of Medicine21 states his primary aim as ‘to were asked to use (Table 1.2) is available on the Website of the Oxford
give in a readable form as much information as I could within a ­limited Centre for Evidence-based Medicine (http://cebm.jr2.ox.ac.uk).22 This

Table 1.2 Oxford Centre for Evidence-Based Medicine levels of evidence (December 2006)
Level Therapy/prevention, Prognosis Diagnosis Differential diagnosis/ Economic and decision
etiology/harm symptom prevalence analyses
study

1a SR (with homogeneity*) SR (with homogeneity*) SR (with homogeneity*) SR (with homogeneity*) SR (with homogeneity*)
of RCTs of inception cohort of level 1 diagnostic of prospective cohort of level 1 economic
studies; CDR† validated studies; CDR† with 1b studies studies
in different populations studies from different
clinical centers

1b Individual RCT (with Individual inception Validating** cohort Prospective cohort Analysis based on
narrow confidence cohort study with study with good††† study with good clinically sensible
interval‡) ≥80% follow-up; CDR† reference standards; or follow-up**** costs or alternatives;
validated in a single CDR† tested within one systematic review(s)
population clinical center of the evidence; and
including multiway
sensitivity analyses

1c All or none§ All or none case-series Absolute SpPins and All or none case-series Absolute better-value
SnNouts†† or worse-value
analyses††††

2a SR (with homogeneity*) SR (with homogeneity*) SR (with homogeneity*) SR (with homogeneity*) SR (with homogeneity*)
of cohort studies of either retrospective of level >2 diagnostic of level 2b and better of level >2 economic
cohort studies or studies studies studies
untreated control
groups in RCTs

2b Individual cohort study Retrospective cohort Exploratory** cohort Retrospective cohort Analysis based on
(including low quality study or follow-up of study with good††† study, or poor follow-up clinically sensible costs
RCT; e.g. <80% untreated control patients reference standards; or alternatives; limited
follow-up) in an RCT; Derivation CDR† after derivation, review(s) of the evidence,
of CDR† or validated on or validated only on or single studies; and
split-sample§§§ only split-sample§§§ or including multiway
databases sensitivity analyses
(Continued)
 Foundations of health and practice

Table 1.2 Oxford Centre for Evidence-Based Medicine levels of evidence (December 2006)—cont’d
Level Therapy/prevention, Prognosis Diagnosis Differential diagnosis/ Economic and decision
etiology/harm symptom prevalence analyses
study
2c ‘Outcomes’ research; ‘Outcomes’ research Ecological studies Audit or outcomes
ecological studies research

3a SR (with homogeneity*) SR (with homogeneity*) SR (with homogeneity*) SR (with homogeneity*)

of case-control studies of level 3b and better of level 3b and better of level 3b and better
studies studies studies

3b Individual case-control Non-consecutive study; Non-consecutive cohort Analysis based on limited

study or without consistently study, or very limited alternatives or costs,
applied reference population poor quality estimates
standards of data, but including
sensitivity analyses
incorporating clinically
sensible variations

4 Case-series (and poor Case-series (and poor Case-control study, poor Case-series or superseded Analysis with no
quality cohort and quality prognostic cohort or non-independent reference standards sensitivity analysis
case-control studies§§) studies***) reference standard

5 Expert opinion without Expert opinion without Expert opinion without Expert opinion without Expert opinion without
explicit critical appraisal, explicit critical appraisal, explicit critical appraisal, explicit critical appraisal, explicit critical appraisal,
or based on physiology, or based on physiology, or based on physiology, or based on physiology, or based on physiology,
bench research or ‘first bench research or ‘first bench research or ‘first bench research or ‘first bench research or ‘first
principles’ principles’ principles’ principles’ principles’

Produced by Bob Phillips, Chris Ball, Dave Sackett, Doug Badenoch, Sharon Straus, Brian Haynes, Martin Dawes since November 1998.
Notes – Users can add a minus-sign ‘−’ to denote the level that fails to provide a conclusive answer because of:
• EITHER a single result with a wide confidence interval (such that, for example, an ARR in an RCT is not statistically significant but whose confidence intervals fail to exclude
clinically important benefit or harm)
• OR a systematic review with troublesome (and statistically significant) heterogeneity.
• Such evidence is inconclusive, and therefore can only generate Grade D recommendations.
* By homogeneity we mean a systematic review that is free of worrisome variations (heterogeneity) in the directions and degrees of results between individual studies. Not
all systematic reviews with statistically significant heterogeneity need be worrisome, and not all worrisome heterogeneity need be statistically significant. As noted above,
studies displaying worrisome heterogeneity should be tagged with a ‘−’ at the end of their designated level.
†
Clinical Decision Rule (these are algorithms or scoring systems which lead to a prognostic estimation or a diagnostic category).
‡
See note #2 for advice on how to understand, rate and use trials or other studies with wide confidence intervals.
§
Met when all patients died before the treatment became available, but some now survive on it; or when some patients died before the treatment became available, but none
now die on it.
§§
By poor quality cohort study we mean one that failed to clearly define comparison groups and/or failed to measure exposures and outcomes in the same (preferably blinded),
objective way in both exposed and non-exposed individuals and/or failed to identify or appropriately control known confounders and/or failed to carry out a sufficiently long
and complete follow-up of patients. By poor quality case-control study we mean one that failed to clearly define comparison groups and/or failed to measure exposures and
outcomes in the same (preferably blinded), objective way in both cases and controls and/or failed to identify or appropriately control known confounders.
§§§
Split-sample validation is achieved by collecting all the information in a single tranche, then artificially dividing this into ‘derivation’ and ‘validation’ samples.
††
An ‘Absolute SpPin’ is a diagnostic finding whose Specificity is so high that a Positive result rules-in the diagnosis. An ‘Absolute SnNout’ is a diagnostic finding whose
Sensitivity is so high that a Negative result rules-out the diagnosis.
†††
Good reference standards are independent of the test, and applied blindly or objectively to all patients. Poor reference standards are haphazardly applied, but still
independent of the test. Use of a non-independent reference standard (where the ‘test’ is included in the ‘reference’, or where the ‘testing’ affects the ‘reference’) implies a
level 4 study.
††††
Better-value treatments are clearly as good but cheaper, or better at the same or reduced cost. Worse-value treatments are as good and more expensive, or worse and
equally or more expensive.
**Validating studies test the quality of a specific diagnostic test, based on prior evidence. An exploratory study collects information and trawls the data (e.g. using a regression
analysis) to find which factors are ‘significant’.
*** By poor quality prognostic cohort study we mean one in which sampling was biased in favor of patients who already had the target outcome, or the measurement of
outcomes was accomplished in <80% of study patients, or outcomes were determined in an unblinded, non-objective way, or there was no correction for confounding factors.
**** Good follow-up in a differential diagnosis study is >80%, with adequate time for alternative diagnoses to emerge (e.g. 1–6 months acute, 1–5 years chronic).
RCT, randomized controlled trial; SR, systematic review.
Grades of recommendation:
A. Consistent level 1 studies
B. Consistent level 2 or 3 studies or extrapolations from level 1 studies
C. Level 4 studies or extrapolations from level 2 or 3 studies
D. Level 5 evidence or troublingly inconsistent or inconclusive studies of any level
‘Extrapolations’ are where data are used in a situation which has potentially clinically important differences from the original study situation.
 Evidence-based child health 

type of approach has been used in other evidence-based textbooks.23 Summary

The editors appreciated that in a number of areas there may not be
level 1 evidence available, but to distinguish this category of evidence Clinical decision making is a complex process. Evidence-based ­medicine
from the remainder, authors were asked to identify references to proponents endorse the importance of clinical expertise, which uses
level 1 evidence with an asterisk (*). This approach may be consid- narrative skills to integrate the information provided by the clinical
ered inadequate by some evidence-based proponents but was felt to be method. In practicing evidence-based medicine, clinicians integrate
an important step in ensuring that the standard of information avail- this information with that provided by sound clinical research. When
able from this textbook was as high as possible. The grading system tensions arise in this approach, it is usually because the ­ narrative/
used has some problems and a number of these are acknowledged intuitive paradigm has been discarded and decision ­making is ­predicated
(http://cebm.jr2.ox.ac.uk). For example, definitions of homogeneity on the ‘evidence’ alone. Systematic reviews are the result of rigorous
(with respect to systematic reviews) and narrow confidence intervals research, using methodology which is designed to reduce bias. As such
(with respect to randomized controlled trials) are open to considerable they can provide reliable summaries of evidence. This methodology has
interpretation and may not adequately distinguish high quality stud- been most comprehensively developed for systematic reviews of ran-
ies from those of poor quality. It is also important to realize that this domized controlled trials, but is being developed for other study designs.
particular ‘hierarchy of evidence’ applies only to questions about the By searching databases of systematic reviews, such as those contained
effectiveness of interventions – other types of questions need different within The Cochrane Library, the clinician can obtain high quality evi-
research approaches. dence rapidly and reliably and enhance his/her clinical practice.

References (*Level 1 Evidence) 7. Greenhalgh T. Narrative based medicine in an 15. Polnay L. Research in community child health.
evidence based world. BMJ 1999; 318:323–325. Arch Dis Child 1989; 64:981–983.
1. Moyer V, Elliott E. Preface. In: Moyer V, Elliott E, 8. Cochrane Library. The Cochrane Library, Issue 3, 16. Smyth RL, Edwards D. A major new initiative to
Davis R, et al, eds. Evidence based pediatrics and child 2001. Oxford: Update Software. improve treatment for children. Arch Dis Child 2006;
health. London: BMJ Books; 2000. 9. Wolf FM, Guevara JP, Grum CM, et al. Educational 91:212–213.
2. Sackett DL, Richardson WS, Rosenberg W, et al. interventions for asthma in children. Cochrane 17. Chalmers I, Haynes B. Systematic reviews:
Evidence-based medicine: how to practice and teach Database Syst Rev 2002; (4):CD000326. reporting, updating, and correcting systematic reviews
EBM. London: Churchill Livingstone; 2000. 10. Jadad AR, Moher M, Browman G, et al. Systematic of the effects of health care. BMJ 1994; 309:862–865.
3. McAlister F, Straus S, Sackett DL. Why we need reviews and meta-analyses on treatment of asthma: 18. Jefferson T. What are the benefits of editorials
large, simple studies of the clinical examination: critical evaluation. BMJ 2000; 320:537–540. and non-systematic reviews? BMJ 1999; 318:135.
the problem and a proposed solution. Lancet 1999; 11. Smyth RL, Weindling AM. Research in children: 19. Moyer VA, Elliott EJ, Davis RL, et al, eds. Evidence-
354:1721–1724. ethical and scientific aspects. Lancet 1999; 354: based pediatrics and child health. London: BMJ Books;
4. Collin P, Kaukinen K, Vogelsang H, et al. (suppl 2):SII21–SII24. 2000.
Antiendomysial and antihuman recombinant 12. Campbell H, Surry S, Royle E. A review of 20. Barton S. Clinical evidence. London: BMJ
tissue transglutaminase antibodies in the diagnosis randomised controlled trials published in Archives Publishing; 2001.
of celiac disease: a biopsy-proven European of Disease in Childhood from 1982–1996. Arch Dis 21. Bristowe J. Theory and practice of medicine.
multicentre study. Eur J Gastroenterol Hepatol Child 1997; 79:192–197. London: Smith, Elder & Co; 1882.
2005; 17:85–91. 13. Cheng K, Smyth RL, Motley J, et al. Randomized 22. Website of Oxford Centre for EBM. Oxford Centre
5. Ashby D, Smith AFM. Evidence-based medicine controlled trials in cystic fibrosis (1996–1997) cate­ for Evidence-based Medicine; 2001. http://cebm.jr2.
as Bayesian decision-making. Stat Med 2000; gorized by time, design, and intervention. Pediatr ox.ac.uk
19:3291–3305. Pulmonol 2000; 29:1–7. 23. Feldman W. Evidence-based pediatrics. Ontario:
6. Lilford R, Braunholtz D. The statistical basis of 14. Feldman B, Giannini E. Where’s the evidence? BC Decker; 2000.
public policy: a paradigm shift is overdue. BMJ 1996; Putting the clinical science into pediatric rheuma-
313:603–607. tology. J Rheumatol 1996; 23:1502–1504.
 10001

2
Epidemiology of childhood diseases
Stuart Logan
Introduction 11 Population determinants of child health 16
Mortality rates 11 Age and sex structure of the population 16
Definitions 11 Genetic factors 16
Patterns of mortality 11 Birth weight and gestation 16
Mortality in poor countries 12 Physical environment 17
Childhood mortality in the industrialized world 13 Social factors 17
Morbidity in childhood 14 Health behaviors 17
Health service use 14 Health services 18
Chronic disease and disability 15

Introduction who fulfill this definition are not in fact registered as live births and that,
as the frontiers of neonatal intensive care have been pushed back over
Children under 15 comprise about one third of the world population, time, this proportion has changed. In the UK a stillbirth is defined now
with three quarters living in less developed countries. Children’s health as being ‘a child born at 24 or more weeks post conception who shows
varies enormously across the world. In 2003 the under-5 mortality rate no signs of life’, although, up until 1992, the definition required that
was 6 per 1000 in industrialized countries but 175 per 1000 in ­­ sub- they were born at least 28 weeks’ post conception.
Saharan Africa.1 The definitions of the various perinatal, neonatal and infant mortal-
Understanding the patterns of health and disease is important in ity rates (IMRs) are shown graphically in Figure 2.1. The denominator
planning health and social policies and in monitoring change over time. for the stillbirth and perinatal mortality rates is the number of still and
It is important to recognize the limitations of the data that are avail- live births while that for the other rates is the number of live births.
able. If the statistics are to be useful for comparing between areas and The under-5 mortality rate, widely used, particularly in poorer
over time, effective systems must be in place to collect the data and there countries, is defined as the annual number of deaths in children under 5
must be consistency of definitions. In many parts of the world, such sys- years of age per 1000 live births. Age-specific death rates are the num-
tems are rudimentary, although considerable progress has been made in ber of deaths in an age group per 1000 individuals in that age group.
ensuring agreed definitions at least for the principal measures of birth
and mortality. The World Health Organization is an excellent source of
regularly updated health data which can be accessed on their website1
Patterns of mortality
and this chapter draws heavily on these data and can be assumed to be There have been dramatic changes in life expectancy in developed
the source where no other reference is given. countries over the last century. In 1901, the average life expectancy for
This chapter will review the patterns of mortality and morbid- women in the UK was 48 years, but by 2004 it was 81 years. Life expec-
ity amongst children and examine the major determinants of health tancy for men in the UK, although lower than for women, at 76 years,
at a population level. These patterns vary widely between countries. has also risen greatly2 (except where specifically referenced, UK data are
This chapter will discuss mainly the extremes, with data drawn from from the Office of National Statistics, whose website gives access to an
the rich industrialized nations and the poorest nations, largely in sub- enormous range of current and historical data). Although death rates
Saharan Africa. Many countries will of course have rates of mortality at all ages have declined, much of this change has been due to the rapid
and ­patterns of disease lying between these extremes. decrease in deaths in childhood, particularly in the first half of the cen-
In describing the health of children at a population level we are usu- tury. Changes in death rates during childhood over time for England and
ally forced to rely on measures of mortality or morbidity. Measures of Wales are shown in Figure 2.2.2
health rather than disease are philosophically attractive but have proved In Tanzania by contrast the life expectancy for women in 2004 was
problematic in practice. Although a number of tools have been devel- 49 and for men, 47. Much of the difference in life expectancy between
oped to measure quality of life in childhood their application at a pop- the UK and Tanzania is driven by differences in childhood mortality, par-
ulation level has been limited and they have mainly proved useful in ticularly mortality under the age of 5. The probability of dying between
clinical trials or in the investigation of the effects of specific conditions. 15 and 60 years of age per 1000 population was 102 in the UK and 5.5
times higher (552 per 1000) in Tanzania. However, the under-5 mortal-
ity rate in the UK in 2004 was 6 per 1000 live births and 126 in Tanzania,
Mortality rates 21 times higher. All these figures are for males, rates for females showing
similar patterns but lower absolute rates. Not only are the relative differ-
Definitions ences larger in childhood but each childhood death contributes more to
Any infant who breathes, has a heart beat or pulsation of the umbilical the total years of life lost and hence the life expectancy figures.
cord is defined as a ‘live birth’, irrespective of gestation or the duration of Not only are absolute mortality rates higher in poorer countries but
the signs of life. It seems likely that at least some very premature infants the pattern of change over time is also different. Globally the under-5 11
12 foundations of health and practice

Table 2.1 Changes in under-5 and infant mortality rates per 1000 live
Post neonatal# births over time by region1
Under-5 mortality rate Infant mortality rate
Late neonatal#
Region 1960 2003 1960 2003
Neonatal#
Industrialized nations 39    6 32    5
Early Developing countries 224 87 142 60
neonatal# Least developed 278 155 127 54
countries
Perinatal* Sub-Saharan Africa* 278 175 165 104
World 198 80 127 54
Stillbirth*
*Most countries in sub-Saharan Africa are included within the category ‘least
developed countries’ but the region is shown separately to emphasize the slow rate
24 weeks post Birth 1 week 28 days 1 year of improvement.
conception

Fig. 2.1 Definitions of mortality rates in the neonatal period and infancy.
Mortality in poor countries
4.0 The burden of mortality in children in poor countries is extraordinary.
In sub-Saharan Africa some 10% of infants die in the first year of life
3.5 1 to 4
(compared to around 0.5% in Europe) and over 15% before their 5th
5 to 9 birthday. The proportions of deaths related to major causal groups is
Death rate per 100 population

3.0 10 to 14 shown in Figure 2.3 for under-5s in Africa and in western Europe3. Nearly
three quarters of these deaths are due to six causes: (1) pneumonia
2.5
(19%), (2) diarrhea (18%), (3) neonatal sepsis or (4) pneumonia (10%),
2.0 (5) preterm delivery (10%) or (6) asphyxia at birth (8%). It is estimated
that undernutrition is an underlying cause in over 50% of deaths in
1.5 under-5-year-olds.
The first year of life is the most dangerous period of childhood and
1.0 globally 40% of under-5 deaths occur in the neonatal period. The pro-
portion is lower in Africa (26%) because of the higher burden of deaths
0.5 in the post-neonatal period, although the neonatal death rates are high-
est here. As in the resource rich world, the consequences of prematu-
0 rity and low birth weight are important causes of neonatal deaths in
1941 1951 1961 1971 1981 1991 2001 2005 resource limited countries but infections (including neonatal tetanus)
Year are also important (Fig. 2.4).3 The high risks related to underlying
Fig. 2.2 Death rates in children in England and Wales over time.2 poverty are exacerbated by the lack of services. In Africa only 40% of
women deliver with skilled care and in South Asia less than 30%.4
Communicable diseases are responsible for around half of the under-
5 deaths in the world and nearly two thirds in Africa. Malaria is particu-
mortality rate fell from 198 in 1000 in 1960 to 80 in 1000 in 2003. larly important in Africa with 94% of all malaria deaths occurring in this
This hides a huge variation in the rate of decline, with the least progress region. A similar picture is seen with deaths from human immunodefi-
generally being made in countries which started with the highest rates. ciency virus (HIV)/acquired immune deficiency syndrome (AIDS) where
In 93 countries containing 40% of the world’s population, < 5 mortal- 89% of deaths are in Africa although this picture is likely to change as
ity is declining fast, in 51 (48% of the population) progress is slow and the prevalence of the infection rises in poor countries of Southeast Asia.
in the remaining 43 (12% of the population), mostly clustered in sub- The deaths of over 400 000 children per year due to measles is depress-
Saharan Africa, rates are stagnant or rising (Table 2.1). ing considering the availability of a cheap and ­effective vaccine which

40 Fig. 2.3 Percentage of deaths of under-5-

year-olds by cause in the world and in Africa
35
2000–2003.3
Percentage deaths by cause

30
25
20
15
10
5
0
Neonatal
Pneumonia Diarrhea Malaria Measles HIV/AIDS Injuries Other
causes
World 19 17 8 4 3 37 3 10
Africa 21 16 18 5 6 26 2 5
 epidemiology of childhood diseases 13

45 Fig. 2.4 Percentage of neonatal deaths

40 by cause in Africa and western Europe
2000–2003.3
Percentage deaths by cause
35
30
25
20
15
10
5
0
Neonatal Congenital Birth
Infection Diarrhea Prematurity Other
tetanus anomalies asphyxia
Africa 26 7 3 28 8 23 7
Western Europe 6 0 0 41 32 15 6

Table 2.2 Proportion < 5 deaths by cause in the 42 countries in which 10–14 years 10%
90% of all < 5 deaths occur which are potentially preventable5
5–9 years 7% Early neonatal 43%
Disease or condition % total deaths % preventable
Diarrhea     22     88
Pneumonia     21     65
Malaria     9     91
HIV/AIDS     3     48
Measles     1    100
Post neonatal 26%
Neonatal disorders     33     55

has virtually eliminated deaths from this condition in countries with

Late neonatal 14%
high vaccine coverage.
The vast bulk of these deaths are related directly or indirectly to pov- Fig. 2.5 Proportion of deaths in childhood by age of occurrence (England
erty. Knowledge already exists about effective interventions for both and Wales 2005).2
treatment and prevention, which could substantially reduce this burden.
It has been estimated that about two thirds of these deaths could be pre-
160
vented by a small number of key interventions of proven effectiveness
which could feasibly be introduced in low-income countries (Table 2.2).5 140 IMR*
The interventions considered are relatively cheap and simple including 120 SBR**
measures such as the use of oral rehydration fluid in diarrhea, insec-
ticide treated materials for the prevention of malaria, antibiotics for 100
Death rate

neonatal sepsis and pneumonia and encouragement of high rates of 80

breast-feeding.
60
40
Childhood mortality in the industrialized
world 20

Of all deaths in childhood in the UK (ages 0–14), 73% occur within the 0
first year of life, 50% within the first month and 38% within the first
–5

–5

–5

–5

–5

–5

–5

–5

–5

–5

–5
01

11

21

31

41

51

61

71

81

91

01

week (Fig. 2.5).2 The decline in the rate of stillbirths and infant mortal-
19

19

19

19

19

19

19

19

19

19

20

ity in England and Wales is shown in Figure 2.6. The rates for stillbirth Year
after 1992 are not strictly comparable because of the change in defini- Fig. 2.6 Changes in stillbirth and infant mortality rates over time
tion mentioned earlier. The decline in infant mortality reflects declines (England and Wales).2 * Per 100 live births. ** Per 100 live and stillbirths.
in both neonatal and post-neonatal mortality rates.
In the neonatal period, a substantial proportion of deaths are related
to congenital anomalies and prematurity (Fig. 2.4). Congenital anoma- injuries and poisoning are responsible for a significant number of deaths
lies are an important cause of death in the neonatal period although in the first year of life, but in this age group, the much higher rates of
the birth prevalence of many anomalies appears to have declined, par- death from other conditions mean that they are responsible for a rela-
ticularly that of anomalies of the central nervous system. While part of this tively small proportion of all deaths.
decline appears to relate to the widespread introduction of screening Even within the industrialized world there are substantial variations
for neural tube defects in pregnancy there has also been a substantial in IMRs (Table 2.3)1. While there is an obvious relationship between
decline in incidence, possibly related to changes in diet or the use of peri- mortality rates and wealth there are some surprising anomalies. For
conceptual folate supplements. example, the USA has an IMR more than twice that of Sweden. To some
Congenital anomalies remain an important cause of death after the extent, this appears to reflect differences in IMRs in different popula-
neonatal period, accounting for 22% of all deaths between 1 month and tion groups within the same country. In the USA, in 2002, the IMR for
1 year of age in England and Wales in 2005. As with older children, infants born to non-Hispanic black mothers was 2.4 times higher than
14 foundations of health and practice

Table 2.3 Infant mortality rates in some selected industrialized 2500−2999 17.1%
countries (2002)1
Country IMR (per 1000 live births) 3000−3499 35%
2000−2499 4.8%
Sweden 3.3 <2000 2.7%
Norway 3.5 1500−1999 1.5%
Austria 4.1 1000−1499 0.7%
France 4.1 <1000 0.5%
Germany 4.2
Italy 4.5 4000+ 11%
The Netherlands 5.1
Ireland 5.1
UK 5.2
3500−3999 28%
Canada 5.4
New Zealand 5.6 Fig. 2.8 Proportion of live births by birth weight in grams, England
United States 7.0 and Wales (2004).2

The risk of death drops rapidly after the first year of life; in the UK
in non-Hispanic white, and different states reported IMRs ranging from to around 21 per 100 000 children aged 1–4 and to 9 per 100 000
5.0 in New Hampshire to 11.3 in the District of Columbia.6 aged 5–14 compared to a risk of death in the first year of life of 520 per
Birth weight, reflecting both gestation and intrauterine growth, is the 100 000 live births. Death rates then begin to rise again after the age of
strongest predictor of the risk of death in the first year of life (Fig. 2.7). In 15, particularly in boys and largely as a result of the increasing risk of
large part, differences in mortality between, for instance, Sweden and the injuries.2
UK and between different ethnic groups in the USA are accounted for by As shown in Figure 2.2, the risk of death from all causes in child-
­differences in the birth weight distribution. The proportion of babies by hood has been falling throughout the last 50 years. Rates of death from
birth weight group in England and Wales in 2004 is shown in Figure 2.8. In unintentional injury in children have also fallen, but injury and poison-
recent decades there has been a small overall increase in mean birth weight ing still remain responsible for the greatest proportion of deaths in older
in many industrialized countries, largely accounted for by an increase in the children. Table 2.4 compares the most common attributable causes of
proportion of babies born weighing more than 3500 g. There has also, how- death in children aged 1–14 in 1955, 1985 and 2005, in England and
ever, been an increase in the proportion of babies born weighing less than Wales. Very similar patterns are seen in the USA6 although uninten-
2500 g, in the USA from 6.7% of live-born infants in 1984 to 8.1% in 2003. tional injury rates are higher than in England and Wales and homicide is
This partly reflects increasing numbers of twins and higher order births, a more important cause of death in children. In 2000 in the USA, homi-
which in the USA now comprise over 3% of all births.6 cide accounted for some 5.8% of deaths in children aged 1–14 compared
to 2.0% in England and Wales.1 Amongst adolescents aged 15–19 in the
USA in 2003, accidents were responsible for 49.7% of deaths, homicide
1000 for 13.9% and suicide for 10.9%.6 In both the UK and USA the risks of
non-intentional and intentional injuries are substantially higher in boys
IMR on logarithmic scale

194 than in girls.

100
Morbidity in childhood
22
Morbidity data are more difficult to obtain, especially in poorer coun-
8.8
10 tries, and to interpret than mortality data, which are generally easily
3.4 available and reasonably reliable. However such data are essential if a
1.8 picture of the health of the population can be built up.
1.1
1
<1500 1500- 2000- 2500- 3000- 4000-
1999 2499 2999 3999 4999 Health service use
Birthweight in g Children are heavy users of health services and routine service data
Fig. 2.7 Infant mortality rate (IMR) by birth weight (England and can provide useful information about patterns of morbidity. A number
Wales 2004).2 of different sources of health service data are available in the UK, some

Table 2.4 Percentage deaths age 1–14 by leading causes in order of frequency in 1955, 1985 and 2005 in England and Wales2

1955 1985 2005

Category % Category % Category %
Injury and poisoning 25.0 Injury and poisoning 33.6 Injury and poisoning 20.5
Respiratory disease* 13.0 Congenital anomalies 15.3 Malignant disease 20.0
Malignant disease 11.7 Malignant disease 14.0 Diseases of the CNS 13.9
Infectious disease 9.4 Diseases of the CNS 9.4 Congenital anomalies 11.3
Congenital anomalies 8.7 Respiratory disease 6.2 Respiratory disease* 8.1
Diseases of the GIT 4.1 Infectious diseases 4.0 CVS disease 6.6
 epidemiology of childhood diseases 15

derived directly from the use of services and some from special ­surveys Estimates of the prevalence of cerebral palsy, the most common
such as the annual ‘General Household Survey’, which includes self- form of serious physical disability in childhood, are available from a
reported health and use of services.7 Somewhat confusingly, these number of countries. Most recent studies in industrialized countries
often cover different components or combinations of components of report rates of cerebral palsy between 2 and 3 per 1000 infants surviv-
the UK. All these routine data have to be interpreted with some cau- ing the neonatal period. There is conflicting evidence about trends over
tion. Changes in admission rates over time for instance may reflect time. In the UK, one study reported a change between 1964–1968 and
changes in classification either by health professionals or those who 1989–1993 from 1.68 to 2.45/1000 neonatal survivors.12 A report
code the data, changes in thresholds for admission or real changes in combining data from five UK registers of cerebral palsy seems to suggest
incidence. that rates rose during the late 1970s and then flattened thereafter but is
Admission of children to hospital is relatively common. In the UK, in unclear about the extent to which this represents changes in ascertain-
2002, around 11% of 0–4-year-olds and 4% of 5–15-year-olds reported ment.13 The rate in babies born weighing > 2500 g has remained virtu-
being admitted to hospital at least once in the previous year.7 There has ally constant over this period while one large study has reported that
been a steady decline in both the proportion and the average length of in western Europe the rate in very low birth weight babies (< 1500 g)
stay in hospital over time. The two most important reasons for admis- fell from 60.6/1000 live births in 1980 to 39.5/1000 in 1996.14 There
sion in 5–14-year-olds are respiratory conditions, (including asthma), have also been increases in the lifespan of children with cerebral palsy,
and injuries, each accounting for about 16% of admissions. which will further increase the overall prevalence of cerebral palsy in
Children are also frequent visitors to Accident and Emergency the population.
departments and to general practitioners (GPs).7 More than 11% of The life expectancy of children with a number of other chronic con-
British children report going to an Accident and Emergency depart- ditions has also increased in the developed world, which is again likely
ment over a 3-month period. Those under 5 visit a GP on average seven to raise the prevalence of such conditions in the population. A survey
times per year and those between 5 and 15, three times per year. Data carried out in one health district in the UK attempted to ascertain what
collected in Scotland suggest that the commonest reasons for these GP proportion of children suffer from nonmalignant ‘life-threatening’ con-
visits are upper respiratory tract infections, otitis media, coughs, sore ditions.15 These were defined as conditions as a result of which the child
throats and other minor conditions. had at least a 50% likelihood of dying before the age of 40 and included
conditions such as cystic fibrosis, chronic renal failure and conditions
causing central nervous system degeneration. The overall prevalence
Chronic disease and disability among children aged 0–19 was 1.2/1000, suggesting a large burden on
Data on chronic disease and disability are relatively poor, even in coun- families and services.
tries with highly developed health systems. In the UK, special studies Mental health problems are extremely common in children and
generally have to be relied upon for this information. Unfortunately adolescents. A large population survey carried out in Great Britain in
there are often problems with the representativeness of the samples and 2004 of children aged 5–16 reported that 9.6% had a mental disorder
with the quality of the definitions used. (based on ICD 10 diagnostic criteria).16 The prevalence rises with age
The General Household Survey, mentioned earlier, asks a repre- and rates are generally higher in boys than girls except for emotional
sentative group of people in the UK to report on their own and their disorders (Table 2.6). Even apparently milder psychological problems
children’s health. One of the questions asks whether they have a long- such as behavioral difficulties which would not fulfill the criteria for
standing illness and whether it limits their activities. In 2002, the a mental disorder in young children can have profound effects on the
parents of 20% of boys and 17% girls aged 0–14 said that their child quality of families’ lives.
had a longstanding illness and 7% said that this limited their child’s
activities.7 Disabling conditions in poor countries
Reliable information on the prevalence of disability in childhood, The prevalence of disabling conditions in children is believed to be dis-
except for the prevalence of a few well-defined conditions such as Down proportionately high in poorer countries although the sources of high-
syndrome, is particularly difficult to find. Such information may also quality data are few.17 Mung’ala-Odera et al18 reported that 61 of 1000
be difficult to interpret, as children’s functional abilities form a con- 6–9-year-old children in one rural area in Kenya had moderate to severe
tinuum, and the point at which the child is labeled as ‘disabled’ is arbi- neurological impairment (Table 2.7). The authors point to the rela-
trary. A national survey in the UK, published in 1989, reported 3% of tively low prevalence of cerebral palsy in their population and suggest
0–15-year-olds were perceived by their parent as disabled.8 Clearly such that this is likely to be related to a high mortality rate in these children
reports depend on the definitions used, the way the data are collected in poor communities. The higher rates seen in girls than boys in this
and the population, leading to widely varying estimates. In 19949 Boyle study are surprising and unexplained. A study of intellectual disability
et al reported that 17% of US children had developmental disabilities in children of the same age group in rural South Africa (defined as IQ
while in 200610 Blanchard et al suggested that the figure was 5%. Using < 80, mild 56–80, severe < 55) reported the more commonly observed
multiple methods of ascertainment a study based in Atlanta (roughly excess in males who were 1.5 times as likely to be affected.19 The overall
50 000 births/year) has reported on the prevalence of four specific reported prevalence was 35.6 of 1000 (mild 29.1 of 1000, severe 6.4
developmental disabilities (Table 2.5).11 of 1000).
Poverty leads to poor nutrition, recurrent illness and inevitably defi-
cient care as families struggle to survive. In addition to the more obvious
Table 2.5 Prevalence of selected developmental disabilities in
Atlanta (2000)11 Table 2.6 Percentage of children with a mental disorder by age
and gender in Great Britain in 200416
Prevalence /1000 children
5–10 years 11–15 years
‘Mental retardation’* 12.0
Disorder Boys Girls Boys Girls
Mild 8.7
Moderate, severe or profound 3.3 Emotional disorder 2.2 2.5 3.9 6.0
Cerebral palsy 3.1 Conduct disorder 6.9 2.8 8.8 5.1
Hearing loss (> 40 dB) 1.2 Hyperkinetic disorder   2.7 0.4 2.6 0.3
Visual impairment (20/70 or worse corrected) 1.2 Autistic spectrum 1.9 0.1 1.0 0.5
disorder
* In Europe the term more commonly used is ‘learning disability’. Mental retardation Any disorder 10.2 5.1 13.1 10.2
was defined here as IQ < 70 (mild = IQ 50–70, moderate/profound = IQ < 50).
16 foundations of health and practice

Table 2.7 Estimated prevalence of moderate to severe neurological 450

impairment in 6–9-year-olds in rural Kenya per 1000 children18 400
Impairment Boys Girls Total 350

Death rates per 100000

Epilepsy 37 45 41 300
Cognitive impairment 27 36 31 250
Hearing impairment 12 15 14
Motor impairment 5 4 5 200
Visual impairment 2 2 2 150
Any impairment 52 79 61
100
50
0
<1 1−4 5−9 15−24 24−34
severe forms of disability this poverty of environment means that chil-
dren fail to reach their developmental potential. Grantham-Macgregor Males 58 37 35 341 424
et al estimate that more than 200 million children under the age of 5 fail Females 67 31 16 99 101
to reach their potential in cognitive development as a result of poverty.20 Fig. 2.10 Death rates form injury and poisoning per 100 000 by age and
This has serious economic and social consequences for them as individ- gender in England and Wales (2004).2
uals and for the societies in which they live.
by ­environmental and behavioral differences. Clearly there are differences
Population determinants between populations in the frequency of some specific single gene disor-
of child health ders – for instance, sickle cell disease is uncommon in northern Europeans
and cystic fibrosis is uncommon in Africans – but these contribute rela-
The health of an individual is determined by a complex interaction tively little to the health experience on a population level. In other words,
between genetic factors, health behaviors and environmental influences. an individual’s genetic makeup is important in determining their risk of ill
health within a particular subpopulation who share common experiences,
but not in explaining the differences between these subpopulations. This is
Age and sex structure of the population
borne out by numerous studies of immigrant populations, which suggest
Between the ages of 5 and 15 the risk of death and severe illness is at that the longer they spend in a new country and the more they adopt the
its lowest, before rising again. Death rates are higher in boys than girls lifestyle of that country, the more nearly their health experience comes
at all ages in both poor and rich countries (Fig. 2.9 shows the rates by to resemble that of the native population. Most differences that ­continue
age and gender for England and Wales). The magnitude of the difference to exist between immigrant groups and the native ­ population can be
varies with age, in the UK rising from around 25% higher in boys from explained on the basis of either behavioral or socioeconomic factors.
birth until the age of 15 and then becoming more than twice as great in
adolescence and early adulthood. Much of this increasing discrepancy
between the sexes is due to the much higher rates of unintentional inju- Birth weight and gestation
ries in males, particularly later in childhood and adolescence. The death Birth weight, reflecting both gestation and intrauterine growth, is a pow-
rates from all injuries and poisoning (which includes the small num- erful predictor of mortality (Fig. 2.7) and morbidity (e.g. cerebral palsy,
bers of deaths due to intentional injuries) in the UK are shown in Figure Fig. 2.11)21 in childhood. Globally 60–80% of neonatal deaths occur in
2.10, broken down by age and gender. low birth weight infants. In recent years it has been recognized that the
effects of suboptimal birth weight may persist throughout life, with links
being demonstrated between birth weight and cardiovascular and respi-
Genetic factors ratory disease in adulthood. It appears that these adverse effects have a
At an individual level there is no doubt that genetic factors play an nearly linear relationship with birth weight rather than simply being
important role in determining health status. At a population level how- associated with the lower extreme of the birth weight distribution.
ever, there is little evidence that they have a significant effect on overall Many of the determinants of both prematurity and poor intrauterine
health. The genetic differences between human subpopulations are in fact growth remain to be elucidated, although some specific conditions such
small and it appears that differences in health are largely accounted for as pre-eclamptic toxemia are important in individuals. Poor ­intrauterine

6
80
Prevalence per 1000 neonatal survivors

72.6
5
Death rate per 1000

70
4 60

3 50

2 40
30
1
20
0 11.1
<1 1−4 5−9 10−14 15−19 20−24 10
Males 5.7 2.3 1 1.4 4.2 6.4 1.2
0
Females 4.5 1.9 0.9 1.1 2.2 2.7 <1500 1500–2499 2500+
Birth weight in grams
Fig. 2.9 Death rates per 1000 by age and gender in England and Wales
(2005).2 Fig. 2.11 Prevalence of cerebral palsy by birth weight in Europe.21
 epidemiology of childhood diseases 17

growth is associated with smoking during pregnancy and with socio- the proportion of infants born weighing less than 2500 g in SES deciles
economic deprivation. Diet has often been suggested as a possible cause in one region of the UK (based on the characteristics of the small area in
of intrauterine growth retardation but good evidence for this is lacking, which they live).22 Perhaps even more striking is the proportion of infants
except for the effects of extreme malnutrition. born weighing more than 3500 g (regarded as being an optimum birth
weight) in different SES deciles (Fig. 2.13). In both figures there is a clear
gradient across the different social groups. The magnitude of this effect
Physical environment is illustrated by the fact that this study reported that, if the whole popu-
While some specific environmental hazards have been clearly identified, lation had the risk seen in the richest 10%, this would avoid 30% of all
the links between others such as damp, overcrowded housing or atmo- births below 2500 g and 32% of births below 1500 g.
spheric pollution and ill health have been difficult to determine. Many of In the UK in 1994–1999 the difference in life expectancy at birth
these, possibly disadvantageous, conditions tend to occur in combina- between children in the 10% of most deprived areas in the UK and those
tion with each other and with socioeconomic disadvantage, making it in the richest 10% was 6 years for boys and 3.2 years for girls.2
difficult to disentangle their effects. Many pollutants in the environment The differences in infant mortality across social classes are shown in
or in food are so widespread that investigation is difficult. It is also possi- Figure 2.14. Particularly marked differences are seen in mortality from
ble that it is the interactions between such pollutants that are important, injuries and poisoning between social groups (Fig. 2.15).23
which further complicates attempts to quantify their effects. It could be Similar differences are seen for most measures of morbidity, where
argued that, in these circumstances, it may be appropriate to accept figures are available, with large differences being shown between social
lower levels of evidence before proceeding to action, sometimes referred groups for the risk of admission to hospital, for severe respiratory infec-
to as the precautionary principle, than would generally be required for tions and for mental disorders (Table 2.8).16 Finally, SES in early child-
reaching conclusions about causal links. hood strongly predicts the likelihood of educational achievement, which
in turn predicts later job opportunities and income.
SES can of course not be said to directly cause ill health, but rather
Social factors acts as a marker for various adverse circumstances and behaviors, which
Socioeconomic status (SES) is a powerful predictor of health outcome are the proximal causes for health outcomes. There is growing evidence
within all societies. Within poor countries, this is unsurprising, given that these circumstances cumulate over the life course and that the lon-
that SES is linked to the availability of basic necessities including food ger a child spends in adverse social circumstances, the greater the risk
and shelter. What is perhaps remarkable is that the link remains strong of poor health outcome.
even in rich industrialized countries. What may seem equally surprising
is that the differences exist, not simply between the poorest members of
society and the rest of the population, but, for many important health Health behaviors
outcomes, there is something approaching a linear relationship between There is considerable evidence of the deleterious effects of a number
SES and adverse outcome. of health behaviors by parents and children on the health of children.
How best to measure SES in childhood remains a source of debate. It Smoking by parents, and by children, in particular, is a major cause of
is clearly a complicated concept and it seems likely that different aspects many adverse outcomes including intrauterine growth retardation,
of disadvantage will be important for different health outcomes. In the sudden infant death syndrome and respiratory disease. At least in the
UK, SES has traditionally been measured using an occupational classifi- UK and USA, smoking is much more common amongst people in poorer
cation. This scheme, first employed around the beginning of the twenti- social circumstances and is likely to be one of the mechanisms through
eth century, assigned all occupations to six (originally five) groups based which SES has its effects on children’s health. This close relationship
on a notion of a hierarchy of status. In 2001 this system was replaced between SES and smoking does, however, hamper efforts to estimate
by a new eight-group classification, the National Statistics Socio-eco- the magnitude of the effect of smoking per se as its effects may be con-
nomic Classification. Other classifications have been developed based on founded by other adverse circumstances.
factors such as maternal education, income, access to material goods Harmful effects of poor diet and lack of exercise in childhood have
such as motor cars or telephones, type of housing and the nature of the also been suggested. Very large increases in the proportion of children
area in which the family lives. Although the strength of the association who are overweight or obese have been reported in the UK and USA.
between SES and a particular outcome may vary according to the mea- Between 1994 and 2004 in the UK the proportion of 11–15-year-old
sure used, the direction of effect is virtually always the same. boys who were obese rose from 14% to 24% and of 11–15-year-old girls
The effects of SES are observable from the beginnings of life, with a from 15% to 26%.24 As there is substantial tracking between fatness
strong relationship between birth weight and SES. Figure 2.12 shows in childhood and adulthood, this may have important implications for

12 50
45 %>3500g
10 %<2500g 40
35
8
Percentage
Percentage

30
6 25
20
4 15
10
2
5
0 0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
SES decile (10th is poorest) SES decile (10th is poorest)
Fig. 2.12 Percentage of births < 2500 g by SES decile in the West Fig. 2.13 Percentage of births > 3500 g by SES decile in the West
Midlands, UK.22 Midlands, UK.22
18 foundations of health and practice

10
9
8

Death rates per 1000

7
6
5
4
3
2
1
0
1.1 1.2 2.0 3.0 4.0 5.0 6.0 7.0 Unclass
Neonatal mortality 2.0 2.7 2.7 3.5 2.7 2.8 4.6 4.3 5.5
Infant mortality 2.8 3.7 3.5 5.1 4.0 3.8 6.5 6.0 8.8
Social class
Fig. 2.14 Neonatal and infant mortality rates by parental occupational social class for jointly registered births in England and Wales (2004).2

30 countries. The obvious exceptions are immunization and public health

measures such as the provision of safe water supplies. Even for vaccine-
25 preventable diseases, much of the decline in mortality in these coun-
Deaths per 100 000

tries preceded the introduction of immunization. It is nonetheless true

20 that immunization has been associated with dramatic declines in deaths
15
from measles, polio, meningococcal disease and other conditions which
were major causes of deaths into recent times.
10 It is clear that the major determinants of child health lie outside the
realm of curative services and are related to social and environmen-
5 tal factors. However, in rich societies, the rates of mortality have fallen
to very low levels and the importance of the effective management of
0 relatively uncommon conditions has become proportionately more
1 2 3 4 5 6 7 8
SES group important in determining mortality rates. For instance, the widespread
use of antenatal steroids in women in preterm labor and of surfactant
Fig. 2.15 Deaths in children aged 0–15 from injury and poisoning per in premature infants has led to substantial declines in mortality from
100 000 by occupational social class, England and Wales (2001–2003).23 idiopathic respiratory distress syndrome in neonates. Similarly, as malig-
nant disease accounts for an increasing proportion of childhood deaths,
improvements in cure rates due to medical management can have
Table 2.8 Percentage of children aged 5–16 with a mental disorder by significant effects on childhood mortality rates.
gender and weekly household income in Great Britain, 200416 As discussed earlier, simple interventions by health services have
Household income Boys Girls the potential to significantly reduce deaths and morbidity in the poor-
est communities in the world. The difficulty has been in trying to intro-
< £100 18.4 13.4 duce such services in the face of a shortage of resources and often weak
£100–199 13.9 13.0 governmental structures which are not always responsive to the needs
£200–299 17.8 11.5 of the poorest members of society. The ‘inverse care law’25 was first
£300–399 14.8 9.8 described in the UK in the 1970s and suggested that ‘The availability
£400–499 10.2 6.8 of good medical care tends to vary inversely with the needs of the pop-
£500–599 10.6 7.1 ulation served.’ Unfortunately this law still applies between countries
£600–770 6.0 3.7
and between population groups, within even the poorest. For instance,
>£770 6.7 3.9
in sub-Saharan Africa and Southeast Asia, amongst the 20% of richest
women 86% have skilled care at the birth of their children compared to
the risk of cardiovascular disease and type 2 diabetes in later life. It has only 14% amongst the poorest 20%.4
been suggested that these increases are likely to reflect declining levels of While health services can significantly ameliorate the consequences
physical activity and the increasing consumption of convenience foods, of many diseases they can also lead to an increased prevalence of chil-
although clear etiological evidence is lacking. dren with significant morbidity. Some of the children who are saved
from death may survive with significant morbidity and may be kept alive
for long periods in spite of their disabilities. It is important that clini-
Health services cians recognize both the importance of factors outside their control in
Until recent years there has been surprisingly little evidence that health determining children’s health and the potential societal consequences
services are an important determinant of children’s health in richer of advances in technology.
 epidemiology of childhood diseases 19

References 11. Bhasin TK, Brocksen S, Avchem RN, et al. Kenya. International Journal of Epidemiology 2006;
Prevalence of four developmental disabilities among 35:683–688.
1. World Health Organization. http://www.who. children aged 8 years – Metropolitan Atlanta 19. Christianson AL, Zwane ME, Manga P, et al.
int/en/. Developmental Disabilities Surveillance Program, 1996 Children with intellectual disability in rural South
2. Office for National Statistics. http://www.statistics. and 2000. MMWR Surveillance Summaries 2006; Africa: prevalence and associated disability. J Intel
gov.uk/default.asp. 55:1–9. Disabil Res 2002; 46:179–186.
3. Bryce J, Boschi-Pinto C, Shibuya K, et al., WHO 12. Colver A, Gibson M, Hey EN, et al. Increasing 20. Grantham-McGregor S, Cheung YB, Cueto S,
Child Health Epidemiology Reference Group. WHO rates of cerebral palsy across the severity spectrum et al. International Child Development Steering
estimates of the causes of deaths in children. Lancet in north-east England 1964–1993. Arch Dis Child Group. Developmental potential in the first 5 years for
2005; 365:1147–1152. 2000; 83:F7–F12. children in developing countries. The Lancet 2007;
4. Lawn JE, Cousens S, Zupan J, the Lancet Neonatal 13. Surman G, Bonellie S, Chalmers J, et al. UKCP: 369:60–70.
Survival Steering Team. 4 million neonatal deaths: a collaborative network of cerebral palsy registers in 21. Surveillance of Cerebral Palsy in Europe (SCPE).
When? Where? Why? Lancet 2005; 365:891–900. the United Kingdom. Journal of Public Health 2006; Prevalence and characteristics of children with
5. Jones G, Steketee RW, Black RE, et al. Bellagio 28:148–156. cerebral palsy in Europe. Develop Med Child Neurol
Child Survival Study Group. How many child 14. Platt MJ, Cans C, Johnson A, et al. Trends in 2002; 44:633–640.
deaths can we prevent this year? The Lancet 2003; cerebral palsy among infants of very low birthweight 22. Bambang S, Spencer NJ, Gill L, et al.
362:65–71. (< 1500 g) or born prematurely (< 32 weeks) in 16 Socioeconomic status and birthweight: comparison
6. Hoyert DL, Mathews TJ, Menacker F, et al. Annual European centres: a database study. The Lancet 2007; of an area-based measure with the Registrar General’s
Summary of Vital Statistics: 2004. Pediatrics 2006; 369:43–50. social class. J Epidemiol Community Health 1999;
117:168–183. 15. Lenton S, Stallard P, Lewis M, et al. Prevalence 53:495–498.
7. Office for National Statistics. Living in Britain: and morbidity associated with non-malignant, life- 23. Edwards P, Green J, Roberts I, et al. Deaths from
Results from the 2002 General Household Survey. threatening conditions in childhood. Child Care injury in children and employment status in family:
http://www.statistics.gov.uk/lib2002/default.asp. Health Dev 2001; 27:389–398. analysis of trends in class specific death rates. BMJ
8. Bone M, Meltzer H. The prevalence of disability 16. Green H, McGinnity A, Meltzer H, et al. Mental 2006; 333:119.
among children. OPCS Surveys of Disability in Great health of children and young people in Great Britain, 24. The Health Survey for England 2004. http://www.
Britain, report 3. London: HMSO; 1989. 2004. Office for National Statistics http://www.statistics. dh.gov.uk/en/Publicationsandstatistics/PublishedSurvey/
9. Boyle C, Decouflé P, Yeargin-Allsopp M. Prevalence gov.uk/downloads/theme_health/GB2004.pdf HealthSurveyForEngland/index.htm
and health impact of developmental disorders in 17. Durkin M. The epidemiology of developmental 25. Hart JT. The inverse care law. Lancet 1971;
children. Pediatrics 1994; 93:399–403. disabilities in low-income countries. Mental Retarda­ 1:405–412.
10. Blanchard LT, Gurka MJ, Blackman JA. tion and Developmental Disabilities Research Reviews
Emotional, Developmental and Behavioral Health 2002; 8:206–211.
of American Children and Their Families: A Report 18. Mung’ala-Odera V, Meehan R, Njuguna P,
From the 2003 National Survey of Children’s Health. et al. Prevalence and risk factors of neurological
Pediatrics 2006; 117:e1202–e1212. disability and impairment in children living in rural
 10001

3
Health care delivery in resource
limited settings
Elizabeth Molyneux

Background 21 Information, surveillance and research 23

The philosophy of care 21 High-impact health interventions 23
Human resources 22 Syndromic management of diseases 23
What is expected of the child health doctor? 22 Expanded program for immunization (EPI) 24
Health care financing 22 HIV/AIDS – prevention, diagnosis prophylaxis
Drug supply – the pharmaceutical pipeline 22 and treatment 24
National health programs 22 Palliative care 24
Health delivery and the referral chain 23 TB treatment programs 24
The health center 23 Malnutrition 24
The district hospital 23 Training 24
The tertiary hospital 23 Dilemmas and constraints 24
Vertical, parallel or horizontal services? 23 Summary 25

Background Funding for primary care has been inconsistent in part because it is
not easy to establish clear indicators of achievements and failures and
In 1978 in Almaty Kazakstan (then Alma-Alta) a meeting on primary without these, financial systems do not function.8 Primary health care
health care was jointly sponsored by the United Nations Children’s Fund is often seen as an emergency response to health disasters by interna-
and the World Health Organization (WHO) and attended by representa- tional donors rather than the bedrock of health care delivery. But in
tives from many countries.1 A call was made for ‘health for all’. Nearly the second Disease Control Priorities in Developing Countries report,
30 years on ‘health for all’ is still more hope than substance.2 In 1990 it is argued that 90% of the health care demand is potentially address-
The Millennium Development Goals were defined.3 Amongst the goals able by primary care.9 Historically most health services have been
are to improve the health of both mothers and children; to reduce under built and powered from the top down. Centrally made decisions have
5s child mortality rate by two thirds, and maternal mortality by three trickled down to primary services in the community. Urban health in
quarters by 2015. At current rates these goals will not be achieved until resource poor countries is nearer to the health model taught in most
2165 in sub-Saharan Africa.4 In 2003 the Bellagio child survival group traditional medical schools, with emphasis placed on specialization
(named after the place that the meeting was held) report5 was published and hospital care.
in which representatives of the donor communities and health services The large hospitals were built, managed and staffed like those in
of resource constrained areas described children’s health in various developed countries. This tended to lead to excellent care for the politi-
parts of the world and suggested that with the present state of knowl- cally powerful urban few, at the neglect of the rural poor. Medical stu-
edge and relatively little financial investment, 60% of the deaths in chil- dents were, and are given bedside teaching which emphasizes diagnostic
dren under 5 years of age could be averted. In research too, the World acumen and knowledge of diseases, but teaches little about health care
Health Forum describes a 90/10 divide in which only 10% of research delivery – for instance why did this particular child get admitted, by
funding goes to poorer parts of the world where 90% of the people live.6 whom, when and why? Why were others not admitted? In the 1970s the
Why is the health divide so wide, and increasing, between well pendulum of donor funding tipped towards primary health care and it
resourced countries and those which are financially constrained? The gradually improved. The use of traditional birth attendants and village
problems are complex. community heath workers was encouraged. Acceptance has not been
universal, with some reports arguing that skilled care at delivery made
The Philosophy of Care the difference in maternal survival with the result that the unskilled
cadre of health workers fell into some disrepute.10,11 More recently there
In the 1970s primary health was promoted as the way forward to reach are many examples of the value of the CHW. They have helped improve
the poorest and most difficult to access (often rural) population. This community uptake of health services, decreasing neonatal mortality in
service was to deploy community health workers (CHWs), traditional India, and in many countries successfully run immunization services
birth attendants (TBAs) and mid-level health staff such as nurses and and health promotion schemes.12–14 They work best when included as
clinical officers. However primary care has been understood variously, part of the health service and not made to feel outsiders, or second best.15
as the central pivot for health systems, as a mere extension of ser- A health service needs all levels of service to be effective. It is recognized
vices to underserved areas, or even as a rather second class service for that the larger hospitals have been neglected, and as these are teaching
the poor.7 The first definition requires social, structural and financial centers and service the community, there is an important swing back to
investment, the other definitions diminish the value of primary care. seeing their improvement as equally important. 21
22 FOUNDATIONs OF HEALTH AND PRACTICE

Human Resources the Under Fives Clinics a nurse is being asked to weigh and measure
the height of a child, assess development, encourage exclusive breast-
Health services are directed by policy and finance but are driven and feeding for 6 months, give immunizations, give vitamin A, nutri-
sustained by people. In the 2006 World Health Report, the WHO high- tional advice, malaria prophylaxis, co-trimoxazole prophylaxis for
lights the crisis in human resources for health.16 The central importance HIV-infected infants and children. Many of these duties can be done,
of human resources to health means that this issue deserves high inter- and are done by less skilled but dedicated and trained staff. The World
national priority.17 To achieve the Millennium Development Goals – Health Report of 2006 urges health services and health training insti-
which are wider than child and maternal health – Africa alone will need tutions to re-look at task allocations within both the professional and
a further 1 million health personnel, and the global need is an extra 4 voluntary cadres.16
million health workers.18 The ratio of health carers (nurses, doctors,
midwives) to population is 10 times greater in Europe than in Africa, Health Care Financing
and yet maternal, neonatal and infant mortality are directly related to
the skilled staff:patient ratio.19 As the cost of and demand for health care has risen various ways of
The need for skilled health carers is at crisis point in much of subSa­ cost sharing have been tried. The Bamako Initiative of 1987 was an
haran Africa and efforts are being redoubled to train as well as to deploy attempt to improve primary care and user fees were suggested as a key
CHWs where possible. In Asia CHWs have worked with great effect. component to make improvements sustainable.25 By raising revenue it
For instance, in the 2005 earthquakes of Pakistan 8000 female health was hoped that the quality of care, coverage and drug availability would
workers were mobilized from the affected areas to assist in the camps improve without the need for persistent external inputs. User fees have
and villages. In India there are plans to train 300 000 more CHWs been successful in some countries (Benin, Senegal, Burkino Faso) but
while in Thailand 60 000 village health workers were trained to sup- not in others and in some places have been withdrawn as they were
port 600 000 voluntary health workers; who in turn looked after 20 seen as a barrier to the poorest people accessing care.26 They seem to
children each.20 It has been suggested that a ratio of one full-time paid help when they are part of a care package that includes insurance and
CHW to 500 people and one to 10–20 for local volunteers is the correct good exemption schemes. Some countries (e.g. Sri Lanka) have contin-
requirement.20 ued successfully to provide free health at point of care while others such
In poorly resourced settings we tend to think of resources as mate- as Bolivia intend to provide a similar service backed up by insurance.27
rial needs – equipment, drugs, good buildings, transport, communi- The World Health Report 2005 ‘Making every mother and child count’28
cations – all of which are necessary and important and which may be rejects user fees and argues that social insurance or tax-based responses
scarce or in poor shape. The health care team is made up of a group are superior.
of people with varying backgrounds, training, skills and expecta-
tions. Each person in the team has a much bigger task in terms of Drug Supply – the Pharmaceutical
quantity and breadth than their counterparts in a well-resourced Pipeline
system. This team may consist of a clinical officer (akin to a nurse
practitioner), nurses/midwives, health care auxiliaries and health In 1975–1997, of 1325 new drugs which came on to the market only
surveillance assistants. If there is a doctor he/she will be expected 11 were for tropical infections.29 WHO has compiled an essential drugs
to lead this team. list (329 drugs and 559 formulations) to help focus drug provision on
basic needs.30 Many governments have central medical stores through
which bulk medical buys are made but worldwide 30% of people lack
What is Expected of the Child access to essential drugs. This varies from 26% in most of SE Asia, to
Health Doctor? 47% in Sub-Saharan Africa (SSA) and 65% in India.31 UNICEF provides
In such circumstances the doctor must be the leader, a teacher, a super- 40% of the global demand for vaccinations and the Global Alliance for
visor and an advocate. He/she needs to think creatively about how to Vaccines and Immunisation (GAVI) and the Global Fund have been very
deploy staff, to prioritize both their time and that of others, and to stim- effective and active in providing a wider range of immunizations for
ulate, enthuse and lead ‘from the front’. In child health, whatever their national EPI programs as well as treatments for TB, malaria and HIV,
own clinical interests and preferences they need to see the wider context respectively.31–33
of the health service and ensure that the team’s efforts reflect the needs Good will donations are sent to many hospitals without any
and demands of the community they serve. clear understanding of the pharmaceutical needs or controls of a
Many doctors trained in poorly resourced countries have stayed to country. Lavy et al 34 gave advice to help potential donors and the
serve their communities for little financial or career rewards. For very WHO has come up with standards to which donated drugs must
many others the offers of specialization, security and financial gain conform.35 In some villages drug revolving schemes have been
have attracted them to resource rich countries.21–23 There is much to be established whereby a responsible villager is chosen by the commu-
gained from medical exchanges, extra specialist training, sabbaticals nity to be custodian of a few simple medicines which he/she dis-
and cross-fertilization of ideas but if these encourage a drain from set- penses at cost and thus recovers the monies to buy replacement
tings of need to places of plenty; then the loss to poor communities is medicines.36,37
very great indeed.24
Government services struggle to implement health care with too National Health Programs
few trained staff. Skill mix has become increasingly important; nurses
and clinical officers have taken on many of the jobs that traditionally Programs require substantial funding which is heavily dependent on
have been carried out by doctors, nurse auxiliaries work under the donor provision.38 Donors have, over the years, channeled funds in a
supervision of trained nurses and Health Surveillance Assistants play variety of different ways – vertical programs, parallel programs, single
key roles in health promotion, disease prevention and monitoring in disease-focused initiatives, funds targeted to one sector of the popula-
the community. Voluntary or paid community health workers, elected tion such as reproductive health or primary care. National health ser-
by the village community assist with bed net provision, simple treat- vices have been molded to meet the requirements of individual donor
ments and health promotion. Nurses have had many tasks put upon states or programs. More recently a sector-wide approach (SWAps)39 has
them in the antenatal clinics beyond the monitoring of a woman’s been introduced in which donors ‘basket’ (put together) funds and give
pregnancy (e.g. human immunodeficiency virus [HIV] prevention, them to a Ministry of Health to deliver a predetermined health program
malaria prophylaxis, nutrition teaching, advice about breast feeding, based on an assessment of needs called the Essential National Package
tetanus immunization, STI treatment and prevention). Similarly in of Health Services.40
 HEALTH CARE DELIVERY IN RESOURCE LIMITED settings 23

Health Delivery and the Referral ing was obtained through the SWAps financing system. District health
Chain management teams set their own priorities and integrated health
informatics into the plan. This experience suggests that investing in the
Health services are divided into different levels of care delivery. It is essen- health system can work but that it takes time. In addition to a burden of
tial to provide care as near to the home as possible, but that care must be disease analysis consideration of cost effectiveness, budgeting, manage-
supported by a tiered system that provides support for the tier below and ment, team building, mapping, transport, and communications were all
receives support and refers to the tier above. There must be a seamless necessary. Even then the effect was only seen after 5 years.49,50
continuum of care within which there is free and good communication
at and between all tiers and appropriate use of each level of care. Information, Surveillance and
Research
The Health Center Comprehensive health management information is important to mea-
At village level, health centers or mobile health clinics provide simple sure quality and coverage of health services. Poor data often leads to
first-line curative treatment, immunizations and health promotion and poor decision making and ultimately poor quality of care.51 In many
do uncomplicated deliveries. These are ideally run by a nurse, midwife countries data is collected in too many different and uncoordinated ways
and medical assistant and Community Health Workers supervised and and there is lack of access to information and to information research.52–54
supported by staff from the local district hospital. These interventions The importance of data and the need to rapidly cycle results back to
collectively reduce many risk factors for ill health. the health care units so as to influence day-to-day decision making has
become clear.55
Detailed country-specific epidemiological data will be available for
The District Hospital each country from WHO so that a national health service can tailor its
The district hospital provides first level referral care and is often efforts, prioritize its needs and set goals for itself.16
staffed by clinical officers and nurses, with or without a doctor. Such
a hospital will typically have about 200 beds and serve a wide catch- High-Impact Health Interventions
ment area. This service can be provided relatively cheaply and is
reported to cost about $9/d in Kenya, $12 in Bangladesh.41 District Some relatively simple interventions influence the health of many. Examples
hospitals will refer their difficult cases to the regional or tertiary are immunization programs, vitamin A supplementation, zinc supplemen-
referral hospital. tation, infant care, simple treatments, bednet provision, improving care at
This level of care is termed the first-level referral and plays a signifi- first level referral hospitals and the integration of health care through the
cant role in the quality of service a population can expect to receive. Integrated Management of Childhood Illnesses (IMCI).56–58
English et al reported on the indifferent quality of care first-level referral Improving inpatient care in referral hospitals and especially emer-
hospitals give in Kenya when staff members are demoralized and unsup- gency care can make an immediate difference to inpatient morbidity
ported and equipment and structures are neglected.42 and mortality. In Blantyre, improving triage and emergency care by
teaching and implementing the WHO Emergency Triage Assessment
and Treatment (ETAT) has led to a reduction in inpatient mortality from
The Tertiary Hospital 11–18% to 6–8%.59,60 Prophylactic co-trimoxazole for HIV-infected chil-
The tertiary hospital or regional hospital will have staff in at least the dren, anti-retroviral therapy for children (and their parents) who have
main specialties of medicine, pediatrics, surgery, obstetrics and gynecol- reached a clinical- or laboratory-based stage of the disease needing
ogy. They are in large towns or cities and may serve as the city district therapy, TB diagnosis and treatments gain from vertically managed pro-
hospital. They are also the teaching units for different medical and nurs- grams which are not mutually exclusive and are imbedded in the health
ing cadres. Staff members of these large hospitals visit district hospitals care system.
to teach and help with problem cases. Many of these hospitals were built Clear consideration of evidence is important as not all apparently
in the 1950s or 1960s and, when policy focused on primary care, were sensible interventions have proved as successful. For instance the blanket
neglected. It is now recognized that these hospitals are where future provision of iron and folic acid to preschool children in a highly malarial
health staff are trained and are a vital link in the continuum of care for area in East Africa proved detrimental to health and outcome.61
patients.16 The public’s views of the quality of a national health service
is colored by the care seen to be provided in its tertiary institutions.
Syndromic Management of Diseases
Vertical, Parallel or Horizontal Where patients are many and staff are few, supported by little or no
Services? laboratory services, a syndromic approach to the diagnosis and man-
agement of particular types of infection has been taken. A diagnosis is
Vertical programs may offer an effective response to address specific made on a brief history and the presence of a few key symptoms or signs.
issues but there is a danger that programs may conflict with each other For instance in sexually transmitted diseases an appropriate history and
or may undermine the provision of primary health care. Maternal and findings elicits a treatment which covers all the common treatable bac-
child health (MCH) are interdependent but in some programs (Safe terial causes. This leads to overtreatment but simplifies the manage-
Motherhood for instance) the baby got little mention, and in some child ment of large numbers of patients by relatively few staff. Clinicians are
initiatives the importance of the mother’s health is unemphasized.43 Of taught this approach and national programs rationalize the availability
the 10 million children who die every year, the vast majority in resource of appropriate drugs for the level of health service.
constrained areas, 4 million are neonatal deaths.44 Half a million moth- IMCI is an integrated approach to the care of children which uses
ers die every year in childbirth, which imperils infant survival and leaves flow diagrams and simple pattern recognition in signs and symptoms to
many orphans, whose health and well being will inevitably suffer.45 treat children.62 IMCI strategy has three main components:
There are over 12 million orphans in Africa.46,47 1. to improve case management skills of health carers using locally
In Tanzania to avoid vertical programs and strengthen subnational adapted clinical guidelines that do not focus on a single diagnosis
health systems the Tanzanian Essential Health Interventions Project but rather on selected signs and symptoms which guide rational
(TEHIP) was started in 1997.48 After 5 years the mix, quality, use and treatment
coverage of health services improved and there was a 40% reduction 2. to provide support through the provision of appropriate therapies
in under-fives mortality. This initiative cost $1 per person and the fund- and referral chains
24 FOUNDATIONs OF HEALTH AND PRACTICE

3. to see the child in the context of the family and the clinical context cases with the onset of the HIV/AIDS epidemic. National control of the
with the acute event used to provide not only immediate curative TB treatments also reduces the risk of poor prescribing leading to the
treatment but also attention to nutrition, immunization, counseling emergence of resistant TB. The directly observed therapy (DOTs) cam-
and other heath needs. paign helps reduce poor compliance.74
In Tanzania this is reported to have led to 13% reduction in cost of care Both TB and HIV/AIDS are well served by vertical national programs.
and better outcomes.63 A recent re-analysis of this data to include the
quality of care given showed that the cost of using IMCI was $4.02/per-
son per annum compared with $25.70 without IMCI.64 Malnutrition
Poor nutrition directly affects 5.6 million deaths every year in children
under 5 years old. In the developing world 25% of all children (about
Expanded Program for Immunization (EPI) 146 million children) are underweight, which remains unchanged since
Immunization remains one of the most cost-effective approaches to 1999. Half these children are in Bangladesh, India and Pakistan, where
health care but the coverage varies greatly from country to country and 45% of the children under 5 years of age are malnourished. It is particu-
even between rural and urban areas. The least accessible children who larly evident in places of rapid population growth, in areas of conflict,
are the most vulnerable in society receive least protection. or with high HIV prevalence, drought or poor agricultural productivity.
EPI programs include periodically giving vitamin A supplements to Diarrhea and poor feeding habits are confounding and causative factors
children. The included vaccines vary. While some countries (e.g. Malawi) and it is of concern that children who do not catch up with growth in
have achieved high rates of uptake and include pentavalent vaccines the first 2 years of life may have continuing cognitive problems. Regular
(DPTHibHepB – diphtheria, pertussis, tetanus, Haemophilus influenzae, vitamin A supplementation and micronutrients help prevent some of
hepatitis B) most countries of subSaharan Africa are still using DPT. the diarrheal diseases that lead to malnutrition.75
Pentavalent vaccine raises the cost of vaccination tenfold and can only Malnourished children need either supplementary feeding or thera-
be sustained with external funding such as from GAVI.65 peutic feeding. This is done in a phased manner so that calorie intake
and protein load are increased over a few days and high loads introduced
once diarrhea and edema have started to resolve and appetite increase.
HIV/AIDS – Prevention, Diagnosis Prophylaxis This is especially important in severe kwashiorkor. Supplementary feed-
and Treatment ing is for moderately malnourished children and is provided as a nutri-
Many high-burden countries with HIV/AIDS are struggling to provide tious porridge flour or a ready-to-use therapeutic feed (RUTF) such as
for the needs of people who are living with AIDS with high proportions plumpynut – a peanut-based high-energy food.62,76–78 It is vital to support
of the population needing care. Fixed combination drugs (FDCs) are nutrition in all areas of health care as malnutrition has a detrimental
now available and provided through the global fund, but every affected effect on every aspect of health outcomes.
country needs a national program with registration procedures, proto-
cols, training, stock control, prescribing controls and monitoring. The Training
human resources required and logistical implications are huge. Children
are expected to form about 10% of the total number of people on anti- Training is required at many levels. University-trained doctors
retroviral therapy (ART) and their access to therapy has been slower and nurses are needed for leadership roles and in the referral hos-
than that of adults. In Malawi they are 5% of the total number on ART.66 pitals but they tend to stay in urban areas. Their training is costly
Reasons for this have been difficulty in deciding how to provide suitable and once qualified their expectations of resources and rewards are
doses for children with a limited variety of FDC tablets available and high. They are not enough to provide for a country’s health needs.
none of them child friendly tablets. Other difficulties include the inability Other cadres of staff must be trained for different tiers of the health
to confirm a diagnosis below the age of 15 months if only HIV antibody service. Medical assistants, clinical officers, patient attendants and
testing is available and complex issues surrounding advice about infant community workers are trained to provide many of these services.
feeding. What advice is best for babies whose chances of acquiring HIV For instance, in Malawi clinical officers receive 4 years of training,
infection with breast-feeding are about 14–42%,67 but whose chances of which includes 1 year’s internship making them able to deal with
dying of gastroenteritis in a poor family if bottle fed are much greater?68 common illness and emergencies.
A Zambian study of co-trimoxazole (CoT) prophylaxis for children with
HIV showed a decrease in mortality.69 This has meant that CoT is recom- Dilemmas and Constraints
mended to all children (and adults) with HIV infection and HIV-exposed
infants. The logistics for this provision are challenging. The means with which to meet large unresolved problems are limited.
There is a dilemma in trying to provide good-quality service – service
which includes kindness, courtesy and explanations – in quantity and for
Palliative Care little cost. Rural primary health care workers in Zambia, 40% of whom
Holistic care of a child and family is incomplete without palliative care, worked without a doctor, identified transport, knowledge, traditional
a stark reality with the HIV/AIDS epidemic. There is a steady increase beliefs, social stigma, poorly trained health staff, poor access to drugs
in awareness of its value and national palliative care programs have and user fees as barriers to good health service provision. Lack of human
started, training courses are available and oral morphine is available resources is a major problem and requires lateral and creative thinking in
in some countries.70–72 Hospital-based and home-based care must liaise job allocations.
to try to provide medical, moral, emotional and spiritual support to Concentrating on infectious diseases, the HIV/AIDS epidemic, TB
the children and their families who approach the end of the children’s and malaria has distracted concern from chronic medical conditions
lives.73 Palliative care supports the families involved and is also a great such as renal disease, hypertension, cardiac disorders and has meant
support to staff who may feel helpless when faced by the obvious needs that trauma has been seriously neglected. Road traffic accidents are ris-
of these families. ing in increasingly urban societies, often against a background of poor
town planning, poor driving ability, poor traffic control and unsafe vehi-
cles. Emergency care has not gained priority and is not recognized as
TB Treatment Programs such in many health services. Triage, resuscitation, stabilization and ini-
National TB programs have served as models for the HIV/AIDS pro- tial good clinical management of all acute pediatric illness will prevent
grams. National TB drug policies, protocols, supervision and monitor- many deaths and much morbidity. Reduction in hospital costs and man-
ing are helping to provide treatment for the increased number of TB power use would be considerable.
 HEALTH CARE DELIVERY IN RESOURCE LIMITED settings 25

Summary remains blurred and unfocused. It is these potent, small, personal efforts
of people at the heart of the service that give it life and clarifies the pic-
Health delivery is a big subject and in this chapter a broad brush has ture. They identify real needs and sustain changes that really matter.
been used to give a picture of common health themes in many coun- Creative thinking and great ingenuity go into local health care to pro-
tries in poorer parts of the world. The fine details are place and person vide what on the surface may not look possible.
specific but training, feedback of outcome, praise, constructive criti- It is to these often unrewarded and unrecognized people that this
cism, commitment and faithful service of members of the health team chapter is respectfully dedicated.
are essential to fill in the details of the picture. Without these the picture

References 21. Migration of health workers: an unmanaged 44. Lawn JE, Cousens S, Zupan J. 4 million
crisis. [Editorial] Lancet 2005; 365:1825. neonatal deaths: When? Where? Why? Lancet 2005;
1. Primary Health Care. Report of the International 22. Hongoro C, McPake B. How to bridge the 265:891–900.
Conference on Primary Health Care, Alma Alta USSR gap in human resources for health? Lancet 2004; 45. Khan KS, Wojdyla D, Say L, et al. WHO analysis of
6–12 September 1978, jointly sponsored by the World 364:1451–1456. causes of maternal death: a systematic review. Lancet
Health Organization and the United Nation’s Children’s 23. Eastwood JB, Conroy RE, Naicker S, et al. Loss 2006; 367:1066–1074.
Fund. Geneva: World Health Organization; 1978 of health professionals from subSaharan Africa: the 46. Monasch R, Boerma JT. Orphanhood and
(Health for All Series No 1). pivotal role of the UK. Lancet 2005; 365:1893–1900. childcare patterns in sub-Saharan Africa: an analysis
2. Marcos C. The promise of primary health care 24. Davey G, Fakade D, Parry E. Must aid hinder of national surveys from 40 countries. AIDS 2004;
[editorial]. WHB 2005; 83:322. attempts to reach the Millennium Development Goals? 18:S55–S65.
3. United Nations Development Programme. Human Lancet 2006; 367:629–631. 47. Sach SE, Sach JD. Africa’s children orphaned by
Development Report 2003 Millennium Development 25. The Bamako Initiative of 1987 by African Health AIDS. Lancet 2004; 364:1404.
Goals: A compact end to human poverty. Oxford: ministers. UNICEF. 48. Marcel T. Strengthening district health systems.
Oxford University Press; 2003. 26. Wiseman V. Reflections on the impact of the Bull World Health Organ 2005; 6:403.
4. World Bank. The Millennium Development Goals Bamako Initiative and the role of user fees. Editorial 49. Neilson S, Smutylo T. The Tanzanian essential
for health: rising to the challenge. Washington: World Trop Doct 2005; 35:193. health intervention ‘Spark’: Planning and managing
Bank; 2004. 27. Regina K. Research for free heath. Real Health health resources at the district level. International
5. Jones G, Steketee RW, Black RE, et al, and the News 2006; 4:March. Development Research Centre, Science for Humanity,
Bellagio Child Survival Group. Bellagio report. How 28. The World Health Report 2005; Making every Canada. TEHIP final report, 2004.
many child deaths can we prevent this year? Lancet mother and child count. Geneva: 2005. 50. De Savigny D, Kasale H, Mbuya C, et al. In focus.
2003; 362:977–988. 29. Jamison D, Breman J, Measham A, et al. Priorities Fixing Health Systems. IDRC; 2004.
6. Global Forum Update on Research for Health, Vol. 2. in health. The International Bank for Development 51. Redman TC. The impact of poor data quality on
Poverty, Equity and Health Research. Global Forum on and Reconstruction. Washington DC: The World Bank; the typical enterprise. Communications of the ACM
Health Research: helping correct the 10/90 gap; 2006. 2006: 145. 1998; 41:79.
7. Werner D, Sanders D. Questioning the solution: the 30. World Health Organization Model list of essential 52. Church M, Moyo C. Health Management Infor­
politics of primary health care and child survival, an medicines, 14th edn. Geneva: WHO; 2005. mation System in Malawi. Ministry of Health; 1999.
in depth critique of oral rehydration therapy. Palo Alto, 31. Jamison D, Breman J, Measham A, et al. Priorities 53. Campbell B. Health management information
CA: HealthWrights; 1997. in health. The International Bank for Development systems in lower income countries: an analysis of system
8. Patel M. An economic evaluation of Health for All. and Reconstruction. Washington DC: The World Bank; design, implementation and utilization in Ghana and
Health Policy Plan 1986; 1:37–47. 2006. 145. Nepal. The Netherlands; Royal Tropical Institute: 1997.
9. English M, Lanata C, Ngugi I, et al. Disease Control 32. Brugha R, Starling M, Walt G. GAVI the first 54. Habibzadeh F. Regional associations of medical
Priorities in Developing Countries. Ch. 25 2nd edn. steps: lessons for the global fund. Lancet 2002; journal editors: moving from rhetoric to reality. WHO
Washington DC: World Bank; 2006. 359:435–438. Bulletin 2005; 83:404.
10. Kruske S, Barclay L. The effect of shifting policies 33. Who are we and what do we do? Investing in our 55. Shibaya K, Scheele S, Boerma T. Health Statistics:
on traditional birth attendant training. J Midwifery future. The Global Fund to fight AIDS, Tuberculosis time to get service. Bull World Health Organ 2005;
Womens Health 2004; 49:306–311. and Malaria. ISBN 92-9224-30-7 http://www. 83:722–723.
11. Sibley L, Ann ST. What can a meta-analysis theglobalfund.org/en/ 56. Technical seminars IMCI the integrated
tell us about traditional birth attendant training and 34. Day JH, Lavy CBD. Pills to Africa: how to donate management of childhood illness. Geneva: World Health
pregnancy outcomes? Midwifery 2004; 20:51–60. effectively. BMJ 2001; 323:1315, Trop Doc 2004; Organization Department of Child and Adolescent
12. Sazawal S, Black RE. Effect of pneumonia case 34:65–66 Health; 2001.
management on mortality in neonates, infants, and 35. Medicines Crossing Borders Project WHO step by 57. Brooks WA, Santosham M, Naheed A, et al. Effects
preschool children: a meta-analysis of community step guide to good drug donation practices. Geneva: of weekly zinc supplements on incidence of pneumonia
based trials. Lancet Infect Dis 2003; 3:547–556. WHO; 2002. and diarrhoea on children younger than 2 years in an
13. Winch PJ, Gilroy KE, Wolfheim C, et al. Teaching 36. Murakami H, Phomasack B, Oula R, et al. urban, low-income population in Bangladesh: rando­
mothers to provide home treatment of malaria in Revolving drug funds at the front-line health facilities in mised controlled trial. Lancet 2005; 366:999–1004.
Tigray, Ethiopia: a randomised trial. Lancet 2000; Vientane Lao. Health Policy Plan 2001; 16:98–106. 58. Tielsch J, Khatry SK, StOltzfus RJ, et al. Effects
356:550–555. 37. Ghebejesus TA, Alemayehu T, Bosman A, et al. of routine prophylactic supplementation of iron and
14. Bang AT, Bang RA, Baitule SB, et al. Effect of Community participation in malaria control in Tigray folic acid on preschool mortality in southern Nepal:
home-based neonatal care and management of sepsis region, Ethiopia. Acta Trop 1996; 2:142. a community-based, cluster-randomised, placebo-
on neonatal mortality: field trial in rural India. Lancet 38. Sall M, Sylla J. African prime ministers take a lead controlled trial. Lancet 2006; 367:144–152.
1999; 354:1955–1961. in child survival. Lancet 2005; 366:166–189. 59. Robertson MA, Molyneux EM. Triage in the
15. Parry E. Volunteers can contribute to health care. 39. Hutton G, Tanner M. The sector-wide approach: developing world – can it be done? Arch Dis Child 2001;
Id 21 insights DFID Tropical Health Education and a blessing for public health? Bull World Health Organ 85:208–213.
Training London: 2004. 2004; 82:893. 60. Molyneux EM, Ahmad S, Robertson A. Improved
16. World Health Report, 2006. Working together for 40. Bobadilla J, Cowley P, Musgrove P, et al. Design, triage and emergency care for children reduces
health. Geneva: WHO. content and financing of an essential national inpatient mortality in resource – constrained setting.
17. The crisis in human resources for health. package of health services. Human Development WHO Bulletin 2006; 84:314–319.
[Editorial] Lancet 2006; 367:1117. Department. Washington DC: World Bank; 1993. 61. Sazawal S, Black RE, Ramsan M, et al. Effects
18. Chen L, Evans T, Anand S, et al. Human resources 41. Jamison D, Breman J, Measham A, et al. Priorities of routine prophylactic supplementation of iron and
for health: overcoming the crisis. Lancet 2004; in Health. The International Bank for Development and folic acid on admission to hospital and mortality in
364:1984–1900. Reconstruction. Washington DC: World Bank; 2006: 136. preschool children in a high malaria transmission
19. Anand S, Barnighausen T. Human resources and 42. English M, Esamai F, Wasunna A, et al. Delivery setting: a community-based, randomised, placebo-
health outcomes. Lancet 2004; 364:1603–1609. of health care at the first-referral level in Kenya. Lancet controlled trial. Lancet 2006; 367:133–143.
20. Jamison D, Breman J, Measham A, et al. Priorities 2004; 364:1622–1629. 62. Integrated Management of Childhood Illnesses.
in Health. The International Bank for Development 43. Safe Motherhood Initiative 1987 Nairobi UNICEF and CAH (WHO). Geneva: WHO; 2001.
and Reconstruction. Washington DC: The World Bank; Kenya. Safe Motherhood Inter-Agency Group. www. 63. Adam T, Manzi F, Schellenberg JA, et al. Does the
2006:131. safemotherhood.org/reviews/publications/html integrated management of childhood illness cost more
26 FOUNDATIONs OF HEALTH AND PRACTICE

than routine care? Results from the United Republic of 69. Chintu C, Bhat GJ, Walker AS, et al. Co- 75. Moszynski P. UNICEF warns of ‘epidemic’ of
Tanzania. Bull WHO 2005; 83:369–377. trimoxazole as prophylaxis against opportunistic childhood malnutrition. BMJ 2006; 332:1051.
64. Bryce J, Cesar V, Habicht J-P, et al. The infection in HIV-infected Zambian children (CHAP): Prognosis for children. A report card on nutrition
multicountry evaluation of the IMCI strategy: lessons a double blind randomised placebo-controlled trial. www.unicef.org.
for the evaluation of public health interventions. Am Lancet 2004; 364:1865–1871. 76. Ciliberto MA, Sandige H, Ndekha MJ, et al.
J Public Health 2004; 94:406–415. 70. Merriman A, Kaur M. Palliative care in Africa: Comparison of home-based therapy with ready-to-
65. Daza P, Banda R, Misoya K, et al. The impact an appraisal. Lancet 2005; 365:1909–1911. use therapeutic food with standard therapy in the
of routine infant immunization with Haemophilus 71. Defippi K, Dowing J, Merriman A, et al. treatment of malnourished Malawian children: a
influenzae type b conjugate vaccine in Malawi, a A palliative care association for the whole of Africa. controlled clinical effectiveness study. Am J Clin Nutr
country with high human immunodeficiency virus Palliat Med 2004; 18:583–584. 2005; 81:864–870.
prevalence. Vaccine 2006; 24:6232–6239. 72. Merriman A. Making drugs available. AIDS 77. Deen JL, Funk M, Guevara VC, et al.
66. Malawi Demographic and Health Survey; 2004. Action 1998; 41:8. Implementation of WHO malnutrition guidelines in
Zomba, Malawi: National Statistical Office. 73. Spence D, Merriman A, Bunagwaho A. Palliative the management of severe malnutrition in hospitals
67. Dunn DT, Newell ML, Ades AE, et al. Risk of care in Africa and the Caribbean. Plos Med 2004; 1:e5. in Africa. Bull WHO 2003; 81:237–243.
human immunodeficiency virus type 1 transmission 74. Balasubramanian VB, Oommen K, Samuel R. 78. Ndekha MJ, Manary MJ, Ashorn P, et al. Home-
through breast feeding. Lancet 1992; 340:585–588. DOT or not? Direct observation of anti-tuberculosis based therapy with ready-to-use therapeutic food is
68. Cutting W. Breast feeding and HIV – a balance treatment and patient outcomes, Kerala State. Int of benefit to malnourished HIV-infected Malawian
of risks. J Trop Paediatr 1994; 40:6–11. J Tuberc Lung Dis 2000; 4:409–413. children. Acta Pediatr 2005; 94:222–225.
 10001

4
Preventive pediatrics
Harry Campbell, Rachael Wood

Determinants of child health 27 Immunization 34

Contribution of health services to improving child health 27 Importance and efficacy of immunization in the
Health promotion and disease prevention 28 prevention of disease 34
Child health advocacy 28 Vaccines: present and future 35
A life course approach to disease prevention 29 Adverse reactions and contraindications to immunization 35
Prevention before conception and during pregnancy 29 Immunization schedules 36
Prevention in neonatal period 29 Developing country schedules 36
Prevention in infancy and early childhood 31 Immunization coverage 38
Prevention in later childhood and adolescence 32 Specific immunizations 38
Support for parenting 33 Information resources relevant to preventive pediatrics 43
Child health screening and surveillance 33

Determinants of Child Health improving child health has been considered to involve ‘placing the
(see also Ch. 2) health of children and their families in its full social, political and eco-
nomic context’ and to be ‘the responsibility of decision makers in all
Any child’s health, development, and welfare reflect the interaction of organizations in all sectors of the economy’.4
various factors including: Nevertheless, the organization and provision of health services can
l the individual’s genetic endowment; and should make a substantial contribution to improving child health.5
l environmental factors including aspects of the direct physical This is acknowledged in the UN Convention on the Rights of the Child,
environment such as housing and ‘behavioral’ factors such as which places a clear responsibility on the State to provide access to pre-
nutrition; ventive care (article 24).6 The remainder of this chapter will therefore
l the quality of interpersonal relationships, particularly with the concentrate on the actions that can be taken principally by health ser-
child’s primary caregiver; vices to promote child health and development and prevent childhood
l the wider social, political, and economic circumstances the child is illness and handicap, and will be considered from the perspective of
living in; health professionals working in pediatric services.
l the availability of health care. In order to maximize their potential to improve child health, health
The health of the child’s mother is also an important factor in deter- services should meet certain key criteria. Services should be:
mining the health of the child, whether this is, for example, in ensuring l universal, i.e. accessible to all members of the population;
proper antenatal care of mothers with diabetes or treating and support- l comprehensive, i.e. focused on health promotion and disease
ing mothers with psychiatric problems or alcohol addiction. prevention as well as curative care;
The impact of this complex web of determinants of health is reflected l integrated with other services that impact on the well-being of
in the marked socioeconomic inequalities that exist in child health.1 children, in particular social services and education;
These inequalities are seen both between countries of the resource l responsive to the health needs of the population;
rich and developing worlds and also within individual countries.2 Even l centered on the individual health needs of children and their
within resource rich countries, almost all indicators of child health show families;
a marked social gradient, with children living in disadvantaged circum- l evidence based and of high clinical quality.
stances having poorer health.3 In Scotland, for example, children from In addition, services should be provided with the aim of reducing
the most deprived areas are more than twice as likely to die compared to inequalities in health. This may mean that additional steps are taken
children from the most affluent areas (Fig. 4.1) with some specific health to ensure very high coverage of essential immunizations in all sec-
problems showing particularly marked inequalities (Ch. 2). tions of the population or that particularly vulnerable children (such
as those in families of asylum seekers or travelers, those from eco-
Contribution of Health Services nomically deprived areas, or ‘looked after’ children) require additional
to Improving Child Health special services to ensure they receive adequate preventive care. The
capacity to monitor the contribution of health services to improving
As many of the factors that influence child health are modifiable, there child health and reducing health inequalities on an ongoing basis is
is considerable scope for improving children’s health and reducing also important.
inequalities in child health. It is clear, however, that much of the activ- It is clear from the above that a focus on health promotion and dis-
ity required to bring about this improvement lies outwith the remit of ease prevention should pervade the day-to-day practice of all health
traditional health services. The profound impact of political action, eco- professionals who work with children. This will require skilled team-
nomic progress, improved education and social change on child health work in working with children and their families, other health workers,
is shown by falling morbidity and mortality long before antibiotics, vac- and colleagues outside the health services such as teachers and social
cines and high-technology medicine became available (Fig. 4.2). Thus workers. 27
28 foundations of health and practice

8.0 Secondary prevention aims to reduce the prevalence of disease

through early diagnosis and treatment that either leads to a cure or to a
Mortality rate per 10,000 population
7.0 reduction in the more serious consequences of disease. Examples of sec-
6.0 ondary prevention interventions include screening for hypothyroidism
in neonates to permit early replacement of deficient thyroxine and pre-
5.0 vention of the disabilities of cretinism, and otoacoustic emission screen-
aged 0–14

4.0 ing for hearing loss in neonates to improve outcomes for deaf children.
The term ‘secondary prevention’ can also be applied to the termina-
3.0 tion of pregnancy to avoid the birth of a fetus with a recognized severe
2.0 impairment.
Tertiary prevention aims to reduce the progress or complications of
1.0 established disease, thus limiting its impact. It consists of measures to
reduce impairments and disabilities from the disease or injury and so
0
1 2 3 4 5 minimize any handicap which may result. For example, tertiary preven-
Least deprived Most deprived tion through the rehabilitation of children with polio can enable them
Deprivation quintile
to take part in daily social life and bring about a great improvement in
Fig. 4.1 Death from all causes by deprivation quintile, children aged the well-being of these children.
under 15 years, Scotland 2004. From ISD (GRO(S) death registrations and
mid-year population estimates). Deprivation quintiles based on Scottish Child Health Advocacy
Index of Multiple Deprivation. ISD (SMR01 and mid-year population
estimates). In addition to improving child health and reducing inequalities through
a focus on health promotion and disease prevention in their day-to-day
clinical practice, pediatricians are well placed to make a valuable con-
Health Promotion and Disease tribution to child health advocacy.10 This is advocacy that strives to pro-
Prevention mote action within society that addresses the broader determinants of
child health (e.g. socioeconomic circumstances, quality of the environ-
Health promotion is usually thought of broadly as encompassing all ment, housing, nutrition and education).
activities aiming to improve health and hence incorporates specific Child health advocacy often begins with an individual child or family
­disease prevention activities along with broader health education and and then may extend into local, regional or even national public health
lobbying for socioeconomic change.7,8 action. It involves taking action to promote health beyond treatment of
Disease prevention is commonly categorized into primary, second- a medical condition. This action can be taken as an individual or col-
ary, and tertiary prevention.9 Primary prevention aims to reduce the lectively. Specific steps in individual pediatrician advocacy have been
incidence of disease by controlling causes or risk factors. Examples of described by Waterston and Tonniges11 as:
primary prevention interventions include vaccination against mea- l identifying a preventable problem in one child;
sles, the use of condoms to prevent human immunodeficiency virus l helping that child overcome the problem;
(HIV) infection, antenatal folate supplementation, and fluoridation of l drawing conclusions in relation to the factors that led to the
­drinking water. problem in the first place;

160 160

Mortality in age group 1–2 years (per 1000 children in this age group)
Meningococcal C vaccine
Whooping cough vaccine

140 140
Infant mortality
Diphtheria vaccine

Measles vaccine

120 120
MMR vaccine
Polio vaccine
Infant mortality (per 1000 live births)

Sulfa drugs

Antibiotics

100 100

80 80

60 60

40 Mortality in age 40
group 1–2 years

20 20

0 0
1851 1871 1891 1911 1931 1951 1971 1981 1991 2000
Fig. 4.2 Infant mortality in England and Wales and important developments in medicine.
 preventive pediatrics 29

l identifying the means to tackle these factors; during and after pregnancy has emerged as a major risk factor for sud-
l influencing government or policy makers to change or reform den infant death syndrome (SIDS) in studies carried out after the wide-
the system that fostered these factors or introduce appropriate spread recommendations on infant sleeping position led to a reduction
legislation. in the number of infants found dead in the prone position.24–26 Reduction
Waterston gives as an example of such individual advocacy the work of smoking levels in the population requires coordinated action includ-
of Hugh Jackson, who cared for a child who died of an accidental drug ing increasing the cost of tobacco, banning advertising, and restrict-
overdose and whose later action led to the development of legislation ing the places where people are allowed to smoke.27 At the individual
requiring the use of childproof medication.11 Collaborative action of level effective interventions to help people to stop smoking are available.
pediatricians and other professionals involved in child health can advo- Pregnant women can benefit from smoking cessation counseling and
cate effectively against the tobacco industry, the motor industry, or baby nicotine replacement therapy, and maternity services should ensure all
milk manufacturers. pregnant smokers have access to specialist stop smoking care.19,28
Well-organized antenatal care makes a vital contribution to mater-
A Life Course Approach to Disease nal and child health. In the UK, current recommendations suggest that
Prevention routine antenatal care for healthy women should include provision of
health information along with the offer of screening for anemia, red
In this section we provide an overview of effective measures to prevent cell alloantibodies, syphilis, hepatitis B and HIV infection, immunity
disease and improve health in childhood. In later sections we go on to to rubella, asymptomatic bacteriuria, pre-eclampsia, Down syndrome,
consider three preventive interventions in more detail, specifically sup- structural anomalies, and fetal growth.20 Women with particular risk
port for parenting, child health screening and surveillance, and immu- factors may require additional antenatal care, for example individu-
nization. We concentrate on measures of relevance in the developed als with a family history of a genetic disorder may benefit from genetic
world, as preventive actions relevant to promoting child health in devel- testing. Over recent years there have been major advances in the under-
oping countries are presented in Chapter 3. standing of the molecular basis of medical disorders and in the develop-
Table 4.1 presents effective disease prevention measures for each ment of new techniques to identify genes associated with disease. Most
stage of childhood using a ‘life course’ framework. Over recent decades of the genes for the more common inherited (single gene or ‘Mendelian’)
there has been a general resurgence of interest in the life course approach disorders have now been identified (see Chapter 14). This in turn often
to health that emphasizes the long-term health implications of the fetal makes prenatal detection and genetic counseling possible.29–31 In con-
and early childhood environment.12–14 It is now widely accepted that junction with this there have been advances in prenatal diagnostic
exposure to adverse environments in early life can irreversibly program techniques, for example using amniocentesis or chorionic villous sam-
higher risk of certain diseases, in particular hypertension, diabetes and pling. Secondary prevention by termination of pregnancy has therefore
ischemic heart disease, later in life.15–18 Thus some of the preventive become more widely practiced.
actions presented here may not only lead to improvement in child health In early pregnancy the avoidance of various teratogens is an
but also represent effective interventions against some of the common important preventive measure. The following drugs and chemicals
diseases of public health importance in the adult population. have been found to be associated with fetal defects: diphenylhydantoin
Some of the preventive interventions presented in Table 4.1 come (phenytoin), trimethadione, paramethadione, valproic acid, carbam-
under the remit of the UK universal child health promotion program azepine, thioureas, carbimazole, methimazole, isotretinoin, vitamin A,
that is mainly delivered by midwives, health visitors and the school etretinate, thalidomide, warfarin, methotrexate, corticosteroids,
health service with support from GPs and specialist pediatricians (see androgens, progestins, diethylstilbestrol, iodine, lithium, mercury and
later section on Child health screening and surveillance) however chlorobiphenyls. A number of other drugs and chemicals have fea-
knowledge of all of the interventions listed is relevant to the day-to-day tured in retrospective studies or case reports but associations with
practice of pediatrics. The centrality of preventive work to pediatrics is fetal defects have not been confirmed by subsequent investigations (see
reflected in the fact that many of the interventions presented in Table Stevenson32 for a further discussion). Exposure to abdominal X-rays
4.1 are discussed in other chapters of this book. For this reason we have should be avoided, but there is no evidence that ultrasound examinations
cross-referenced other chapters in addition to providing references to are harmful to the fetus.
the evidence base for many of the interventions listed.

Prevention in neonatal period

Prevention before conception and during Good intrapartum obstetric care and subsequent effective monitor-
pregnancy ing, investigation and treatment of the many disorders from which the
Genetic factors in either parent, the present and past health of the newborn infant may suffer are important preventive measures. Such
mother, her age, her habits (e.g. smoking, alcohol), the frequency of disorders include asphyxia, birth injury, low birth weight and hyperbili-
her pregnancies, her previous immunizations (e.g. rubella and tetanus) rubinemia. Neonatal screening procedures are discussed in the section
and her social class are all factors which can potentially influence the on Child health surveillance and screening.
health of her infant before it is conceived, and control or elimination of The promotion of breast-feeding is a crucial preventive measure.
adverse factors will play an important part in prevention. A mother’s Breast-feeding reduces the risk of necrotizing enterocolitis, diarrheal
own birth weight and growth through childhood and general nutrition disease, lower respiratory infections, otitis media, and other serious neo-
during adulthood and pregnancy are being increasingly recognized as natal infections.33,34 It also appears to reduce the risk of childhood obe-
important determinants of her child’s fetal growth patterns and hence sity, probably through better development of appetite control.35 Recent
subsequent risk of chronic diseases.15–18 This developing understanding evidence has further linked lack of breast-feeding with poorer intellec-
of ‘biological programming’ demonstrates the importance of focusing tual development, possibly due to the lack of certain long chain fatty
on health promotion for women and young children in order to break acids, essential for normal brain development, in most breast milk sub-
intergenerational cycles of poor health. It also shows how complex dis- stitutes, although it is difficult to totally exclude the possibility of con-
ease prevention interventions can be, for example it is not yet clear how founding from these studies.36,37 Frequent breast-feeds given over a
best to optimize maternal nutrition.19 prolonged period also significantly reduce fertility and increase the birth
The effects of a pregnant woman’s smoking or alcohol consumption interval, with indirect benefits to both mother and infant.38
on her fetus are now well recognized and demand preventive action.20 WHO and UNICEF are coordinating a global initiative (the Baby
Smoking can lead to intrauterine growth retardation, certain congeni- Friendly Hospital Initiative) to promote breast-feeding and to improve
tal anomalies, fetal loss, and preterm delivery.21–23 In addition, smoking health service support for breast-feeding mothers.39,40 Hospital routines
Discovering Diverse Content Through
Random Scribd Documents
 Then the detective had passed out one of the drawings, putting it
face upward on the desk in front of the chief; and the latter had
exclaimed at once:
“‘The Leopard!’ Of course it is ‘The Leopard.’” And then he had
added as we have noted it down here. “Without a doubt.”
Nick Carter raised his brows, interrogatively.
“‘The Leopard?’” he repeated questioningly.
“Yes. That is the name by which she is best known, Carter. She
had a different name for use in each one of the capitals of Europe,
but ‘The Leopard’ is the one by which she is best known, and more
generally recognized. I had been wondering what had become of
her.”
“So you know all about her, chief?”
“All about her? No, indeed; very little about her, as a matter of
fact—and a very great deal about her, too.”
“Isn’t that statement of yours rather ambiguous?”
“Yes, it is; but it comprehends precisely what I wished to say.”
“Would you mind being more direct about it?”
“I’ll be as direct as I can. What do you desire to know about
her?”
“Everything.”
“Ah, my dear friend, Carter, but that is impossible.”
“Then all that you can tell me about her.”
“That is different. Yet—if you were to give me the precise line
concerning which you wish information we might get at it sooner.”
“Why? What is the matter with general information concerning
her?”
“There is no such thing as general information concerning her.
She is not ‘general’ in any sense of the word. She is a many-sided
woman. Young, you will say? Yes; but not so young as she appears
to be. Beautiful? Ah, as a witch! Fascinating? She is an houri. There
are no words to describe her accurately. What is the circumstance
which leads you to make inquiries concerning her, my friend?”
The detective hesitated a moment; then he said:
“She is just now engaged, in company with a man whom we
know on our side of the water as Bare-Faced Jimmy, in rather a large
scheme. Jimmy has stolen the identity of a young Virginian who is
doubtless dead—it isn’t unlikely that Jimmy murdered him and that
this woman helped him do it. ‘The Leopard,’ as you call her, has
married Jimmy——”
“Wait, wait, wait! Married, you say? Impossible!”
“Eh? Why impossible?”
“Because, why should she have married a criminal, when she
could have had her pick of titles over here many times?”
“That is a question I cannot answer; only there is no doubt that
she is married to Jimmy.”
The chief of the secret police of Paris shook his head with
emphasis.
“Impossible!” he said again, with conviction.
“Why?” repeated the detective.
“Because—ah, who can give a reason for what women do, or
refuse to do? Not I, although I have been studying them for years.”
“But you have a reason for such a decided opinion, chief.”
“Assuredly. Of course, Carter. There is a reason. We have it set
down in our dossier of her in the books; there it is a cut and dried
opinion; just a practical one. Perhaps that is the answer you want to
your question.”
“Perhaps it is. Let me hear it.”
“Wait. I will send for the book.”
“No. Tell me about it, and her, first. I would rather have your
version of it. Later, if you will permit it I will read the dossier.”
“Assuredly you shall read it.”
 “Now, what is that cut and dried reason? Tell me that. I have an
idea that it will supply some sort of a pointer in the investigation I
am making.”
“Possibly. Who knows? I have just told you that she might have
had her pick of titles, here in France; or Austria; or Germany; or
Italy; or even in Russia. Everybody who came in contact with her fell
in love with her. She has been the ruin of a score of good men in the
secret police of several countries. Two of my own men committed
suicide because of her. She led them to betray their trusts, and so,
dishonored them. Ah, she is a wonder, that woman! That leopard!”
“But that is not the reason you spoke about for her not
marrying.”
“No, it is this. There was a Duc de Luvois—a rich man with an
honored name, which he offered to bestow upon her, together with
his fortune. He laid them both at her feet, and she refused them and
him. It was her reply to him that is used now in the dossier, as the
reason why she will never marry.”
“Good. What was that reply, chief?”
“The duke repeated it to a friend of his before he shot himself
after her refusal. ‘She told me,’ he said to his friend, ‘that there is
only one name in all the world which she will ever consent to bear,
and that as there is small chance of that name ever being offered to
her is not likely that she would ever marry.’ Now you have it, Carter.
That is the cut and dried reason. Cannot you read between the
words all that they imply?”
“Yes; I think so. Still, I would like to have your version, chief.”
“You shall, then.”
“Thank you.”
“I told you a moment ago that while we know a great deal about
her, we know, in fact, a very little. When I made that remark I meant
that we know absolutely nothing concerning her history before she
arrived at womanhood. In other words, we know everything about
her, for the past eight years—and we know absolutely nothing
concerning her before that time. We do not know where she came
from or what her country is. Have you got that in your mind?”
“Yes.”
“Well, now refer again to what she told the duke when he asked
her to be his wife.”
“I do.”
“There can be only one explanation of that expression. It meant,
if it meant anything at all, that once, before we knew anything about
her, she loved a man who was the cause, directly or indirectly, of her
entering upon a career that brought her to the notice of the police.
It meant that the man is still alive, and that he might yet offer her
his name, and that if he did so, she would accept it; and that if he
failed to do so, she would never accept any other name. I know that
I am a romantic Frenchman, but that is the way I read that answer
she made to the Duc de Luvois.”
“Very well, chief, I accept your version.”
“Yet you say that she is married.”
“Yes; there is no doubt of it.”
“But you suggest that the man she has married is a criminal.”
“He is one.”
“Then I do not believe——”
“Wait, chief.”
“Well?”
“Suppose that the man was not always a criminal? Suppose that
once upon a time he bore a splendid name, and was in line to
succeed to a title? Suppose——”
The chief half started from his chair, then sank back again into its
depths.
“You mean that she has married the only man she would give her
liberty to—the man to whom she referred in that talk with the
duke?”
“Yes.”
 “And that he is now a criminal, and that you are on his track—
and hers?”
“Yes, again.”
“By Jove, Carter, I cannot believe it! Do you know who she is?”
“No.”
“But you are on the track of finding out?”
“Yes; if you will assist me, chief.”
“You may be sure that I will do that, Carter, to the extent of my
ability; and of everything that this office can supply. Where is she
now? In America?”
“Yes.”
“Look here, Carter; before you go more deeply into this affair I
should tell you one thing about the woman.”
“Well?”
“Although she has been the cause of many a crime, and is
responsible for the sudden taking off of many men, although she has
filled a large place in our records for a long time, there is not, to-
day, a thing against her for which she could be arrested, with the
least chance of conviction. It has even been said of her that she has
lived a spotlessly moral life, and so far as my own knowledge goes,
it is the truth. But, she is none the less a dangerous woman.”
“I know. I have seen her.”
“Ah; then you do know. Why, my friend, she even tried her wiles
upon me—and I nearly fell. It was chance that saved me, rather
than my own good sense.”
“And you are more than half in love with her yet, chief. I could
see that when you looked at her picture.”
“No; I am not in love with her; not in the least. But I am
fascinated by her. And there is a difference, Carter.”
 CHAPTER XIX.
TRAILED BY FATALITIES.

Although the detective had been in London before he visited

Paris, he had not sought Nan, who was in that city on the quest he
had given her; just now, in talking with the chief, he half wished that
he had done so.
But he was satisfied that he had not made a mistake in his
conjectures concerning Juno.
She had been a dangerous woman always. According to the
conversation the detective had had with Nan—always provided that
this woman was the same who had once been Siren—she had begun
to scatter danger around her, even when she was still a child. The
peril of her propinquity had grown greater with the physical
development until now even the chief of the secret police of Paris
acknowledged that he had nearly been one of her victims.
While the detective was studying the chief, the chief was
contemplating him. It was the latter who spoke first.
“I wish, Carter, that you would take me more into your
confidence,” he said. “If I was made aware of precisely what you
want, it might be that I could help you. Or, will you have a look at
the dossier first?”
“The dossier, please; after that I will try to be quite frank. But,
first—you assure me that this woman has no criminal record?”
“None that could be designated as such. You will discover all that
when you read the dossier.”
“Will you tell me just why she has made herself so prominently a
figure for the police to study?”
 “Ah! That is different. ‘The Leopard’ has been what you might call
‘a near-criminal’ ever since she first came to our notice; but she has
never been quite one, that we can ascertain, or prove.”
“I see. Will you tell me how she first attracted the attention of
the police of Paris? That might be interesting.”
“Yes. That was a curious case. There was a certain Prince
Turnieff here in Paris as a diplomatic agent of the Russian
government. We were keeping half an eye upon the prince, not
being quite sure of his status, and we noticed that he was frequently
seen in the company of a beautiful young woman who was a
stranger to us.”
“And that woman was Juno?”
“Eh? Who?”
“The woman you call ‘The Leopard.’ I know her by the name of
Juno.”
“Ah; an apt name. It fits her. Yes; that woman was—Juno.”
“Well?”
“It was known that the prince was a very rich man in his own
right. Whatever his capacity was here in representing his
government, he lived in regal style, spent money lavishly, possessed
a fortune in jewels and precious stones, and did what in your
country you would call cutting a wide swath. He was also a
handsome man, young and gifted. Just the sort of a man that the
average woman would admire. See?”
“Perfectly.”
“He lived in a palace in the Faubourg St. Germain, which he had
leased, furnished; he maintained a retinue of servants and lived like
royalty. One day, at four o’clock in the afternoon, he was found dead
in the library of that palace. There was a bullet hole in his right
temple, and he had died instantly.
“But, Carter, there seemed to be no doubt that he had killed
himself. Everything in the room bore evidence of that, even to a half-
written note that he had left on the table near where the body was
discovered.
“But the last person who was known to have been with him was
the woman you call Juno, and whom I call ‘The Leopard.’ She was
arrested, questioned, subjected to every art that the French police
employ to force her to tell all she knew of the circumstances, but we
might just as well have left her alone, so far as any result was
obtained from her. She smiled at us, defied us, bewitched us,
fascinated all of us. That was the time when I so nearly fell under
her spell myself. She was permitted to go; but ever after that we felt
it our duty to keep her under close surveillance.
“But that is not all of the story, as it relates to Prince Turnieff.
“I have said that he made a great display of wealth; that he had
in his possession several fortunes in jewels. I should have added
that he did no business at the banks, and that because of that, it
was assumed that he had brought with him all the cash he required.
“It was estimated that he must have had a million francs or more
in cash in his house at the time of his death, to say nothing of the
jewels. When I tell you that after his death there was no trace found
of either cash or jewels, and that none of it has ever been seen
since, you will understand how it was that Juno, as you call her, was
suspected.”
The detective nodded.
“Still,” continued the chief, “there was nothing against the
woman. It could not even be established that she had been other
than a friend and a companion of the rich prince. On that day when
he killed himself—or was presumed to have done so—she had been
with him in his library only a short time, and it was not until more
than two hours after her departure that the body was discovered.
“You wished to know what it was that brought her first to our
attention; that was the circumstance.”
“And the next one? What was that?”
“It happened to be another victim—if I may use the term. This
time it was an Austrian. We did not know till after his death that he
was a spy in the service of Austria, but that developed later.”
“Did the Austrian also kill himself?”
“No; he was murdered in cold blood. But it happened more than
an hour after he had parted with your Juno, and there was not a
thing to connect her with the crime, save that she had been with
him an hour previously—and that all his money and valuables had
disappeared at the time of his death. He was killed while he was the
occupant of a closed carriage in which they had been riding
together; the carriage stopped at the door where she lived to put
her down, and the driver testified that the man was alive after she
left him.
“He was a diplomatic agent, also; and, as in the other case, it
was said that papers of great value, as well as other things, had
disappeared from his person.
“I could give you other incidents of the same sort, Carter, that
have happened in her career. The police of St. Petersburg could do
the same. Vienna, Berlin, and other centres of activity could each
add a quota; and there you are. Now, I will ask you to read her
dossier, and after that we will discuss her further, if you desire it.”
The chief touched a button and gave a direction; and, presently,
with an open volume before him, Nick withdrew into a corner and
passed half an hour in studying what had been written down in it
under the name of “The Leopard.”
At the end of that time he closed the volume and drew his chair
forward.
“The dossier tells me nothing new,” he said to the chief. “It gives
me no further information than that already supplied by you, save
that it goes into details rather more particularly.”
“Exactly.”
“I find that the part which interests me most is in that sentence
already quoted, used by her in her rejection of the suit of the Duc de
Luvois. I believe, chief, that I can establish the identity of the
woman, through an agent of mine who is now in London. I think
that I can do so, but I am not certain as yet.”
 “Then you will accomplish more than all my force has been able
to do, Carter.”
“Through an accident, believe me; not through lack of zeal, or
because of deficient ability on the part of your men.”
“That’s as it may be. Are you willing to tell me who you think she
is?”
“Unless you insist, I would rather keep that to myself until I am
certain.”
“Very well. I cannot blame you for that. Does the woman know
that you are on her track? Does she suspect that you are searching
out her record?”
“She suspects; I do not think she knows.”
“You are positive that she suspects?”
“No; I am not positive; but I suppose she does suspect. I am
morally certain of it.”
“Carter, have you read that dossier very carefully?”
“Yes.”
“Have you taken careful account of the number of fatalities to
others that have followed in her wake wherever she has gone?”
“Yes.”
“Very well. There is one thing which is not written down in words
upon that record, and it is to that I now call your attention.”
“What is it, chief?”
“I have said that we have never been able to prove anything
against your Juno. The principle reason for that is that in every case
where it has been supposed that she could no longer escape us,
death by violence, self-inflicted, or otherwise, has removed the
person, or persons, whose testimony might have convicted her. Does
that statement convey an idea to you, Carter?”
“Yes. You mean to tell me, in that roundabout fashion, that if
Juno suspects that I am on her track my own life is in danger. Is that
it?”
 “My dear fellow, I am speaking only in generalities; you may call
it superstition if you like; but fatalities have pursued those who have
been inimical to the peace and liberty of that woman. Whether it is
the result of coincidence or of design, I am not prepared to say; but
if I were on that woman’s trail and had unearthed anything which
could be used against her, and knew that she suspected it, I would
make my will and all arrangements for a sudden taking off,
confidently expecting that death might overtake me at any moment.”
“You would make Juno out a murderer, chief.”
“I would make her out what she is called, a leopard, who
destroys. Whether she strikes the necessary blows herself or has
them delivered for her—who can say? If she plots the fatalities and
arranges them—who can tell? If the fiend himself protects her and
drives men mad and makes them kill themselves—who can
determine? The facts remain. Those who are hostile to that woman
die. There you are.”
“I have lived rather a long time, chief, and I am not dead yet.”
“No; but if that woman suspects that you are on her trail with
any chance of doing her a permanent injury, I wouldn’t give a
centime for your life; not one.”
The detective shrugged his shoulders and smiled; but he made
no further comment.
Then the chief took the receiver from the telephone hook, in
answer to a call that came in at that moment.
 CHAPTER XX.
THE SIREN EXERTS HER SKILL.

Had the detective known that Juno was in Paris at that moment
he might possibly have paid more attention to the remarks of the
chief of police.
But he was soon to know it.
When the chief took down the telephone, Nick picked up a paper
that was lying on the desk and was scanning the front page, when
an ejaculation from his companion caused him to turn his head and
regard the man attentively.
“Very well,” he heard the chief say over the phone, in French, “let
nothing throw you off the scent, Mouquin. Keep me informed. Let
me know everything concerning her with the least possible delay. It
is vitally important, just at this time.”
He replaced the receiver on its hook and turned to Nick Carter.
“My friend,” he said, “you will admit that perhaps I am not an
unwise prophet. You are here in Paris on the trail of The Leopard;
The Leopard is here in Paris on your trail. I am so informed by one
of my best men, Mouquin by name. What will you?”
“Do you mean to tell me that Juno is in Paris?” asked Nick,
interested.
“She has only just left the train at the gare du nord,” was the
reply.
“Are you sure that there can be no mistake, chief?”
“Perfectly.”
“And she has entered the city openly? Without any attempt at
disguise?”
 “Yes. Why should she disguise herself? There is nothing for which
we can apprehend the woman. She knows that every footstep she
takes while she is here will be watched. She has known that for a
long time. I think she rather likes it; so why should she not come
here openly?”
“What has brought her here, I wonder?” mused the detective.
“You have brought her here, my friend,” replied the chief.
“But, why should she follow me here? I know why she would like
to have me out of the way—dead, if you will—but——”
“Listen here, Mr. Carter. On the other side of the water, in your
own country, you are something of a celebrity. Murder is as common
there as here, but it is done differently, as a rule. Believe me, you
are a dangerous man to that woman, and, being dangerous, she
desires to overcome that danger. Very well; there is no place on
earth where she would rather see you just now, than here in Paris.
In coming here you have played directly into her hands.”
“Well, admitting that it is so, what then?”
“What then? He asks me, what then? Death, then, my friend!”
“And you, the chief of the secret police of Paris, sit here, in your
chair, in your own private office, and tell me that? And you still
permit such a woman to run at large in the streets of your city!”
Nick smiled when he made that remark; smiled tauntingly.
The chief hunched his shoulders, spread out his hands, palms
upward, screwed his face into an indescribable expression, and
replied:
“What can I do? What could you do, in my position? Nothing.
Nothing at all.”
“I could at least keep the woman under such close surveillance
that she would not make a move that I did not know about. She
would not——”
“Ah! Ah! Ah! Well, I will do that. I have already given directions
to that effect. But I have done it before, times without number—and
it has always been the same.”
 “The same what? Do you mean that she gives you the slip?”
“I mean that, although my men believe that they can put their
hands upon her at any moment of the day or night, while they watch
her, yet—yet the things that I have attempted to describe, happen.”
“Who is the man who telephoned to you just now, chief?”
“Louis Mouquin; one of my best men. There is no better
detective in Paris to-day, and not another one who is as good at
shadowing.”
“He has shadowed her before, has he not?”
“Yes. Many times.”
“Then you can rely upon it she is ‘onto his curves,’ all right.”
“She is—what?”
“She knows the man and his methods. No matter how good he
is, he is no good so far as she is concerned.”
“What, then, would you do, Carter? I am quite willing to take any
suggestion from you that you can make.”
“Very well, then, I will make this one. I’ll take the job of trailing
that woman myself.”
“You, Carter?”
“Yes.”
“For me? For this department?”
“Certainly; only, if I do that—for you, and not wholly for myself—
you must call off your own men and leave it all to me alone.”
For the fraction of a moment the chief hesitated. Then a quiet
smile stole across his face, and he replied:
“Very well. It shall be done. I appoint you—without pay; eh?”
“Certainly. I am now a special, under your orders; but there is
one other thing I must have, chief.”
“What is that?”
“Authority.”
“Eh? What sort of authority?”
 “A written appointment over your signature; a badge; anything. I
don’t care what it is, so long as it bestows the authority I want and
gives me the command over any of your men whom I may chance to
meet.”
“You shall have that. I will give you the badge which will place
you next in authority to me. That place happens to be vacant just
now. It shall be yours so long as you remain in Paris. And I will send
you to Mouquin. He will show you to what place——”
“Pardon me, chief, but I would prefer it in another way.”
“Well? As you please. What, then?”
“I will sit right here until Mouquin telephones to this office again.
When he does so, you will tell him to bring The Leopard here to you.
If she should hesitate to come——”
“Oh, she will not do that. She will probably be delighted to come
here.”
“Indeed? Well, I will remain here till she arrives. I will see her
here.”
“And then——”
“Chief, you and I work on different plans and by different
methods; but we work to the same ends.”
“Assuredly.”
“In the case of a person like this woman something new and
entirely original has to be undertaken. There are circumstances
where I think it is best to play your cards, face up, on the table, and
this is one of them. When she arrives here you will see what I
mean.”
“Do you intend to let her know that you are to take the trail after
her, and that——”
“Exactly, chief. But, wait until she arrives. You have other
business to attend to now. I will amuse myself with these books
here until Mouquin telephones. After that, he will not be long in
bringing the woman here, will he?”
“No. Very well. As you say, my friend.”
 It was an hour later when Mouquin called again and notified the
chief of the street and number to which Juno had betaken herself;
and then he received his orders to bring her to the private office of
the chief without delay. When another half hour had passed she was
ushered into the room.
Nick had felt no doubt that Juno was aware of his presence in
Paris, but he did expect her to manifest some surprise at finding him
in that office in consultation with the chief of the secret police.
But, quite on the contrary, her eyes sought him at once after she
had greeted the chief, and she turned to him with a smile and
exclaimed, quite as if she had confidently expected to see him:
“How do you do, Mr. Carter? We meet again, in a strange place,
after our last interview; is it not so?”
“I must confess that I did not expect to find Mrs. Dinwiddie in
Paris so soon,” replied the detective, rising and stepping toward her.
Then, addressing the chief, he added: “Chief, I wish to present you
to madam by her true name; a name which I fancy you have not
known—Mrs. Dinwiddie, of Virginia.”
She laughed, and with a gayety which did not appear to be
assumed.
“I am a respectable married woman now, chief,” she said. “I am,
indeed, Mrs. Ledger Dinwiddie, of Virginia, and if you are perhaps
wondering why I am here, or in Paris at all, I will enlighten you
without delay. Mr. Carter, as you doubtless know, is that famous
detective whom all America praises. He is supposed to be
exceptionally great in his class, and his profession has brought about
a strange circumstance.
“Notwithstanding the great ability of Mr. Carter, he has committed
a grievous error. He has mistaken my husband for a man who was
once a criminal, but who is now dead. The highest court in the State
of New York has adjudged that criminal to be dead and accepted the
proofs of identity offered by my husband. Yet Mr. Carter persists in
asserting that Ledger Dinwiddie is that dead criminal. Remarkable,
isn’t it?”
 The chief did not reply. He preferred not to commit himself. He
waited; but, in the meantime, he devoured the beauty of the woman
with his eyes. It was quite true that she had brought with her into
that obscure office a radiance, a fascination, and an atmosphere of
influence which affected every person there.
It was not her beauty alone; it was a certain magnetism which
seemed to shed energy around her like the particles that spring
spontaneously from radium.
As no one spoke, she continued:
“Not long ago, Mr. Carter appeared, during an evening, at my
home in Virginia. He did not say that he was a detective, then; and I
did not suspect it till after he had gone away. I remembered that Mr.
Carter possessed an international reputation, and thought it not
unlikely that he knew something about me, as you have known me
here. I followed him to New York when he returned there—and then
I followed him here.”
She paused for a moment and turned squarely toward Nick
Carter. Then she spoke directly to him.
“I have followed you here, Mr. Carter, for my own protection,” she
said, using her eyes with all the art she possessed, and lowering her
voice until it purred like the animal for which the Paris police had
named her. “The chief will tell you that I am not a criminal, and that
there is nothing against me, although many ugly things have been
said about me.
“Mr. Carter, I do not want to have all these matters discussed
over there in your country, where I have married, and am happy, so
I have come here after you to plead with you to spare me. Surely
that is not a great boon to ask at your hands. I ask you now to come
with me to my hotel so that I may tell you the story of my
chequered life—so that I may prevail upon you to become my
champion instead of my traducer. Will you go there with me?”
 CHAPTER XXI.
THE SIREN AT WORK.

“Madam,” said Nick Carter, “let us understand each other. I came

here to trace out the career you have pursued, not because I
expected to make you the victim of my researches, but because I
believed that through you I would be able to prove the identity of
Bare-Faced Jimmy Duryea. That is the reply to your request for me
to tell you why I am in Paris. I will add this: My work here is already
finished. I have found the information that I expected to find.” He
looked at his watch. “In three hours from now I shall leave for
London.”
They were seated opposite each other in the parlor of the suite
she had taken on her arrival in the city, where he had accompanied
her from the office of the chief.
Nick had placed no confidence whatever in her stated wish to
reveal her life history to him, but he had thought that she might say
or do something to betray herself, or to give him the cue he needed
in following out his plans. So he had accepted her proposal that he
should go with her.
She left her chair and crossed the door toward him. He arose as
she did so and stood facing her.
He knew her to be a dangerous woman; he believed her to be a
treacherous one; he had no doubt that just now she was a
desperate one.
To what ends she might dare to venture in this interview with
him he had no idea, but he was thoroughly on his guard. That she
would dare to attempt violence of any sort was farthest from his
thoughts, for they had gone there together with the full knowledge
of the chief of police—and that they had been trailed from the office
of the chief Nick did not for a moment doubt.
But he did expect that she would try her feminine wiles upon
him.
He knew that to be her most effective weapon. He had heard
enough to understand that she did not hesitate to make use of it
when occasion demanded.
There are women in the world who have been gifted wondrously,
and she was the personification of them all.
The word beautiful does not describe her. She was alluring. She
drew men to her as a charged magnet draws particles of steel. Once
amenable to her influence, they were apparently as powerless to
resist her as those same bits of steel are helpless under that
attraction.
She halted directly in front of him.
Her eyes, luminously bright, glowed upon him. He felt the thrill of
them. He realized that there was something more than mere
magnetism in that gaze, too. There was a quality about it that was
hypnotic. He knew that at that moment she was exerting all the
latent powers within her to bring him under the spell of her charms.
Nick Carter had anticipated something of this sort, and he was
prepared for it.
He had suggested to the chief that there were occasions when it
was well to play one’s cards face up on the table, and up to this
point he had done that very thing. He had purposely thrust himself
in this woman’s way in order that she might have all the opportunity
she desired to exert her powers of fascination—for Nick Carter
intended all along to appear to yield to them.
That was the game he had intended, from the moment her
presence in Paris was known to him, to play. He meant to let her
suppose that he was the same sort of weakling as the others had
been, who were inimical to her interests.
In a word, he meant to appear to become her willing victim.
 He realized that he had an extremely difficult part to play. He
knew what her intelligence was and that she would be as shrewd as
he in every move that she might make—unless her supreme
confidence in her own powers, so many times successful with
others, should lead her astray.
But egotism, too much self-confidence, is the rock upon which
many a one has foundered. Nick Carter believed it to be the one
which would be the undoing of this brilliant woman, who had so
successfully defied the police departments of all of Europe, and,
figuratively, snapped her fingers at them.
As she approached him across the floor he arose and faced her.
When she smiled into his eyes he compelled his own to glow with
an answering fire.
When she reached out one hand toward him, in a half pathetic,
half pleading gesture, he extended both his own and took it between
them and held it there.
“Mr. Carter,” she said.
“Yes?” he replied.
“You will spare me, won’t you?”
“Spare you? From what?”
“From the consequences of the investigation you are making.
See, I throw myself upon your mercy; I plead with you; I am
pleading with you now.”
She held his eyes with her own. Her other hand reached forward
and joined the first one resting upon both of his. They were standing
in the middle of the room. They were very close together. Juno’s
eyes were glowing strangely, and Nick, playing his part, wondered if,
after all, he had not dared too much.
She was alluring. She was fascinating. She had the power of
casting a spell, and already he was cognizant of the force of it.
Her eyes never left his face. He knew that she was exerting all
the hypnotic power she possessed to subject him to her will.
 He had no doubt that she had cultivated that power to the
utmost for years, under competent teachers, until she had become a
master in its use. The power of exerting hypnotic influence is an
attainment which is the consequence of study and practice; it is not
a gift. One may learn it just as one may learn to be a doctor, or a
lawyer, or a dentist.
Here, then, was the secret of what this woman had been able to
accomplish in her defiance of authority and in her undoing of the
men who had stood in her way in the past.
Her weapon had been hypnotism, and with it she had lured that
Russian prince, that Duc de Luvois, that Austrian, and others to their
death.
Slowly that free hand of hers stroked the backs of his.
Brightly, almost with a suggestion of living fire, her big eyes
burned into his.
He felt a tightening at his throat. There was a sensation as if a
rubber band wound tightly around his brows; but he controlled
himself. He managed to fix his mind upon the object of his presence
there, and he felt that he could resist her, even unto the end.
Her victims had not suspected this quality in her. They had been
men who had thought that she was succumbing to them rather than
they to her.
He forced his eyes to express all that she wished to show, or to
give out that lack of expression which would assure her that she was
succeeding.
All the while that they stood there, facing each other, with their
hands clasped, she kept on murmuring to him in a low voice, but
uttering words that would have been meaningless under any other
circumstances.
It was the droning of her voice, the soft cadences of it, that tided
in what she had undertaken to do. She was so accustomed to
success, so entirely unfamiliar with failure, that she could not see
that her effort was failing, and that Nick Carter was just as much the
master of himself now as when he entered the parlor with her.
 In the struggle that Nick had with himself when she first began
to attempt the exertion of her power, beads of perspiration came out
upon his brow. The effort with his own will brought to his face that
strained expression which she had expected to see there as a result
of her own influence.
Presently she drew him toward one of the large armchairs that
were in the room. She forced him gently down upon it, standing
before him, holding his eyes still, and now stroking his forehead with
her velvety touch.
Nick knew then why other men had become her willing victims.
He realized the depth of the pitfall that had been spread out for
them, and how entirely willing they had been to cast themselves into
it.
He appeared to struggle against her power for a time, and then
he permitted his eyes to close, as if he were indeed hypnotized—and
then he heard a quick sigh of satisfaction escape her.
She drew away from him. He heard her cross the room, but he
did not dare to peer at her between his lashes, lest she should be
watching and see him do it.
He knew that she sank upon a chair, and rested there for a time,
breathing heavily, as if the effort to which she had put herself had
fatigued her greatly. After a time, when she seemed entirely to have
recovered, she approached him again.
This time she spoke commandingly, as if she were ordering some
menial to do her will.
“Nicholas Carter!” she said sharply. “Answer me!”
“Yes,” he replied dully.
“You are my slave, are you not? Answer.”
“Yes. I am your slave.”
“You will do my will, as I direct? Answer.”
“I will do your will—as you direct,” he replied.
“Hereafter, when I raise my hand—so—open your eyes that you
may see me now—when I raise my hand, so, you will lose the power
of resisting me. Answer.”
“Hereafter, when you raise your hand, so, I will lose the power of
resisting you,” he repeated in a sing-song voice.
Again she walked away from him, crossing the room to one of
the windows, and this time, as the detective’s eyes were open, he
could watch her.
He saw her stand there for a time looking out upon the street,
and he heard her murmur broken sentences to herself.
“To think that this man should also succumb to me! He seemed
too strong, at first. I have overpowered him. When will it all end.”
And more to that effect; then, with a deep sigh she turned about
again and went to him.
“Nick Carter,” she said, standing in front of him.
“Yes?” he replied.
“I am about to awaken you now. You will forget that you have
slept. You will remember, only, a belief that I shall give you now.”
“Yes.”
“You will believe that you have held me tightly in your arms; that
my head has been pillowed upon your breast; that you have told me
of your love for me; that I have confessed my love for you. You will
implicitly believe all that, when you awaken.”
“Yes. I will believe all that.”
“And, when you are seemingly in your right mind, you will look
your love for me; you will feel it, too, through all your being—but
you will make no effort to demonstrate it. You will not so much as
touch me with your hands. You will be the slave to my will. You will
love me with all your strength, and you will believe that I return that
love—poor fool. And now, awake! I command it! Awake, Nick
Carter!”
But it was no part of Nick Carter’s policy to wake up just then. He
believed there was more to be learned if Juno believed him to be in
the hypnotic trance, and he realized that there would never be
another opportunity like the present one for learning things that he
wished to know about. So, instead of starting fully awake, as she
had commanded him to do, he sprang up only half alive to things
about him, seemingly; and he did the very thing which he knew
would compel her to reduce him again, as she would suppose, to the
hypnotic trance.
He seized her in his arms.
For just one instant of time she did not resist him. Then she
jerked herself away, out of his grasp, and he made no effort to
prevent her doing so. He sank backward upon his chair as if he were
again fully under her influence.
He saw that for a moment she turned her back to him, and that
she seemed to struggle with herself—and in that instant he recalled
the brief interval of hesitation on her part when he had seized her in
his arms. He thought no more of it then, although there was to
come another time when he would remember it. But that is another
story.
She turned toward him again, and between his eyelids he could
see that she was very pale. She raised her hands and made passes
over him, and he permitted himself to sink into a state which had
every appearance of being a deep, hypnotic sleep.
Having reduced him, as she supposed, to that utterly
unconscious and helpless state, she crossed again to the window
and looked out. Nick watched her furtively, suddenly possessed with
the idea that she was expecting somebody.
He saw her start, glance hurriedly toward him, then hasten from
the room.
Nick remained quietly where he was, not making a move, and in
a moment she was back again—but not alone. The identity of the
man who accompanied her, notwithstanding the disguise he wore,
was instantly apparent to the detective.
Jimmy Duryea—Bare-Faced Jimmy—stood just inside the door
beside her; Jimmy, perfectly made up to represent a man long past
middle age, and French, at that; but, Jimmy, nevertheless. Jimmy
was looking down upon the detective with an ironical smile upon his
lips, and a self-satisfied air that made the detective long to leap to
his feet and seize the fellow.
“Look at him,” said Juno, indicating Nick. “I have done what you
could not do; I captured Nick Carter. I have made him my slave. I
could command him to start for the far North pole, and he would
awake and start, nor would he turn back.”
“Then for goodness’ sake, Juno, command him to go and drown
himself in the Seine,” was the quick reply. “The world will be well rid
of him—and I will be able to live in peace. Could you do that, Juno?”
“I could.”
“And would he obey you?”
“He would.”
“Then do it. Do it, and we will go to Havre to-night and catch the
French line steamer that steams away to-morrow. We will be in New
York in six days. Do it. Do it. I have been sorry ever since I arrived
that I came here at all. It is too close to old times. I’m not healthy,
or healthful, here. I seem to feel a string around my neck, and to
see a huge knife falling from above. I am going back, anyhow,
whether you go or not, so do it, Juno; do it.”
For a moment she was silent. Then she replied:
“Go over there and stand near the window, then. Stand with your
back this way. Help me, yourself, by saying over and over to
yourself, ‘Obey! Obey! Obey! Obey!’ Will you do that, Jimmy?”
“Yes. Go ahead.”
Jimmy crossed to the window, and Nick could see that she kept
her eyes upon him as he did so. Nick dared to peep between his
eyelids toward her, and he was amazed to see that she had tiptoed
half the distance to the window behind Jimmy, and was making
passes in the air toward the man she had married; not toward Nick
Carter.
In his amazement the detective opened his eyes wider. He could
see that Juno’s whole mental effort was at that moment
concentrated upon Bare-Faced Jimmy, and he was utterly astounded
by it.
But a greater astonishment followed when he saw Jimmy’s arms
suddenly fall limply at his sides, after which the man turned slowly
around on his heels and faced Juno, every vestige of expression
gone from his face.
“Come nearer to me, Howard Drummond!” she commanded him;
and he obeyed, drawing nearer to her, keeping his eyes riveted upon
hers. Nick knew that she had accomplished this control of the man
merely by having induced him to concentrate his mind upon one
thing when she asked him to repeat over and over again the word,
obey. But why had she done it? Nick was soon to know.
For a moment she sank back against a chair, half exhausted, but
keeping her eyes upon Jimmy. Then Nick heard her muttering words
to herself, and yet toward Jimmy. They were:
“You had the thought; fulfill it. You spoke of the Seine; go there.
You talked of drowning; drown yourself.”
Suddenly she wheeled toward the detective, who managed to
close his eyes again before she discovered that they had been
unclosed.
“And you, Nick Carter,” she said, half fiercely, “go with him. Jump
into the river with him. Seize him in your arms and hold him so that
neither of you can unclasp from that embrace. Leap into the river
together; drown together. Go, go, go, go! I will be well rid of both of
you!”
Nick was so amazed by the turn that affairs had taken that he did
not move, although Jimmy turned obediently toward the door. She
cried out at him again, “Go, go, go!” and he pretended to obey.
He started to his feet and moved toward the door after Jimmy,
who was already passing the threshold. Juno darted after Nick,
seized him, held him for an instant, pulled his head partly around,
and whispered another command into his ear.
“Jump into the river with him, but do not drown yourself,” she
commanded in a whisper. “See to it that he dies beneath the water,
but save yourself. Save yourself, Nick Carter, and then—return here
to me. I have other work for you. Obey me.”
She thrust him through the now open doorway. She closed the
door upon him. Jimmy, under her influence, was already halfway to
the street. For a moment the detective hesitated. Then, realizing
that she had ordered him to return there, he followed slowly after
Jimmy, not doubting that he would find Juno when he did choose to
return, and realizing that he must keep this other man from
throwing himself into the river Seine.
The hour was late in the afternoon. Darkness was falling.
They passed up the street and Nick seized Jimmy by one arm
and guided him around the first corner they came to—and there, as
he had half suspected would be the case, he encountered the
detective, Mouquin, who had taken Juno to the office of the chief.
Mouquin had been sent after Nick to keep watch over him, for the
chief of the secret police of Paris feared “The Leopard.”
“Mouquin,” said the detective sternly, “you know that I am
temporarily in authority over you. Here is the badge of authority
given to me by the chief. Here, also, is some money; sufficient for
your needs. Now, listen to my orders, and obey them, literally.”
“At your service,” was the calm reply.
“This man here”—Nick had seized Jimmy by the arm and was
holding him—“is under a hypnotic trance, or spell. He has been
ordered to jump into the Seine, but you must lead him to the river
Thames, instead; to London. All rivers will be the same to him, and
that one will do as well as another. Speak soothingly to him; tell him
that you are taking him to the Seine, and he will go with you quietly
enough. Wait for me in London at Gray’s hotel, in Dover Street. I
may be there as soon as you are.”
“Am I to put him into the river? Did you mean that?” asked
Mouquin.
“No. Wait for me at Gray’s. Go, now. If you have to address the
man by name, call him Jimmy. Can you pronounce it? Good. Go,
now.”